WO2012057198A1 - 画像処理装置およびコンピュータプログラム - Google Patents
画像処理装置およびコンピュータプログラム Download PDFInfo
- Publication number
- WO2012057198A1 WO2012057198A1 PCT/JP2011/074662 JP2011074662W WO2012057198A1 WO 2012057198 A1 WO2012057198 A1 WO 2012057198A1 JP 2011074662 W JP2011074662 W JP 2011074662W WO 2012057198 A1 WO2012057198 A1 WO 2012057198A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- pattern
- image
- scum
- edge
- data
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 76
- 238000004590 computer program Methods 0.000 title claims abstract description 20
- 238000007689 inspection Methods 0.000 claims description 94
- 238000000034 method Methods 0.000 claims description 80
- 238000013461 design Methods 0.000 claims description 67
- 238000001514 detection method Methods 0.000 claims description 62
- 230000008569 process Effects 0.000 claims description 36
- 239000004065 semiconductor Substances 0.000 claims description 15
- 238000003860 storage Methods 0.000 claims description 15
- 238000011156 evaluation Methods 0.000 claims description 14
- 239000002245 particle Substances 0.000 claims description 5
- 238000004088 simulation Methods 0.000 claims description 4
- 230000002950 deficient Effects 0.000 claims 6
- 230000007547 defect Effects 0.000 abstract description 32
- 238000005259 measurement Methods 0.000 description 55
- 238000010586 diagram Methods 0.000 description 21
- 238000010894 electron beam technology Methods 0.000 description 21
- 238000001878 scanning electron micrograph Methods 0.000 description 21
- 238000004519 manufacturing process Methods 0.000 description 6
- 230000003746 surface roughness Effects 0.000 description 6
- 238000003384 imaging method Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000004140 cleaning Methods 0.000 description 3
- 239000002131 composite material Substances 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000007788 liquid Substances 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000001364 causal effect Effects 0.000 description 2
- 238000003708 edge detection Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000013041 optical simulation Methods 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 238000000059 patterning Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000010884 ion-beam technique Methods 0.000 description 1
- 238000001000 micrograph Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000007261 regionalization Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 239000010409 thin film Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B15/00—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons
- G01B15/04—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons for measuring contours or curvatures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
- H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
- H01J37/02—Details
- H01J37/22—Optical, image processing or photographic arrangements associated with the tube
- H01J37/222—Image processing arrangements associated with the tube
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
- H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
- H01J37/26—Electron or ion microscopes; Electron or ion diffraction tubes
- H01J37/28—Electron or ion microscopes; Electron or ion diffraction tubes with scanning beams
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30141—Printed circuit board [PCB]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30152—Solder
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
- H01J2237/00—Discharge tubes exposing object to beam, e.g. for analysis treatment, etching, imaging
- H01J2237/245—Detection characterised by the variable being measured
- H01J2237/24571—Measurements of non-electric or non-magnetic variables
- H01J2237/24578—Spatial variables, e.g. position, distance
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
- H01J2237/00—Discharge tubes exposing object to beam, e.g. for analysis treatment, etching, imaging
- H01J2237/245—Detection characterised by the variable being measured
- H01J2237/24592—Inspection and quality control of devices
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
- H01J2237/00—Discharge tubes exposing object to beam, e.g. for analysis treatment, etching, imaging
- H01J2237/26—Electron or ion microscopes
- H01J2237/28—Scanning microscopes
- H01J2237/2813—Scanning microscopes characterised by the application
- H01J2237/2817—Pattern inspection
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
- H01L22/10—Measuring as part of the manufacturing process
- H01L22/12—Measuring as part of the manufacturing process for structural parameters, e.g. thickness, line width, refractive index, temperature, warp, bond strength, defects, optical inspection, electrical measurement of structural dimensions, metallurgic measurement of diffusions
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2924/00—Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
- H01L2924/0001—Technical content checked by a classifier
- H01L2924/0002—Not covered by any one of groups H01L24/00, H01L24/00 and H01L2224/00
Definitions
- the present invention relates to an image processing apparatus that executes image processing of image data obtained by a charged particle beam apparatus or the like, and a computer program that realizes the image processing, and in particular, based on the obtained image data,
- the present invention relates to an image processing apparatus for detecting defects and a computer program.
- Scum is a typical defect caused by the manufacturing process.
- the scum is a thin film resist remaining at the bottom of the resist pattern after the development process. When scum is generated, the scum portion is not etched and the pattern may be short-circuited. Therefore, it is necessary to inspect the position where the scum occurs and the frequency of occurrence depending on the process conditions.
- a defect when detecting a defect, there is a method of comparing the shape of a normal circuit pattern (hereinafter referred to as a reference pattern) in which no defect has occurred and a circuit pattern to be inspected (hereinafter referred to as an inspection pattern).
- a reference pattern a circuit pattern to be inspected
- an inspection pattern a circuit pattern to be inspected
- an operator selects a circuit pattern having an ideal shape from among circuit patterns formed on a wafer, and the reference pattern image is created by photographing the circuit pattern.
- an inspection pattern is photographed, and the luminance difference between images is calculated by superimposing the inspection pattern and the reference pattern.
- Citation 1 discloses a technique for detecting a defect by comparing a reference pattern image formed based on design data with an inspection pattern.
- the luminance at the scum position is different from the luminance of the reference pattern image, so the difference value of the luminance is larger than that at the non-scum position. Using this property, it is conceivable to detect a point on an image having a large luminance difference value as a scum.
- an image processing apparatus and a computer program for acquiring image data and detecting edge branch points branched in at least three directions from the image data are proposed.
- the flowchart explaining the process of extracting an edge branch point from image data The figure explaining an example of the measurement of a pattern containing a scanning electron microscope, or an inspection system. The figure explaining an example of a scanning electron microscope image. The figure explaining the example which set the inspection coordinate on design data. The figure explaining the example which performed the pattern matching between design data and image data.
- the figure explaining the example of a display of output data The flowchart explaining the process of performing pattern matching, after patterning pattern data.
- the figure explaining the detection process of an edge branch point The figure which shows the template group used for the detection of an edge branch point.
- the figure explaining the example which performs a scum detection using two outline data The flowchart explaining the process of specifying the groove part of a pattern by matching with the outline data of a pattern, and design data, and detecting a scum.
- the flowchart explaining the process of detecting a scum using a backscattered electron image The figure explaining the example of the fall of a pattern.
- an image processing apparatus that mainly acquires image data of an electronic device, detects edge points branched in at least three directions from the image data, and generates information about the detected edge points; And a computer program is proposed.
- coordinate data of the edge point can be given.
- number of edge points is mentioned as another example of the information regarding edge points.
- the inspection area of the image data corresponding to the coordinate data of the electronic device in which the occurrence of a defect is estimated by design-based layout verification or model-based layout verification is specified, and is limited to the inspection area. Then, an image processing apparatus and a computer program for generating information related to the edge point are proposed.
- the inspection data may be classified based on the design layout data corresponding to the circuit pattern close to the edge point.
- the inspection data based on the edge point of the image data may be generated based on the coordinate data of the electronic device to generate a data file in which the inspection data is recorded.
- data for displaying information on edge points on the screen may be generated.
- the target of screen display may be the data file described above.
- an image processing apparatus mounted on a scanning electron microscope (SEM) or an image processing apparatus connected to the SEM via a communication line or the like will be described as an example.
- the present invention is not limited to this, and a process as described later may be performed using a general-purpose arithmetic device that executes image processing by a computer program.
- the method described later can also be applied to the case where defect detection is performed based on an acquired image by another charged particle beam apparatus such as a focused ion beam (FIB) apparatus.
- FIB focused ion beam
- edge points branched in at least three directions from image data obtained by photographing an electronic device by detecting edge points branched in at least three directions from image data obtained by photographing an electronic device, and generating inspection data based on the edge points, it is possible to perform inspection more than inspection using a reference pattern. Accurate inspection can be performed with high throughput.
- FIG. 2 is a schematic configuration diagram of a semiconductor inspection system including an SEM.
- the semiconductor inspection system includes an SEM 201 that acquires circuit pattern image data and a control device 214 that detects defects by analyzing the image data.
- the SEM 201 irradiates a sample 203 such as a wafer on which an electronic device is manufactured with an electron beam 202, captures electrons emitted from the sample 203 with the secondary electron detector 204 and the reflected electron detectors 205 and 206, and performs A / D
- the converter 207 converts it into a digital signal.
- the digital signal is input to the control device 214 and stored in the memory 208.
- the image processing hardware 210 such as the CPU 209, ASIC, or FPGA performs image processing according to the purpose, and a circuit pattern defect such as scum is inspected.
- the control device 214 also has a function of creating a line profile based on the detection signal and measuring a dimension between peaks of the profile. That is, the inspection system illustrated in FIG. 2 also functions as a semiconductor device measurement apparatus.
- control device 214 is connected to a display device 211 having an input means, and has a function such as a GUI (Graphical User Interface) that displays images, inspection results, and the like to the user.
- GUI Graphic User Interface
- control in the control device 214 can be assigned to a CPU or an electronic computer equipped with a memory capable of storing images and processed and controlled.
- control device 214 manually performs an imaging recipe including inspection coordinates of an electronic device required for inspection, a pattern matching template used for inspection positioning, imaging conditions, or the like, or uses design data 213 of the electronic device. Connected to the recipe creation device 212 created via the network or bus.
- FIG. 7 is a flowchart illustrating a process of creating a recipe, which is an operation program of the SEM 201, using the recipe creation device 212.
- the inspection conditions are the imaging magnification of the SEM 201, the coordinate information of inspection points, and the like.
- the user refers to the design data, and uses the coordinate information 400 of the sample on which the abutting circuit pattern is manufactured.
- Set the inspection conditions It is also possible to set coordinate information in which a defect such as a scum detected by a model base check using a design rule check or optical simulation performed at the time of semiconductor circuit design is predicted.
- a shooting recipe is generated (step 702).
- the imaging recipe is data for controlling the SEM 201, and a template for specifying an inspection point from coordinate information, magnification, and image data is defined.
- a circuit pattern is photographed by the SEM 201 based on the recipe (step 703).
- pattern matching between template data such as design data and simulation data and image data is performed to specify inspection coordinates of image data corresponding to coordinate information registered in the imaging recipe (step 704).
