CN107545572A - The treating method and apparatus of target in image - Google Patents
The treating method and apparatus of target in image Download PDFInfo
- Publication number
- CN107545572A CN107545572A CN201610463501.2A CN201610463501A CN107545572A CN 107545572 A CN107545572 A CN 107545572A CN 201610463501 A CN201610463501 A CN 201610463501A CN 107545572 A CN107545572 A CN 107545572A
- Authority
- CN
- China
- Prior art keywords
- image
- target
- primary election
- gray level
- split
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of processing method and processing device of target in image, this method obtains marginal information by the gray level image of the image to be split to acquisition using edge detection method, and according to edge extraction to primary election target, to realize the preliminary extraction to target.Further, on the basis of the primary election target tentatively extracted, predeterminable area where extraction primary election target obtains area image corresponding to each primary election target, by area image binaryzation, and according to the primary election target of the image after binaryzation, second extraction from primary election target to extract the target of each area image, to realize accurate extraction target.Due to being to carry out target second extraction respectively on the area image where each primary election target, rather than second extraction is carried out on the basis of original image, substantially reducing needs dimension of picture to be processed, it is possible to increase the processing speed of Objective extraction in image.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind for the treatment of method and apparatus of target in image.
Background technology
With the development of image processing techniques, Micro-Medical Images processing is widely used in blood cell classification, cyto-diagnosis, dye
The medical domains such as colour solid type analysis.Micro-Medical Images treatment technology is advantageous to mitigate medical worker's onerous toil, is medical matters
Personnel provide reliable diagnosis basis, greatly improve the operating efficiency of medical worker.
Parasite egg identification based on image is a key areas in Micro-Medical Images treatment technology, passes through excrement
Just parasite egg quantity and worm's ovum type in microscopy detection fecal sample.The conventional means of stool examination are automatic microscopy, from
Index glass inspection refers to utilize MIcrosope image, splits worm's ovum from MIcrosope image automatically, calculate the quantity of worm's ovum.
However, the situation of stool image is complicated, usually there is the phenomenon of worm's ovum and impurity adhesion, in order to detect worm's ovum,
Wide-field use is often used, therefore the resolution ratio of image is larger.Thus, the image data amount for causing sampling to obtain is big, place
It is slow to manage speed.
The content of the invention
Based on this, it is necessary to provide a kind for the treatment of method and apparatus of target in image for improving processing speed.
The processing method of target in a kind of image, including：
Obtain image to be split；
The image to be split is subjected to gray processing processing, obtains gray level image；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
Predeterminable area where the primary election target is extracted from the image of association obtains area image, the figure of the association
As including one kind in image to be split and the gray level image；
Binary conversion treatment is carried out to the area image and obtains bianry image；
According to the primary election target of the bianry image, extraction obtains each administrative division map from the primary election target respectively
The target of picture.
In one embodiment, the image of the association is image to be split；
The image to be split is subjected to gray processing processing described, before the step of obtaining gray level image, in addition to：It is right
The image to be split carries out diminution processing；
After the step of primary election target in the gray level image according to the edge extraction, it is described from
Before predeterminable area where the primary election target is extracted in the image of association obtains area image step, in addition to：
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election target corresponding to being extracted according to primary election target after the amplification in image to be split, obtains image to be split
In primary election target.
In one embodiment, the step of the primary election target in the gray level image according to the edge extraction
After rapid, the step of predeterminable area where the primary election target is extracted in the image from association obtains area image it
Before, in addition to：
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, corresponding primary election target in the gray level image is rejected.
The processing method of target in a kind of image, including：
Obtain image to be split；
Gray processing processing is carried out to the image to be split, obtains gray level image；
Diminution processing is carried out to the gray level image, the gray level image after being reduced；
Using edge detection algorithm, the marginal information in the gray level image after the diminution is obtained；
The primary election target in gray level image after the diminution according to the edge extraction；
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election corresponding to being extracted according to primary election target after the amplification in the image to be split or the gray level image
Target, obtain the primary election target in image or gray level image to be split；
Binary conversion treatment is carried out to the image to be split or the gray level image, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
In one embodiment, the primary election in the gray level image after the diminution according to the edge extraction
After the step of target, the described the step of primary election target is amplified to primary election target after initial size is amplified it
Before, in addition to：
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, reject corresponding in the image to be split or gray level image
Primary election target.
The processing method of target in a kind of image, including：
Obtain image to be split；
The image to be split is subjected to diminution processing, the image to be split after being reduced；
Gray processing processing is carried out to the image to be split after the diminution, obtains gray level image；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election target corresponding to being extracted according to primary election target after the amplification in the image to be split, is obtained to be split
Primary election target in image；
Binary conversion treatment is carried out to the image to be split, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
The processing method of target in a kind of image, including：
Obtain the gray level image of image to be split；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, corresponding primary election target in the gray level image is rejected；
Binary conversion treatment is carried out to the gray level image, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
The processing unit of target in a kind of image, including：
Acquisition module, for obtaining image to be split；
Gray processing module, for the image to be split to be carried out into gray processing processing, obtain gray level image；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Area image processing module, obtained for the predeterminable area where the primary election target is extracted from the image of association
Area image, the image of the association include one kind in image to be split and the gray level image；
Binarization block, bianry image is obtained for carrying out binary conversion treatment to the area image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target
Obtain the target of each area image.
In one embodiment, the image of the association is image to be split；The processing unit of target is also in described image
Including：
Image processing module, for carrying out diminution processing to the image to be split；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for primary election mesh corresponding to being extracted according to primary election target after the amplification in image to be split
Mark, obtains the primary election target in image to be split.
In one embodiment, the processing unit of target also includes in described image：
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Module is rejected, for when the area of the primary election target is less than given threshold, it to be right in the gray level image to reject
The primary election target answered.
The processing unit of target in a kind of image, including：
Image collection module, for obtaining image to be split；
Gray processing module, for carrying out gray processing processing to the image to be split, obtain gray level image；
Image processing module, for carrying out diminution processing to the gray level image, the gray level image after being reduced；