- Figure 5 shows an example of pattern matching results.
- the inspection coordinates 500 of the image data corresponding to the coordinate information 400 specified as the inspection conditions by pattern matching can be specified.
- the evaluation area 501 for detecting edge points branched in three or more directions is set to be large with respect to the area (edge branch point evaluation area) 501 centered on the inspection coordinates 500 of the image data. As a result, a desired inspection point can be inspected even when pattern matching shift or circuit pattern deformation occurs.
- the inspection area is located in a background area where scum as shown in FIG. 5 does not normally occur, and noise included in the image is erroneously contoured. It is possible to prevent detection of a pseudo defect due to such a pseudo contour 503. Furthermore, the inspection time can be shortened compared with the case where the entire image data is inspected.
- FIG. 1 is a flowchart showing a process of detecting edge points branched in three or more directions. Such processing can be executed by software processing using the CPU 209 of the control device 214, the memory 208, and the like. However, the image data from the SEM 201 can also be executed by software processing using a CPU, memory, and the like of an electronic computer that can input image data via a LAN or bus, or via a storage medium such as a portable memory or a hard disk. Also, part or all of the flowchart shown in FIG. 1 can be executed by the image processing hardware 210.
- FIG. 3 shows an example of an image of a circuit pattern.
- This image includes three circuit patterns 301-303.
- a scum 304 exists between the circuit pattern 301 and the circuit pattern 302
- a scum 305 exists also between the circuit pattern 301 and the circuit pattern 303. Since a large amount of electrons are emitted by electron beam irradiation at the edge portion of the circuit pattern and the edge portion of the defect, the image is brighter than the background region.
- a branch point of a white edge is detected (step 102).
- the edge branch point is a portion where the circuit pattern and the scum are connected.
- the edge branch point is an intersection of the circuit pattern 301 and the scum 304 (edge branch point 306).
- the edge branch point can be detected by a procedure as shown in FIG.
- edge branch point detection procedure An example of the edge branch point detection procedure will be described using the image data shown in FIG.
- filtering computer image processing, Ohm, p. 110
- a smoothing operator step 901
- the smoothed image data is filtered by a line detection operator (computer image processing, Ohm, p. 200) to emphasize edges (step 902). Since noise is superimposed on the image data obtained from the SEM 201, image data ((2) in FIG. 10) in which the circuit pattern and the edge of the scum are emphasized can be generated by performing such filtering processing. it can.
- binarization processing (computer image processing, Ohmsha, p.138-142) is performed on the image data after edge enhancement using a fixed threshold value or a threshold value determined by image analysis (step 903).
- binary image data ((3) in FIG. 10) in which the edge portion is a white pixel is generated.
- contour data ((4) in FIG. 10) indicating the edge as an edge having a width of one pixel is generated.
- pattern matching between the edge branch pattern and the contour data is performed to detect an edge branch point (step 905).
- Pattern matching is a technique for searching for a place where a pattern having a specific shape that can be regarded as the same as a reference image called a template exists.
- the edge branch pattern is, for example, a binary image template showing the circuit pattern shown in FIG. 11 and the contour shape of the connection part of the defect (only a representative example is shown).
- An enlarged view of the circuit pattern 1000 of the contour data ((4) in FIG. 10) and the connection area 1002 of the scum 1001 is shown in 1003.
- a 3 ⁇ 3 pixel region 1005 centered on the edge branch point 1004 matches the edge branch pattern ((1) in FIG. 11).
- the edge branch point included in the contour data can be detected by using a plurality of edge branch patterns that match the contour shape of the edge branch point and performing pattern matching with the contour data.
- the outline of (4) of FIG. 10 is expressed using a black line so that it can be easily recognized, when the template illustrated in FIG. 11 is used, the outline is expressed in white. And good.
- FIG. 24 is a diagram showing a pattern of edge branch points.
- B and d in the figure are scum generation points.
- an edge pixel position black rectangular portion
- the number of pixels of the edge existing in the 3 ⁇ 3 pixel centered on the target pixel is counted.
- the target pixel is determined as an edge branch point.
- a, b, c, and d are points at which three or more edge pixels exist in 3 ⁇ 3 pixels centering on the own pixel.
- edge branch point detection method compared to the case of using the template described above, a point where a scum and a pattern are complicated, such as a and c, is detected as an edge branch point, but the scum occurrence position is roughly determined. In this case, high-speed processing is possible and effective compared to the case of using a template.
- FIG. 8 shows an example of inspection data output to the display device 211.
- This is a screen on which a map 801 showing a wafer on which an electronic device is manufactured and the number of edge branching points 802 detected at points corresponding to inspection coordinates on the map 801 are drawn.
- the number of templates to be prepared can be suppressed by detecting edge branch points using templates after the SEM image is contoured.
- the contour line simplifies the shape of the scum part, so it is not necessary to prepare many templates to deal with complex edge shapes, and simple shapes can be matched with each other. Therefore, it becomes possible to identify the scum portion with high accuracy.
- design data of the electronic device to be inspected and process simulation data by optical simulation hereinafter, described as design data in order to simplify the text
- circuit pattern type information obtained by the analysis By classifying the inspection results, an inspection method that enables specification of a circuit pattern type that is likely to generate scum is shown.
- FIG. 6 shows a flowchart of this embodiment.
- the inspection shown in the first embodiment is performed (step 601).
- the shape of the design data used for pattern matching is analyzed, and the shape of the tip portion (a) of the line, the corner portion (b), the straight portion (c) of the line, etc. is different as shown in FIG. Classification into parts (step 602). Since the design data is generally data of a line coordinate group for expressing the circuit shape with lines, the above classification can be easily performed by analyzing the relationship between a plurality of continuous line segments. .
- a part of design data close to the detected edge branch point is specified at the position of superposition obtained by pattern matching, and inspection data is classified for each type of part (step 603).
- the classified inspection data is output to the display device 211 or the memory 208.
- the classified inspection data is displayed so that the number of edge branch points 802 shown in FIG. 8 is displayed for each circuit pattern type.
- the scum generated at the tip of the line and the scum generated at the straight line portion of the line can be detected as separate defects, so that the operator can easily find a circuit pattern portion where scum is likely to occur. Can be specified.
- an edge is detected from image data obtained by photographing an electronic device, an edge point branched from at least three directions from the edge is detected, and inspection data based on the edge point is generated. Therefore, it is possible to perform inspection with higher throughput and accuracy than the method of inspecting defects using a reference pattern.
- FIG. 13 is a diagram for explaining an example in which a library is provided for each template having a different shape, and the types of scum are classified according to the library to which the template matched with the scum included in the contour line data belongs.
- the library A has a template for detecting an edge branch point when the contour line corresponding to the pattern edge is formed in the X direction and the scum is formed in the Y direction and extending downward. It is registered.
- the library B a template for detecting an edge branch point when the contour line corresponding to the pattern edge is formed in the X direction and the scum is formed in the Y direction and extending upward is registered.
- the scum specified by the template included in the library A is formed, for example, at the upper pattern end of the abutting pattern illustrated in FIG. Further, it is conceivable that the scum specified by the template included in the library B is formed at the pattern end of the lower pattern of the abutting pattern.
- the scum type can be classified according to the library type. According to such a method, scum can be classified without referring to design data. Further, more detailed classification may be performed by using the classification based on the design data and the classification according to the type of the library.
- FIG. 14 is a diagram illustrating an example in which scum detection is performed on contour data in which a plurality of abutting patterns are arranged.
- FIG. 15 is a flowchart for explaining a process of detecting scum using templates stored in a plurality of types of libraries.
- an SEM image is acquired (step 1501), and outline data 1402 is created based on association with design data 1401 (step 1502).
- four scums 1403, 1404, 1405, 1406 are generated.
- a template is read from a certain library (step 1503), and matching processing is executed (step 1504).
- matching processing is executed (step 1504).
- the template stored in the other library is read and the matching process is executed.
- the matching locations 1407 and 1408 are specified by the template registered in the library A
- the matching locations 1409 and 1410 are specified by the template registered in the library B
- the matching locations 1411 and 1412 are registered in the library C.
- An example in which matching points 1413 and 1414 are specified by a template registered in the library D is described.
- the template registered in the library A detects a scum extending downward in the drawing
- the template registered in the library B detects a scum extending upward in the drawing
- the library C The template registered in is for detecting a scum extending toward the right side of the drawing
- the template registered in the library D is for detecting a scum extending toward the left side of the drawing.
- the scum type can be determined according to the library type.
- the scum identified as described above is stored in a predetermined database together with information related to the scum such as identification information, coordinate information, or information on the type of scum corresponding to the type of library (step 1505).
- a step of setting a predetermined area 1415 in relation to the template and determining whether or not another matching portion is included in the predetermined area 1415 is provided. May be. For example, when a plurality of matching locations exist in the predetermined area 1415 and a plurality of templates specifying the plurality of matching locations satisfy a predetermined relationship, the matching locations are connected by the same scum, etc. It becomes possible to make a judgment.
- the inspection area 1416 is superimposed on the design data 1401, and the design data 1401 and the outline data 1402 are overlapped by pattern matching or the like, so that the inspection area set on the design data is placed at an appropriate position on the outline data. It becomes possible to set.
- FIG. 16 is a diagram for explaining an outline of a database for storing scum identification information (ID) and information related to scum in association with each other.
- ID scum identification information
- Information information related to scum in association with each other.
- coordinate data for example, position information obtained by matching with the position on the contour line data or design data is registered.
- pattern data identification information of a pattern in which a scum has occurred is registered.
- the pattern shape (Pattern Shape) is registered with information on the type of library described above and the pattern shape of the place where the scum is generated based on the design data.
- the pattern line segment identification information includes the identification information of the part for each partial area of the design data pattern, and pattern matching between the design data and the contour line data is performed.
- the partial area of the contour line data belonging to each partial area of the design data can be obtained. Further, as illustrated in FIG. 23, matching processing is performed between the design data 2301 and the contour line data 2302, and then correspondence is performed between the line segment of the design data 2301 and the line segment of the contour line data 2302. By attaching, identification information (line segment information) may be added to each line segment of the contour line data 2302.
- T represents an upper part
- B represents a lower part
- L represents a left part
- R represents a right part
- L1 to L6 represent line segment numbers.
- the line segment included in the region 2303 of the contour line data 2302 corresponds to the line segment of the design data to which the identification information “L2” is attached.
- the line segment of the contour line data included in the region 2304 corresponds to the line segment of the design data to which the identification information “L5” is attached, it is conceivable to add the identification information “L5”. Of course, other identification information may be added based on a predetermined rule.
- the closest position on the contour line data 2302 is specified from a predetermined position on the design data 2301, and that position is used as a corresponding point.
- the library type described above and the scum identification information of the template are registered in the Scum Type (Scum Pattern) column. Further, in the “Connect” column, when a plurality of matching locations are included in the predetermined area 1415 of FIG. 14, it is determined that the plurality of matching locations are connected by scum, and information indicating that is provided. sign up.
- the data stored in these databases does not necessarily require all data, and information necessary for performing scum evaluation as described later may be registered.
- FIG. 17 is a diagram for explaining the outline of an arithmetic unit that detects edge branch points.
- an SEM image acquisition unit 1701 for acquiring image data from the circuit pattern 401 and the like
- an outline extraction unit 1702 for extracting outline information from the image data
- pattern matching using a predetermined template a predetermined template.
- a storage unit 1706 for storing information is provided.
- the identification information (ID), coordinate data (Address), part information of the pattern in which the scum has occurred, the type of library in which the template specifying the edge branch point is stored, and other scum Is added, or information regarding the number of scums in the contour line data is added and stored in a table format as exemplified in FIG. Further, the data folder prepared for each scum type may be classified and stored according to the scum type.
- a storage medium is built in the arithmetic device, and the arithmetic unit is configured to be accessible to the storage unit as necessary.
- the storage medium is installed outside and necessary. Depending on the situation, the arithmetic unit may access.
- FIG. 19 is a diagram for explaining an output example of the statistical value of the scum detection result.
- FIG. 19A shows a graph in which the interval between patterns is on the horizontal axis and the frequency of occurrence of scum is on the vertical axis.
- the inter-pattern interval information may be obtained from design data, or may be obtained from actual measurement values as will be described later. By performing such display, it is possible to determine how much the inter-pattern spacing contributes to scum generation.
- FIG. 19B illustrates a graph in which the vertical axis represents the occurrence frequency and the horizontal axis represents the pattern shape.
- the pattern shape include a line end, a corner, and a straight line portion.
- a composite condition of a plurality of patterns for example, an abutting pattern, a portion where the line end and the corner are close to each other, a portion where the straight line portion and the line end are close to each other, or the like
- the bar graph may be divided and displayed in accordance with the distance between the patterns in order to display the occurrence frequency for each distance between the patterns.
- (C) in FIG. 19 shows a distribution diagram of the occurrence frequency for each pattern direction.
- the scum generation direction and the scum length are displayed in a distributed manner. According to such a display form, it is possible to confirm the relationship between the scum generation direction and the scum length.
- the scum distance is replaced with the inter-pattern dimension, the relationship between the inter-pattern spacing and the occurrence of scum can be grasped.
- the information on the scum generation direction is extracted based on the determination as to where the generated scum is present in the pattern, for example.
- FIG. 19D is a distribution diagram showing the relationship between the scum and the scum length specified by the template capable of specifying the scum direction described in FIG. 13 and the like.
- the scum direction is specified using a plurality of templates capable of specifying four scum directions, and the occurrence frequency in each direction is shown. In the case of this example, it can be determined that scum extending in the vertical direction is relatively frequent.
- FIG. 20 is a block diagram for explaining the outline of the image processing apparatus 2001.
- An image processing apparatus 2001 includes a data input interface 2002 for inputting data from design data 2004, contour data 2005, and template library 2006 stored in an external storage medium, and design data 2004 and contour data 2005.
- a pattern identification information adding unit 2009 for adding a pattern and identification information of each part of the pattern based on the matching.
- the pattern identification information adding unit 2009 performs an operation of adding the pattern identification information included in the design data 2004 and the pattern part identification information to the contour line data.
- the scum detection unit 2008 performs scum detection on the contour line data 2005 using the edge branch point detection template stored in the template library 2006.
- the scum detected in this way is classified by the scum classification unit 2010 based on the identification information added by the pattern identification information addition unit 2009 and / or the identification information stored in the template library 2006. Is done.
- the scum related information, contour line data, and design data are converted into a predetermined data format by the output data generation unit 2012 and stored in the database 2011 via the data output interface 2003.
- FIG. 18 is a diagram for explaining an example of measuring a dimension between edge branch points based on detection of scum.
- FIG. 18A illustrates an example in which scum is generated between a plurality of line patterns.
- a scum 1803 is generated between the first outline 1801 and the second outline 1802.
- a length measurement point 1804 as a length measurement start point and a length measurement end point
- such an edge branch point can selectively select a point where scum has occurred.
- the selection of the measurement start point and measurement end point based on the extraction of the edge branch point selects the part corresponding to the defect as the measurement target, so the measurement recipe creation can be made more efficient and automated. It can be realized.
- FIG. 18 is a diagram for explaining an example in which scum is generated between line ends, and is a diagram for explaining an example in which a length measurement location 1805 is selectively set at a location where scum has occurred. In this way, efficient measurement can be performed by selectively setting a place where the scum is generated as a measurement target.
- FIG. 18 is a figure explaining the example in which the scum 1806 generate
- the first edge branch point 1807 and the second edge branch point 1808 are set as the length measurement start point and the length measurement end point.
- the measurement object can be the interval in the X direction ( ⁇ x), the interval in the Y direction ( ⁇ y), the interval in the two-dimensional direction ( ⁇ ), or the length of the scum 1806 itself between the two branch points.
- FIG. 22 is a flowchart illustrating a process of automatically setting a measurement location based on detection of an edge branch point. Steps 2201 to 2204 are the same as the edge branch point detection step. Step 2205 is a step of extracting a length measurement point from the detected edge branch point. At this time, the above-described pattern part classification information is used in order to measure between related edge branch points.
- FIG. 21 is a diagram for explaining a method of adding pattern part identification information to the contour lines 2103 and 2104.
- FIG. 21A illustrates L-shaped pattern and line pattern design data 2101 and 2102. Identification information for identifying a part of the pattern is added to each of the design data 2101 and 2102.
- “A” of “Aa1” is an identification code of a pattern
- “a” is an identification code of a pattern part type
- “1” is a plurality of pattern parts of the same type. It is an identification code given to each part.
- “A” indicates an L-shaped pattern
- “B” indicates a line pattern.
- “A” indicates a line end
- “b” indicates a corner
- “c” indicates a straight line portion.
- FIG. 21 shows a state in which pattern matching is executed between the design data to which the identification information as described above is added and the contour line data 2103 and 2104, and the pattern part identification information is added to the contour line data 2103 and 2104. Illustrated in (b). By adding identification information to contour lines belonging to each identification region, the contour line has unique information for each part.
- both combinations of pattern parts considered to be connected between edges by scum (in this example, for example, “Ab2” of the contour line data 2105 and “Ba1” of the contour line data 2104)
- scum in this example, for example, “Ab2” of the contour line data 2105 and “Ba1” of the contour line data 2104)
- step 2205 of FIG. 22 the length measurement start point and end point are set based on the above-described rules and stored as a measurement recipe (step 2206).
- FIG. 25 shows scum detection using profile edges.
- FIG. 25 shows a line pattern in which a scum 2500 exists in the lower part of the pattern edge (hereinafter referred to as pattern edge lower part) 2503, a cross-sectional shape of the line pattern (hereinafter referred to as cross-sectional profile) 2504, and an image of the line pattern corresponding to the cross section.
- pattern edge lower part a scum 2500 exists in the lower part of the pattern edge
- cross-sectional profile a cross-sectional shape of the line pattern
- image profile design data 2501 based on the result of pattern matching
- the scum 2506 existing below the pattern edge also causes a fatal defect, and therefore needs to be detected.
- the profiles with and without scum are shown.
- the broken line shows the profile of the pattern without scum, and the solid line shows the profile of the pattern with scum.
- an image signal 2508 (hereinafter referred to as the top edge) 2508 and pattern edge lower part 2503 corresponding to the upper part of the pattern edge (hereinafter referred to as the upper part of the pattern edge) 2502 are analyzed by analyzing the image profile 2507
- Corresponding image signals (hereinafter referred to as under edges) 2506 are individually extracted.
- contour data composed of the top edge 2508 and contour data composed of the under edge are generated, and the presence or absence of scum is detected by comparing the shapes of the contour data.
- a shape difference occurs between the contour data composed of the top edge and the contour data composed of the under edge, and therefore, a pattern having a large shape difference is detected as a scum pattern.
- FIG. 26 shows a cross-sectional profile 2601 and an image profile 2608 of a part of the line pattern.
- the pattern area 2600 indicates a groove area of a line pattern in which scum is generated, and is designated by the user as area information for detecting scum. Further, the groove region may be set based on the result of pattern matching between design data having information on the groove and non-groove and image data.
- the peak signal 2603 of the image profile 2608 near the boundary of the groove region 2600 is obtained, and the peak position 2607 is obtained. This point is set as the position of the top edge corresponding to the upper part of the pattern edge.
- the signal difference h1 between the peak signal 2603 and the bottom peak signal 2605 is calculated.
- a point 2606 of the image profile 2608 corresponding to the threshold value 2604 corresponding to the range between the peak 2603 and the bottom peak 2605 is obtained. This point is set as the position of the under edge corresponding to the lower part of the pattern edge.
- the above top edge and under edge are detected for the image area to be scum detected.
- the reference direction of the image profile for edge detection may be specified by the operator for each point for detecting the edge point, or may be set based on the continuous direction of the design pattern used for pattern matching.
- top edge contour data hereinafter, upper contour data
- under edge contour data hereinafter, lower contour data
- FIG. 27 shows an example of the upper contour data 2701 and the lower contour data 2702.
- the interval (error) 2703 between the upper contour data 2701 and the lower contour data 2702 is comprehensively measured for the scum detection target region 2704.
- the measurement interval can be set arbitrarily.
- the threshold processing of the index is performed based on the statistical amount of error between the upper contour data and the lower contour data, and patterns above the threshold are detected as scum patterns.
- scum detection may be selectively performed for the portion.
- the scum detection is performed and the variation or the like is equal to or less than a predetermined value, the detected scum is not scum (for example, noise may be detected), and the scum detection result is filtered.
- the image signal of the pattern groove portion where the scum is present has a larger change than the image signal of the pattern groove portion where the scum is not present. For this reason, the pattern groove portion is specified, the variation of the image signal in the region is calculated by the variance, the standard deviation, etc., and the presence or absence of scum can be determined from the size.
- FIG. 28 shows contour data 2801 at the upper part of the pattern edge and contour data 2802 at the lower part of the pattern edge in the scum detection using the luminance distribution.
- a scum 2803 is generated in a region surrounded by the contour data 2801 above the pattern edge and the contour data 2802 below the pattern edge. For this reason, the scum evaluation area 2804 is limited by using the pattern edge upper portion 2801 and the pattern edge lower contour data obtained by the pattern edge detection described with reference to FIG.
- the evaluation area can be limited by using either contour data below the pattern edge or contour data inside the pattern edge, or may be set by the operator. Furthermore, the evaluation area may be limited based on the result of pattern matching between design data (or simulation data) having information on groove portions and non-groove portions and image data.
- a statistic indicating the variation of the image signal such as the variance of the image data and the standard deviation corresponding to the evaluation area 2804 is calculated.
- threshold processing is performed on the statistic indicating the variation of the image signal, and a pattern equal to or higher than the threshold is detected as a pattern having scum.
- the scum detection may be selectively performed on the portion, or the variation or the like may be predetermined after the scum detection is performed.
- the value is equal to or less than the value, the detected scum is not scum (for example, noise may be detected), and the scum detection result may be filtered.
- the causal relationship between the degree of divergence between contour lines and the occurrence of scum can be obtained. It is also possible to specify.
- FIG. 36A is a diagram showing a pattern arranged normally
- FIG. 36B is a diagram showing a pattern collapse.
- the pattern collapse may occur because forces (surface tension) attracting each other between patterns due to the use of a cleaning liquid or the like.
- FIG. 36 (c) is a top view of a pattern having a local pattern collapse at the top.
- step 3701 After an SEM image is acquired by the SEM 201 (step 3701), pattern upper contour data 2701 and lower contour data 2702 are respectively acquired (step 3702).
- the intervals (error values) 2703 between the upper contour data 2701 and the lower contour data 2702 are comprehensively measured for a plurality of regions in the scum detection target region 2704 along the edge of the pattern (step 3703).
- An average value and standard deviation of error values measured for each region are calculated (step 3704). Here, only one of the average value and the standard deviation may be obtained, or both may be calculated and used as evaluation targets in the subsequent steps.
- Determination is made (step 3706). If it is smaller than the threshold A, it is determined whether or not a predetermined threshold B (B ⁇ A) or more is reached (step 3707). (Step 3708). If it is smaller than the threshold value B, it is determined that detection other than scum (for example, noise) is a factor (step 3709). In this way, when the interval between the upper contour data 2701 and the lower contour data 2702 has a certain size and variation, the pattern collapses whether the cause is due to the occurrence of scum by making a determination using the threshold value. It is possible to determine whether this is due to other factors such as
- the predetermined range refers to, for example, obtaining an average value or standard deviation of the error values for each partition area defined to have a certain size, and determining whether the number of partition areas having large variations exceeds a predetermined number. It can also be determined as
- the determination process is performed using the error regarding the interval between the upper contour data 2701 and the lower contour data 2702.
- the frequency of the edge of the pattern acquired from the SEM image is analyzed. It is also possible to discriminate pattern collapse or scum. Specifically, when the frequency in a predetermined region in the SEM image is higher than a certain value, there is a locally protruding scum, and when the frequency is relatively low, it depends on the inclination of the pattern. It can be judged that it is fluctuation.
- the width of the white band acquired based on the luminance information of the pattern image is different between the left and right side walls. Therefore, as shown in FIG. It can be detected by measuring the dimensions of the profile. Specifically, the left and right side walls of the pattern are identified by superimposing the SEM image and design data, and a line profile based on luminance information is created from the SEM image in a predetermined area on each of the left and right sides, and white lines are obtained from this peak position. If the band widths are compared and a result of the comparison shows that a difference of a certain level or more is produced on the left and right, it can be determined that the pattern collapses.
- FIG. 44A is a flow showing a procedure for creating the contour line data.
- an SEM image of a pattern to be evaluated is acquired (step 4401), and a line profile is created based on the luminance information of the image (step 4402).
- the profile tends to be sharper as the inclination of the pattern 4405 increases as the signal intensity increases.
- a plurality of threshold values are set in advance for the acquired profile, and for each threshold value, a portion on the image indicating the threshold value is connected to extract a contour line (step 4403).
- a plurality of length measurement contour lines are extracted by performing this for a predetermined number of set threshold values (step 4404). Further, it is possible to omit the step 4404 and set a predetermined number of threshold lines to be extracted in advance.
- FIG. 44C shows an example (enlarged view) in which a plurality of threshold values are set in a profile and a plurality of contour lines are extracted. In the case of the example in FIG. 44C, when the peak portion of the profile is set to 100% (Th1), the contour line is extracted by changing it at a constant rate from T2 to T5.
- the threshold value can be set by various methods.
- FIG. 44D is a diagram seen from above with the contour lines extracted for each of the predetermined number of thresholds superimposed.
- the interval (error) 2703 between the upper contour data 2701 and the lower contour data 2702 is determined for the scum detection target region 2704 in the scum detection and pattern collapse determination as described above.
- step 2901 As design data, information in which information inside and outside the pattern is added by a vector or the like is used.
- design data information in which information inside and outside the pattern is added by a vector or the like is used.
- contour data indicating the edge of the pattern as a one-pixel width edge is extracted by the procedure shown in FIG. 10 (step 2902).
- step 2903 After positioning by pattern matching between design data and contour data (step 2903), the edge of the side wall is searched (step 2904), and the groove portion of the pattern is specified (step 2905).
- contouring in step 2902 may be performed.
- an inspection area is set based on the information on the groove portion of the specified pattern (step 2906).
- step 2907 pattern matching between edge branch patterns and contour data is performed (step 2907), and edge branch points are extracted (step 2908). Then, it is determined whether or not the extracted edge branch point is within the inspection region (step 2909). If it is within the region, inspection is performed as a scum generated outside the pattern (step 2910). In some cases, setting is made so that the inspection is not performed as surface roughness occurring inside the pattern (step 2911). Through the above steps, statistics of scum generated outside the pattern and surface roughness generated inside the pattern can also be obtained. Note that the determination at step 2909 can be omitted, and the condition can be set in advance so that edge branch point extraction is performed only on the inspection region as an inspection target.
- the determination can be performed by using the LR image of the sample instead of the design data used for the pattern matching process in Step 2903 of the above flow (FIG. 30).
- the contour data of both are obtained. It can be matched by using (step 3005).
- steps 3006-3012 are the same as steps 2904-2910 in FIG.
- the edge branch point is extracted for all the areas without setting the inspection area in step 2906, and the identification information as to whether or not the groove is the pattern specified from the edge information obtained by the matching process with the design data or the like Can be stored in the storage unit of the database shown in FIG. 16 in association with the extracted information related to the edge branch point, and statistical processing can be performed. By doing so, the relationship between the surface roughness of the pattern and the number of occurrences of scum can be acquired.
- FIG. 33 is a cross-sectional view and a top view showing the degree of occurrence of scum generated between the wiring patterns (grooves), and a diagram showing the relationship between the risk level of scum.
- the scum shown in this figure is an example, and naturally the form of scum generation is not limited to this.
- the scum generated in the groove portion of the wiring pattern has a form that locally protrudes to the bottom portion of the pattern at the initial stage of generation, and faces the groove portion from the adjacent pattern. Even when scum occurs, there is an interval between them (risk level 1).
- the scum generation level becomes higher than in the above case, the two become connected to each other in the vertical-horizontal direction or the height direction, and the same phenomenon is observed at multiple locations.
- risk level 4 When the generation level further increases, a scum having a certain width or height is formed (risk level 4).
- FIG. 38 shows the relationship between the degree of the width and height of the generated scum and the level of danger.
- the degree of danger here means the ease of short-circuiting the pattern, the degree of difficulty in removal, and the like. That is, a minute scum at a low generation level can be removed in the manufacturing process, but the removal becomes difficult as the level increases. Therefore, it is required to quantitatively evaluate the generated scum.
- FIG. 39 shows an example of a procedure for measuring the width of a scum generated between a plurality of line patterns.
- outline data is created (step 3902).
- the creation of the contour line data can be performed using the above-described threshold processing, association with design data, or the like.
- scum 4002, 4003, and 4004 are generated between a plurality of line patterns 4001.
- FIG. 40B shows an example of measuring the width of the generated scum.
- templates are read from a plurality of types of libraries according to the procedure shown in FIG.
- a length measurement point 4005 is set for an area not including the original pattern, that is, including the edge caused by the scum (Step 3906). At this time, preferably, the length measurement point 4005 is set in the direction in which the peak of the profile created in step 3908 can be most sharply acquired. In particular, by setting the measurement point 4005 so as to be perpendicular to the edge caused by the scum, a highly accurate measurement result can be obtained as described later.
- an electron beam is scanned perpendicularly to the edge caused by the scum at the set measurement point 4005 (step 3907), a profile is created from the detected secondary electron signal, and the dimension of the scum width is set. Measure (Steps 3908 and 3909).
- the measurement point is set for the region including the edge caused by the scum and the electron beam is scanned.
- the region is obtained based on the luminance information of the SEM image acquired in advance in step 3901 in the region.
- a profile may be created from the signal. At this time, it is desirable to acquire the luminance distribution in a direction perpendicular to the direction of the edge caused by the scum.
- the procedure for measuring the width of the scum uniformly with respect to the extracted edge branch point has been described, but the above method may be selectively used when the branch point satisfies a predetermined condition.
- a predetermined condition For example, when a plurality of edge branch points are extracted, the edge branch points detected by the features of the template are connected by a single line, that is, whether they are connected as the same scum. If there is one that is determined to form the same scum, the length measuring point 4005 is set to be perpendicular to the edge caused by the scum. In addition, when it is determined that the same scum is not formed, the scum width dimension can be excluded.
- the said conditions are not limited to this, Various conditions can be set according to a condition.
- the scum length is measured by, for example, the method shown in FIG. It is also possible to set a criterion for determination so as to be a dimension measurement target.
- An arrow 4006 in FIG. 40B indicates the scanning direction of the electron beam.
- a scum having a wide width is generated like a scum 4004 shown in FIG. 40 (a)
- two contour data are created for both end portions of the scum.
- a predetermined condition such as a distance between the created contour line data is smaller than a predetermined distance, for example, these are determined to form the same scum.
- Measure the dimensions In the case of the scum 4002 and 4003 in FIG. 40A, it is considered that when a profile is created from the luminance information of the SEM image, a single peak is generated as in the profile 4007 in FIG. 40C.
- the dimension of the scum width can be measured by measuring the distance between the start point and the end point that form the peak.
- an algorithm may be used so that the distance between the two peaks is measured as the scum width dimension.
- the method for measuring the scum width can be appropriately selected depending on the state of the generated scum.
- FIG. 41 is a diagram showing the relationship between the scanning direction 4102, 4103, and 4104 of the electron beam and the profile created from the generated secondary electron signal with respect to the contour line data 4101 representing the edge of the scum.
- the profile shapes also differ as shown in FIGS. 41A, 41 ⁇ / b> B, and 41 ⁇ / b> C.
- the scanning direction 4102 of the electron beam is 90 ° with respect to the scum outline data, that is, a direction perpendicular to the scum contour data, and a profile with the largest peak is created as shown in FIG. Can do.
- FIG. 42 it is assumed that there is a certain tendency (relative relationship) between the detected secondary electron signal intensity and the angle formed by the scanning direction of the electron beam and the scum. From this figure, it can be seen that the highest signal intensity can be obtained by scanning the electron beam in the direction perpendicular to the edge of the scum.
- the above example has the same tendency as the scanning direction of the electron beam with respect to the edge direction of the pattern.
- FIG. 43 shows an example of a variation of the method for setting the length measurement location.
- the generation of the scum 4302 as described above.
- a pattern 4301 as shown in FIG. May occur in a direction that is oblique with respect to the curve and may be deformed so as to draw a curve, or may occur between patterns located in places that are not adjacent to or facing each other as shown in FIG. .
- FIG. 43C shows a case where a plurality of generated scums 4302 are not connected.
- the measurement of the scum width dimension may be excluded from the determination of step 3906 in FIG. 39, and the length measurement portion 4303 may not be provided.
- the length of the scum 4302 is a predetermined length.
- the length measuring point 4303 can be set in the vertical direction with only the scum as a measurement target. Furthermore, using the design data, information related to the position of the pattern in advance, for example, information on whether there is a close pattern in advance is acquired in advance, and when the scum is detected, It is also possible to scan the electron beam by setting the length measurement location 4302 so as to include the proximity pattern grasped in advance. As described with reference to FIG. 33, since the scum may be connected to an adjacent pattern, the edge branching is provided by providing the length measurement location 4302 based on the positional relationship of the pattern obtained from the design data. Depending on the point extraction, only a part of the scum can be detected, and even if the part continuing to the scum cannot be detected, the scum that actually exists can be detected with high accuracy. become.
- the method for quantitatively evaluating the width of the scum has been described.
- a method for evaluating the height of the scum will be described.
- the height direction is actually It is possible to recognize scum that has not grown and has a high risk level, that is, that can be removed. Further, when only the height of the scum significantly affects the risk level, it is also effective to perform the evaluation alone after detecting the scum.
- the procedure can be applied to various scums, such as measurement of dimensions and width, with respect to the scum detected by the above-described method. For example, it can be selectively performed only when it is determined that the detected scum has a certain width.
- contour data indicating the edge of the pattern as one pixel edge is extracted by the procedure shown in FIG. 10 (step 4502).
- a template which is an edge branch point pattern is read from the above library (step 4503), matching processing is performed (step 4504), edge branch points are extracted, and scum is detected (step 4505).
- a line profile of an area including the detected scum is created (step 4506).
- an image may be acquired by setting the measurement point again in the area where the scum is detected, or only the portion of the scum detection area in the profile created from the already acquired image. May be selectively used.
- contour lines are extracted for each predetermined number of thresholds (step 4507), and all the predetermined number of threshold lines are extracted (step 4508). At this time, by setting the predetermined number of thresholds so that the height of the profile becomes a constant interval, a plurality of contour lines obtained are contour lines.
- the degree of risk having a certain height is set. It is determined that the scum is high, and is set as an inspection target (step 4510). Otherwise, it can be set as a non-inspection target (step 4511).
- an SE image secondary electron image
- a BSE image reflected electron image
- the detector when acquiring the BSE image, it is desirable to use a detector that detects electrons reflected in a direction indicated by an arrow 3401 in FIG. 34 that is parallel to the pattern. Since BSE is signal electrons having higher energy than SE, detection using a detector set in a desired position in this way is possible. In particular, in the scum generated in the groove portion of the wiring pattern described above, the detected amount of electrons reflected in the parallel direction reflects the degree of scum more than the electrons reflected in the direction perpendicular to the pattern. Therefore, the reflected electrons reflected in such a direction are selectively detected. Next, contour data is extracted from both images using the above-described method (steps 3503 and 3504).
- the SE image is aligned by matching processing with design data (step 3505), the edge of the side wall is searched (step 3506), and the groove portion of the pattern is specified (step 3507).
- the SE image and the BSE image are overlaid, and the focus, contrast, etc. are adjusted to obtain a composite image (step 3508).
- an inspection area is set with respect to the image obtained by the synthesis (step 3509), an edge branch point is extracted by matching with the edge branch pattern (step 3510) (step 3511), and a scum is detected (step 3511).
- Step 3512 According to the above procedure, it becomes easier to detect the scum more accurately than when only the SE image is used by superimposing the SE image and the BSE image.
- the scanning direction of the electron beam is set to a direction parallel to the pattern, thereby causing scum relative to the edge of the pattern.
- Many BSE can also be detected.
- scum detection using the SE image can be performed for the region.
- the BSE image is relatively easy to detect scum as compared to the SE image, it is effective to first determine the region by performing rough detection using the BSE image.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Quality & Reliability (AREA)
- Manufacturing & Machinery (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Power Engineering (AREA)
- Electromagnetism (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
- Length-Measuring Devices Using Wave Or Particle Radiation (AREA)
- Analysing Materials By The Use Of Radiation (AREA)
Abstract
Description
本発明の他の目的、特徴及び利点は添付図面に関する以下の本発明の実施例の記載から明らかになるであろう。
このように、上部輪郭データ2701と下部輪郭データ2702の間の間隔がある程度の大きさ及びばらつきを有する場合において、閾値を用いた判断を行うことによって原因がスカムの発生によるものなのか、パターン倒れのようなその他の要因によるものなのかを判断することが可能になる。
このように抽出した複数の輪郭線を用いることで、上述のようにスカム検出やパターン倒れの判別において、スカム検出の対象領域2704について上部輪郭データ2701と下部輪郭データ2702の間隔(誤差)2703を網羅的に計測することに加えて、各々の輪郭線に対して任意の輪郭線間の間隔を計測することも可能になる。
まず、SEM201によってSEM画像を取得したのちに(ステップ2901)、図10に示した手順によってパターンのエッジを1画素幅のエッジとして示す輪郭データを抽出する(ステップ2902)。次に設計データと輪郭データのパターンマッチングによって位置合わせをしたのち(ステップ2903)、側壁のエッジを探索して(ステップ2904)、パターンの溝部を特定する(ステップ2905)。なお、ここでステップ2903のマッチングを行ったのちに、ステップ2902の輪郭線化を行うようにしても良い。次に、特定したパターンの溝部の情報に基づいて、検査領域を設定する(ステップ2906)。ここで、検査領域については、図32に示すようにパターンの溝部を含むとなるように設定するのが望ましい。このように設定することによって、パターンの内側に発生する表面の荒れを検査対象から除外することができるからである。次に、エッジ分岐パターンと輪郭データのパターンマッチングを行い(ステップ2907)、エッジ分岐点を抽出する(ステップ2908)。そして、抽出したエッジ分岐点が検査領域内であるかどうかを判断し(ステップ2909)、領域内である場合には、パターンの外側に発生したスカムとして検査を行い(ステップ2910)、領域外である場合には、パターンの内側に発生した表面の荒れとして検査を行わないように設定する(ステップ2911)。以上の工程により、パターンの外側に発生したスカムと、パターンの内側に発生した表面の荒れの統計を求めることもできる。なお、このステップ2909の判断を省略し、予め検査領域のみを検査対象としてエッジ分岐点抽出を行うように条件を設定することもできる。
また、ステップ2906の検査領域を設定することなく全ての領域についてエッジ分岐点を抽出し、設計データ等とのマッチング処理によって得られたエッジ情報から特定されたパターンの溝部であるかどうかの識別情報を、抽出したエッジ分岐点に関する情報と関連付けて図16に示したデータベースの記憶部に記憶させ、統計処理を行うこともできる。このようにすることによって、パターンの表面荒れと、スカムとの発生数の関係を取得することができる。
本図に示される通り、配線パターンの溝部に発生するスカムについては、発生初期の段階ではパターンの底部に局所的に張り出したような形態をしており、隣り合うパターンから溝部において対向するようにスカムが発生した場合であっても、両者の間には間隔が存在している(危険度レベル1)。次の段階として、スカムの発生レベルが上記の場合よりも高くなり、お互いに縦-横方向、または高さ方向へ大きくなったことによって両者が接続し、さらには複数箇所において同様の現象が見られることがある(危険度レベル2-3)。そして、さらに発生レベルが高くなると、一定の幅、または高さを持ったスカムが形成される(危険度レベル4)。図38は、発生したスカムの幅、及び高さの度合いと、危険度のレベルとの関係を示す。ここでいう危険度とは、パターンのショートのしやすさや、除去の困難性の程度等を意味する。すなわち、発生レベルの低い段階における微小なスカムについては製造工程において除去が可能であるが、レベルが高くなるにつれて除去が困難になる。したがって、発生したスカムを定量的に評価することが求められる。
そこで、次に、このように発生したスカムの幅、及び高さを評価する方法について説明する。図39は、複数のラインパターン間に発生したスカムの幅を寸法測定する手順の一例を示す。
まず、SEM201によってSEM画像を取得したのちに(ステップ3901)、輪郭線データを作成する(ステップ3902)。ここで、輪郭線データの作成には、上述の閾値処理や、設計データとの対応付け等を用いて行うことができる。図40の(a)の例では、複数のラインパターン4001の間にスカム4002、4003、4004が発生している。また、図40の(b)は、発生したスカムの幅を寸法測定する例を示す。次に、図15に示した手順によって複数種類のライブラリからテンプレートを読み出し(ステップ3903)、それぞれのテンプレートを用いたマッチング処理(ステップ3904)を行って、エッジ分岐点を検出する(ステップ3905)。次に、エッジ分岐点を構成する複数のエッジのうち、本来のパターンに起因しない、すなわちスカムに起因したエッジを含んだ領域について測長個所4005を設定する(ステップ3906)。このとき、好ましくはステップ3908にて作成するプロファイルのピークを最もシャープに取得可能な方向に測長個所4005を設定する。特に、当該スカムに起因するエッジに対して垂直となるように測長個所4005を設定することにより、後述するように高精度な測長結果を得ることができる。そして、設定した測長個所4005にてスカムに起因するエッジに対して垂直に電子線の走査を行って(ステップ3907)、検出される二次電子信号からプロファイルを作成し、スカム幅の寸法を測定する(ステップ3908、3909)。
上記のフローでは、スカムに起因するエッジを含む領域について測長個所を設定し、電子線を走査しているが、当該領域においてステップ3901にて予め取得したSEM画像の輝度情報に基づいて得られる信号からプロファイルを作成するようにしても良い。このとき、当該スカムに起因するエッジの方向に垂直となる方向に輝度分布を取得することが望ましい。
ここでは抽出したエッジ分岐点に対して画一的にスカムの幅を測長する手順を説明したが、当該分岐点が所定の条件を満たす場合に選択的に上記の手法を用いるようにしてもよい。例えば、複数のエッジ分岐点が抽出された場合に、当該テンプレートの有する特徴によって検出されたエッジ分岐点が1本の線で結ばれている、すなわち同一のスカムとして接続された状態にあるかを判断し、同一のスカムを形成していると判断されたものがある場合には、当該スカムに起因するエッジに対して垂直となるように測長個所4005を設定する。また、同一のスカムを形成していないと判断された場合には、スカム幅の寸法測定の非対象とすることができる。また、当該条件はこれに限定されるものではなく、状況に応じて種々の条件を設定できる。例えば上記のように同一のスカムを形成していない場合においても、例えば図18にて示した方法によってスカムの長さを測定した結果、危険度のレベルが高いと判断された場合にはスカム幅の寸法測定対象とするように判断の基準を設けることもできる。図40の(b)の矢印4006は、電子線の走査方向を示している。上記の手順により、評価対象とするスカムの寸法を選択的に測定することができる。
また図42に示されるように、検出される二次電子信号強度と、電子線の走査方向とスカムとがなす角度との間には一定の傾向(相対関係)があることが想定される。本図から、スカムのエッジに対して垂直な方向へ電子線走査を行うことで最も大きな信号強度を得ることができることがわかる。上記の例は、パターンのエッジ方向に対する電子線の走査方向と同様の傾向である。従って、信頼性の高いスカム幅の寸法測定を行うために、上述のように選択的に測長個所を設定することが望ましい。
図43に、測長個所の設定方法のバリエーションの例を示す。スカム4302の発生については上述の通り種々の形態があり、図40の(b)にて説明したような1方向に発生する例の他にも、例えば図43の(a)のようにパターン4301に対して斜めとなる方向に、かつ曲線を描くように変形して発生する場合や、図43の(b)のように隣接も対向もしていない場所に位置するパターン間にわたって発生する場合もある。このような場合、発生したスカムの複数の領域ごとに測長個所4303を設定し、どの領域においても常に垂直方向に電子線が走査されるようにすることもできる。また、1つの測長個所4303内においても、電子線の走査方向を調整することで常にスカムに対して垂直になるように走査することもできる。また、図43の(c)は発生した複数のスカム4302が接続状態にない場合を示す。このような場合においては、図39のステップ3906の判断によってスカム幅の寸法測定の非対象とし、測長個所4303を設けないこともできるが、例えば上述のようにスカム4302の長さが所定の条件よりも大きくなった場合には、当該スカムのみを測定対象として垂直方向に測長個所4303を設定できる。
さらに、設計データを用いて、予めパターンの位置に関する情報、例えば、近接しているパターンが存在するか等の情報を予め取得しておき、スカムを検出した場合に、スカムの発生位置から、当該予め把握している近接パターンを含むように測長個所4302を設定し、電子線を走査することもできる。図33にて説明したように、スカムは近接するパターンに接続する可能性があるため、このように、設計データから得られるパターンの位置関係に基づいて測長個所4302を設けることで、エッジ分岐点の抽出によっては一部分のみのスカムを検出し、当該スカムに連続する部分については検出することができなかった場合においても、実際に存在している状態のスカムを高精度に検出することが可能になる。
図45を用いて、スカムのSEM画像から輪郭データを抽出し、高さを評価する手順の一例について説明する。なお、当該手順は上述の手法によって検出したスカムに対して寸法や幅の測定など種々の評価と併せて適用可能である。例えば、検出したスカムが一定の幅を有していると判断された場合にのみ選択的に行うことができる。まず、SEM画像を取得したのちに(ステップ4501)、図10に示した手順によってパターンのエッジを1画素のエッジとして示す輪郭データを抽出する(ステップ4502)。次に、上述のライブラリからエッジ分岐点パターンであるテンプレートを読み出して(ステップ4503)、マッチング処理を行い(ステップ4504)、エッジ分岐点を抽出してスカムの検出を行う(ステップ4505)。そして、SEM画像において、当該検出したスカムを含む領域のラインプロファイルを作成する(ステップ4506)。当該プロファイルの作成においては、スカムが検出された領域において再度測長個所を設定して画像を取得しても良いし、既に取得した画像から作成されるプロファイルのうち、スカムの検出領域の部分のみを選択的に用いるようにしても良い。次に、図44Aにて説明したように、予め定めた所定数の閾値ごとに輪郭線を抽出し(ステップ4507)、所定数の閾値の輪郭線を全て抽出する(ステップ4508)。このとき、当該所定数の閾値について、プロファイルの高さが一定の間隔となるように設定することで、得られる複数の輪郭線は等高線となる。このとき、どのような条件においても常に一定の基準によってプロファイルの高さを評価するために、画像のコントラスト/ブライトネスの自動調整等、プロファイルの形態に影響を及ぼす操作を行わないようにすることが望ましい。
そして次に、例えば、スカムの下部輪郭データから、上部輪郭データまでの間に所定の本数以上の輪郭データが形成されている場合には(ステップ4509)、ある程度の高さをもった危険度の高いスカムであるとして判定し、検査対象とし(ステップ4510)、それ以外の場合には非検査対象とすることができる(ステップ4511)。
まず、SE像とBSE像をそれぞれ取得する(ステップ3501、3502)。ここで、BSE像の取得の際には、パターンに対して平行となる図34中の矢印3401にて示す方向に反射される電子を検出する検出器を用いることが望ましい。BSEは、SEよりも高いエネルギーを有した信号電子であるため、このように所望の位置に設定した検出器を用いた検出が可能となる。特に、上述の配線パターンの溝部に発生したスカムにおいては、パターンに対して垂直な方向に反射される電子よりも、平行な方向に反射される電子の検出量の方がスカムの程度をより反映したものであるため、このような方向に反射される反射電子を選択的に検出する。次に、両画像に対して上述の手法を用いて輪郭データを抽出する(ステップ3503、3504)。ここで、SE像については設計データとのマッチング処理によって位置合わせをし(ステップ3505)、側壁のエッジの探索を行って(ステップ3506)パターンの溝部を特定する(ステップ3507)。次に、SE像とBSE像とを重ね合わせ、焦点及びコントラスト等について調整を行って合成画像を取得する(ステップ3508)。次に、合成して得られた画像に対して検査領域を設定し(ステップ3509)、エッジ分岐パターンとのマッチングによって(ステップ3510)エッジ分岐点を抽出後(ステップ3511)、スカムを検出する(ステップ3512)。
以上の手順によれば、SE像と、BSE像との重ね合わせにより、SE像のみを用いた場合よりも正確にスカムを検出することが容易となる。さらに、両者の合成画像を適切に取得することによって、スカムの検出とパターンの寸法計測の両方を1枚の画像に基づいて効率良く行うことが可能になる。また、上記の手順において、BSE像を取得する際には、電子線の走査方向を、パターンに対して平行となる方向に設定することにより、パターンのエッジに対して相対的にスカムに起因したBSEを多く検出することもできる。
上記記載は実施例についてなされたが、本発明はそれに限らず、本発明の精神と添付の請求の範囲の範囲内で種々の変更および修正をすることができることは当業者に明らかである。
202 電子線
203 試料
204 二次電子検出器
205 反射電子検出器1
206 反射電子検出器2
207 A/D変換器
208 メモリ
209 CPU
210 画像処理ハードウェア
211 表示装置
212 レシピ作成装置
301,302,303 回路パターン
304,305 スカム
306,307,308,309 エッジ分岐点
400 座標情報
401,402 回路パターン
500 画像データの検査座標
501 評価エリア
503 擬似輪郭
3101 パターン
3102 パターンの表面の荒れ
3201 設計データ
3202 SEM画像
3203 スカム
3401,3601,4301,4405 パターン
3402 溝部
3403 パターンに平行な方向を示す矢印
3601 パターン
3602 洗浄液
4001 パターン
4002,4003,4004 スカム
4005,4303 測長個所
4006 電子線の走査方向
4007,4008 スカムの輝度プロファイル
4101 スカムの輪郭線データ
4102,4103,4104 電子線の走査方向及び形成されるプロファイル
4302 スカム
Claims (35)
- 画像から抽出された輪郭線データ上にて、特定形状のパターンを検出する検出部を備えた画像処理装置において、
当該検出部は、少なくとも3方向に分岐したエッジの分岐点を検出し、当該検出された分岐点を所定の条件に基づいて識別する識別部を備えたことを特徴とする画像処理装置。 - 請求項1において、
前記識別部は、前記分岐点の座標データと共に、当該分岐点に関する情報を、記憶媒体に記憶させることを特徴とする画像処理装置。 - 請求項1において、
前記識別部は、前記分岐点の数の情報を、記憶媒体に記憶させることを特徴とする画像処理装置。 - 請求項1において、
テンプレートを用いて特定形状のパターンを探索するマッチング処理部を備え、当該マッチング処理部は、前記輪郭線データの内、所定の領域について、選択的に分岐点の検出を行うことを特徴とする画像処理装置。 - 画像から抽出された輪郭線データ上にて、テンプレートを用いて特定形状のパターンを探索するマッチング処理部を備えた画像処理装置において、
当該マッチング処理部は、前記輪郭線データと、当該輪郭線の設計データとのマッチングに基づいて、輪郭線の各部位の識別を実行し、当該部位毎に識別された輪郭線データ上で、少なくとも3方向に分岐したエッジの分岐点を検出する検出部と、当該検出された分岐点が存在する輪郭線の各部位の識別情報と、当該分岐点に関する情報を関連付けて記憶媒体に記憶させる識別部を備えたことを特徴とする画像処理装置。 - 請求項5において、
前記識別部は、前記分岐点の座標データと共に、当該分岐点に関する情報を、記憶媒体に記憶させることを特徴とする画像処理装置。 - 請求項5において、
前記識別部は、前記分岐点の数の情報を、記憶媒体に記憶させることを特徴とする画像処理装置。 - 請求項5において、
前記マッチング処理部は、前記輪郭線データの内、所定の領域について、選択的に分岐点の検出を行うことを特徴とする画像処理装置。 - 画像から抽出された輪郭線データ上にて、特定形状のパターンの検出を、演算装置に実行させるコンピュータプログラムにおいて、
当該プログラムは、前記演算装置に、少なくとも3方向に分岐したエッジの分岐点を検出させ、当該検出された分岐点を所定の条件に基づいて識別させることを特徴とするコンピュータプログラム。 - 請求項9において、
前記プログラムは、前記演算装置に、前記分岐点の座標データと共に、当該分岐点に関する情報を、記憶媒体に記憶させることを特徴とするコンピュータプログラム。 - 請求項9において、
前記プログラムは、前記分岐点の数の情報を、前記演算装置に記憶させることを特徴とするコンピュータプログラム。 - 請求項9において、
前記プログラムは、前記演算装置に、前記輪郭線データの内、所定の領域について、選択的に分岐点の検出を行わせることを特徴とするコンピュータプログラム。 - 画像から抽出された輪郭線データ上にて、テンプレートを用いたパターンマッチングを、演算装置に実行させるコンピュータプログラムにおいて、
当該プログラムは、前記演算装置に、前記輪郭線データと、当該輪郭線の設計データとのマッチングに基づいて、輪郭線の各部位を識別させ、当該部位毎に識別された輪郭線データ上で、少なくとも3方向に分岐したエッジの分岐点を検出させ、当該検出された分岐点が存在する輪郭線の各部位の識別情報と、当該分岐点に関する情報を関連付けて記憶媒体に記憶させることを特徴とするコンピュータプログラム。 - 請求項13において、
前記プログラムは、前記演算装置に、前記分岐点の座標データと共に、当該分岐点に関する情報を、記憶媒体に記憶させることを特徴とするコンピュータプログラム。 - 請求項13において、
前記プログラムは、前記分岐点の数の情報を、前記演算装置に記憶させることを特徴とするコンピュータプログラム。 - 請求項13において、
前記プログラムは、前記演算装置に、前記輪郭線データの内、所定の領域について、選択的に分岐点の検出を行わせることを特徴とするコンピュータプログラム。 - 画像からパターンの形状を検査する検査部を備えた画像処理装置において、当該検査部はパターン上部に相当するエッジとパターン下部に相当するエッジを画像に含まれるパターンの傾斜部に対応するピーク信号に基づき検出し、当該パターン上部に相当するエッジとパターン下部に相当するエッジの間隔値を検査データとすることを特徴とする画像処理装置。
- 請求項17において、前記検査部は、計測された少なくとも2点以上のエッジの間隔値の統計量を検査データとすることを特徴とする画像処理装置。
- 請求項18において、前記検査部の前記統計量は、間隔値の最大、または平均、または分散に基づく数値であることを特徴とする画像処理装置。
- 請求項18において、前記検査部は前記統計量に対して所定の閾値を用いた判定処理を行い、パターン形状の良品、不良品の判定をすることを特徴とする画像処理装置。
- 画像からパターンの形状を検査する検査部を備えた画像処理装置において、当該検査部はパターンの傾斜部に対応するピーク信号に基づき、パターンのエッジを検出し、前記エッジによって形成される評価領域を画像に設定し、前記評価領域内の前記画像信号のばらつきに関する統計量を算出し、前記統計量を検査データとすることを特徴とする画像処理装置。
- 画像からパターンの形状を検査する検査部を備えた画像処理装置において、当該検査部は設計データやシミュレーションデータと前記画像とのパターンマッチング結果に基づき、評価領域を画像に設定し、前記評価領域内の前記画像信号のばらつきに関する統計量を算出し、前記統計量を検査データとすることを特徴とする画像処理装置。
- 請求項21において、前記統計量は画像信号の分散に基づく数値であることを特徴とする画像処理装置。
- 請求項22において、前記統計量は画像信号の分散に基づく数値であることを特徴とする画像処理装置。
- 請求項21において、前記検査部は前記統計量に対して所定の閾値を用いた判定処理を行い、パターン形状の良品、不良品の判定をすることを特徴とする画像処理装置。
- 請求項22において、前記検査部は前記統計量に対して所定の閾値を用いた判定処理を行い、パターン形状の良品、不良品の判定をすることを特徴とする画像処理装置。
- 画像から抽出された輪郭線データ上にて、特定形状のパターンを検出する検出部と、当該検出された特定形状のパターンを検査する検査部を備えた画像処理装置において、
前記検出部は、前記輪郭線データのうち、少なくとも3方向に分岐したエッジの分岐点を検出し、
前記検査部は、前記3方向に分岐したエッジのうち、半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジを含んだ領域を画像に設定し、
前記領域内の輝度に基づく信号波形を取得することを特徴とする画像処理装置。 - 請求項27において、
当該取得した信号波形に基づいて、
前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの幅を測定することを特徴とする画像処理装置。 - 請求項27において、
前記検出部は、当該分岐点を複数検出し、
前記検査部は、当該検出された複数の分岐点のうち、少なくとも2つの分岐点間を結ぶエッジが存在する場合には、当該エッジを含んだ領域について、
当該画像の輝度に基づく信号波形を取得することを特徴とする画像処理装置。 - 請求項27において、
前記検査部は、
前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの方向に対して垂直となる方向について、
前記領域内の当該画像の輝度に基づく信号波形を取得することを特徴とする画像処理装置。 - 請求項27において、
前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジに交差するように荷電粒子線を走査して、当該領域の画像を取得し、当該取得した画像の信号波形を取得することを特徴とする画像処理装置。 - 請求項27において、
前記検査部は、当該領域について、前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの長手方向に対して垂直に荷電粒子線を走査して、当該領域の画像を取得し、当該取得した画像の信号波形を取得することを特徴とする画像処理装置。 - 請求項27において、
前記検査部は、当該領域について、前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの長手方向に対して常に垂直に荷電粒子線を走査して、当該領域の画像を取得し、当該取得した画像の信号波形を取得することを特徴とする画像処理装置。 - 画像から抽出された輪郭線データ上にて、特定形状のパターンを検出する検出部と、当該検出された特定形状のパターンを検査する検査部を備えた画像処理装置において、
前記検出部は、少なくとも3方向に分岐したエッジの分岐点を検出し、
前記検査部は、前記3方向に分岐したエッジのうち、半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジを含んだ領域を画像に設定し、前記領域内の前記画像信号に基づくラインプロファイルを作成し、
当該作成されたラインプロファイルに基づいて、前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの幅を測定することを特徴とする画像処理装置。 - 請求項34において、
前記検査部は、
前記半導体デバイスの回路を構成するパターンを形成するエッジを除いたエッジの方向に対して垂直となる方向について、前記領域内の前記画像信号に基づくラインプロファイルを作成することを特徴とする画像処理装置。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020137010641A KR101549314B1 (ko) | 2010-10-27 | 2011-10-26 | 화상 처리 장치 및 기록 매체 |
JP2012540902A JP5548780B2 (ja) | 2010-10-27 | 2011-10-26 | 画像処理装置およびコンピュータプログラム |
US13/882,141 US9183622B2 (en) | 2010-10-27 | 2011-10-26 | Image processing apparatus |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010-240254 | 2010-10-27 | ||
JP2010240254 | 2010-10-27 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012057198A1 true WO2012057198A1 (ja) | 2012-05-03 |
Family
ID=45993904
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/074662 WO2012057198A1 (ja) | 2010-10-27 | 2011-10-26 | 画像処理装置およびコンピュータプログラム |
Country Status (5)
Country | Link |
---|---|
US (1) | US9183622B2 (ja) |
JP (3) | JP5548780B2 (ja) |
KR (1) | KR101549314B1 (ja) |
TW (1) | TWI475187B (ja) |
WO (1) | WO2012057198A1 (ja) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015141130A (ja) * | 2014-01-29 | 2015-08-03 | 株式会社日立ハイテクノロジーズ | パターン形状検査装置及びパターン形状検査方法 |
WO2016152582A1 (ja) * | 2015-03-20 | 2016-09-29 | 株式会社日立ハイテクノロジーズ | 電子線式パターン検査装置 |
WO2017042932A1 (ja) * | 2015-09-10 | 2017-03-16 | 株式会社日立ハイテクノロジーズ | 検査装置 |
CN113076907A (zh) * | 2021-04-16 | 2021-07-06 | 青岛海尔电冰箱有限公司 | 冰箱内物品信息标识方法、冰箱和计算机存储介质 |
CN118175237A (zh) * | 2024-04-01 | 2024-06-11 | 中创智元信息技术有限公司 | 基于实景三维的空间轮廓构建方法 |
Families Citing this family (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TW201430336A (zh) * | 2013-01-23 | 2014-08-01 | Huang Tian Xing | 缺陷檢測方法、裝置及系統 |
JP2016058465A (ja) * | 2014-09-08 | 2016-04-21 | 株式会社日立ハイテクノロジーズ | 欠陥定量化方法、欠陥定量化装置、および欠陥評価値表示装置 |
JP6867122B2 (ja) * | 2016-08-10 | 2021-04-28 | 株式会社Screenホールディングス | パターン倒壊状況画像およびその生成方法 |
US10706522B2 (en) * | 2016-11-08 | 2020-07-07 | Kla-Tencor Corporation | System and method for generation of wafer inspection critical areas |
JP7322549B2 (ja) * | 2019-06-28 | 2023-08-08 | セイコーエプソン株式会社 | 画像処理方法および画像処理装置 |
CN111027118B (zh) * | 2019-11-19 | 2024-01-19 | 广东博智林机器人有限公司 | 一种实测实量任务点搜索与任务派发方法及系统 |
JP2022165649A (ja) | 2021-04-20 | 2022-11-01 | 株式会社日立ハイテク | 欠陥検査装置、及び欠陥検査方法 |
DE102021113764A1 (de) * | 2021-05-27 | 2022-12-01 | Carl Zeiss Smt Gmbh | Verfahren und Vorrichtung zur Analyse eines Bildes einer mikrostrukturierten Komponente für die Mikrolithographie |
CN114022569B (zh) * | 2021-11-18 | 2024-06-07 | 湖北中烟工业有限责任公司 | 一种基于视觉测量箱体方正度的方法及装置 |
US20230267574A1 (en) * | 2022-02-23 | 2023-08-24 | Applied Materials Israel Ltd. | Reducing backscattered electron induced errors |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62184579A (ja) * | 1986-02-10 | 1987-08-12 | Toshiba Corp | パターン検出方法 |
JPS63223977A (ja) * | 1987-03-13 | 1988-09-19 | Mazda Motor Corp | 画像処理方法 |
JPH01106180A (ja) * | 1987-10-19 | 1989-04-24 | Nec Corp | パターン検査装置 |
JPH03252546A (ja) * | 1990-03-02 | 1991-11-11 | Matsushita Electric Ind Co Ltd | 配線パターン検査装置 |
JPH11339046A (ja) * | 1998-05-27 | 1999-12-10 | Nec Ibaraki Ltd | 自動外観検査装置と方法 |
JP2008047664A (ja) * | 2006-08-14 | 2008-02-28 | Hitachi High-Technologies Corp | パターン検査装置及び半導体検査システム |
JP2010129599A (ja) * | 2008-11-25 | 2010-06-10 | Toshiba Corp | パターン形状の評価方法及びこれを利用したパターン形状の評価装置 |
Family Cites Families (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0617875B2 (ja) * | 1986-11-14 | 1994-03-09 | 株式会社日立製作所 | パターン検査方法およびその装置 |
JP3317030B2 (ja) * | 1994-06-27 | 2002-08-19 | ヤマハ株式会社 | 寸法測定装置 |
JPH09295250A (ja) | 1996-04-30 | 1997-11-18 | Komatsu Ltd | 工作機械の工具異常検出方法 |
JP3031613B2 (ja) * | 1996-11-12 | 2000-04-10 | 株式会社つくばソフト研究所 | カラー/濃淡画像入力出力装置と入力出力方法 |
JP4209525B2 (ja) | 1998-12-17 | 2009-01-14 | 三菱電機株式会社 | パンタグラフすり板の厚み測定方法 |
US6326618B1 (en) * | 1999-07-02 | 2001-12-04 | Agere Systems Guardian Corp. | Method of analyzing semiconductor surface with patterned feature using line width metrology |
US6868175B1 (en) * | 1999-08-26 | 2005-03-15 | Nanogeometry Research | Pattern inspection apparatus, pattern inspection method, and recording medium |
JP2002195955A (ja) * | 2000-12-25 | 2002-07-10 | Matsushita Electric Ind Co Ltd | 半導体欠陥検査方法及び半導体欠陥検査装置 |
JP3870044B2 (ja) | 2001-07-25 | 2007-01-17 | 株式会社日立製作所 | パターン検査方法及びパターン検査装置 |
US7220990B2 (en) * | 2003-08-25 | 2007-05-22 | Tau-Metrix, Inc. | Technique for evaluating a fabrication of a die and wafer |
JP2005242071A (ja) * | 2004-02-27 | 2005-09-08 | Fuji Photo Film Co Ltd | 平版印刷版の処理方法及び平版印刷版の処理装置 |
JP4599110B2 (ja) * | 2004-07-30 | 2010-12-15 | キヤノン株式会社 | 画像処理装置及びその方法、撮像装置、プログラム |
JP4792331B2 (ja) | 2006-05-31 | 2011-10-12 | 株式会社日立ハイテクインスツルメンツ | 電子部品装着装置 |
DE102006044595B4 (de) * | 2006-09-19 | 2011-06-22 | Baumer Optronic GmbH, 01454 | Bildverarbeitungsvorrichtung zur Segmentierung anhand von Konturpunkten |
JP2009222454A (ja) | 2008-03-14 | 2009-10-01 | Hitachi High-Technologies Corp | パターン測定方法及びパターン測定装置 |
JP2009244058A (ja) * | 2008-03-31 | 2009-10-22 | Nireco Corp | 品質検査装置及び品質検査方法 |
-
2011
- 2011-10-20 TW TW100138088A patent/TWI475187B/zh not_active IP Right Cessation
- 2011-10-26 JP JP2012540902A patent/JP5548780B2/ja not_active Expired - Fee Related
- 2011-10-26 US US13/882,141 patent/US9183622B2/en not_active Expired - Fee Related
- 2011-10-26 KR KR1020137010641A patent/KR101549314B1/ko active IP Right Grant
- 2011-10-26 WO PCT/JP2011/074662 patent/WO2012057198A1/ja active Application Filing
-
2013
- 2013-12-06 JP JP2013252618A patent/JP2014077798A/ja active Pending
-
2015
- 2015-10-14 JP JP2015202843A patent/JP6134366B2/ja not_active Expired - Fee Related
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62184579A (ja) * | 1986-02-10 | 1987-08-12 | Toshiba Corp | パターン検出方法 |
JPS63223977A (ja) * | 1987-03-13 | 1988-09-19 | Mazda Motor Corp | 画像処理方法 |
JPH01106180A (ja) * | 1987-10-19 | 1989-04-24 | Nec Corp | パターン検査装置 |
JPH03252546A (ja) * | 1990-03-02 | 1991-11-11 | Matsushita Electric Ind Co Ltd | 配線パターン検査装置 |
JPH11339046A (ja) * | 1998-05-27 | 1999-12-10 | Nec Ibaraki Ltd | 自動外観検査装置と方法 |
JP2008047664A (ja) * | 2006-08-14 | 2008-02-28 | Hitachi High-Technologies Corp | パターン検査装置及び半導体検査システム |
JP2010129599A (ja) * | 2008-11-25 | 2010-06-10 | Toshiba Corp | パターン形状の評価方法及びこれを利用したパターン形状の評価装置 |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015141130A (ja) * | 2014-01-29 | 2015-08-03 | 株式会社日立ハイテクノロジーズ | パターン形状検査装置及びパターン形状検査方法 |
WO2016152582A1 (ja) * | 2015-03-20 | 2016-09-29 | 株式会社日立ハイテクノロジーズ | 電子線式パターン検査装置 |
JPWO2016152582A1 (ja) * | 2015-03-20 | 2017-05-25 | 株式会社日立ハイテクノロジーズ | 電子線式パターン検査装置 |
WO2017042932A1 (ja) * | 2015-09-10 | 2017-03-16 | 株式会社日立ハイテクノロジーズ | 検査装置 |
JPWO2017042932A1 (ja) * | 2015-09-10 | 2018-06-14 | 株式会社日立ハイテクノロジーズ | 検査装置 |
CN113076907A (zh) * | 2021-04-16 | 2021-07-06 | 青岛海尔电冰箱有限公司 | 冰箱内物品信息标识方法、冰箱和计算机存储介质 |
CN118175237A (zh) * | 2024-04-01 | 2024-06-11 | 中创智元信息技术有限公司 | 基于实景三维的空间轮廓构建方法 |
Also Published As
Publication number | Publication date |
---|---|
KR101549314B1 (ko) | 2015-09-01 |
TW201237363A (en) | 2012-09-16 |
JPWO2012057198A1 (ja) | 2014-05-12 |
KR20130065717A (ko) | 2013-06-19 |
JP2016028252A (ja) | 2016-02-25 |
JP2014077798A (ja) | 2014-05-01 |
TWI475187B (zh) | 2015-03-01 |
US20130279793A1 (en) | 2013-10-24 |
JP5548780B2 (ja) | 2014-07-16 |
JP6134366B2 (ja) | 2017-05-24 |
US9183622B2 (en) | 2015-11-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6134366B2 (ja) | 画像処理装置およびコンピュータプログラム | |
US10445875B2 (en) | Pattern-measuring apparatus and semiconductor-measuring system | |
TWI512684B (zh) | Defect observation method and defect observation device | |
US10297021B2 (en) | Defect quantification method, defect quantification device, and defect evaluation value display device | |
US9311697B2 (en) | Inspection method and device therefor | |
KR101588367B1 (ko) | 하전 입자선 장치 | |
JP5543872B2 (ja) | パターン検査方法およびパターン検査装置 | |
US20120099781A1 (en) | Method and apparatus of pattern inspection and semiconductor inspection system using the same | |
JP6759034B2 (ja) | パターン評価装置及びコンピュータープログラム | |
JP5988615B2 (ja) | 半導体評価装置、及びコンピュータープログラム | |
WO2013140907A1 (ja) | パターン計測装置、及び半導体計測システム | |
US20170372464A1 (en) | Pattern inspection method and pattern inspection apparatus | |
KR101615843B1 (ko) | 반도체 계측 장치 및 기록 매체 | |
US8045807B2 (en) | Pattern edge detecting method and pattern evaluating method | |
US20230194253A1 (en) | Pattern Inspection/Measurement Device, and Pattern Inspection/Measurement Program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11836331 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2012540902 Country of ref document: JP Kind code of ref document: A |
|
ENP | Entry into the national phase |
Ref document number: 20137010641 Country of ref document: KR Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13882141 Country of ref document: US |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11836331 Country of ref document: EP Kind code of ref document: A1 |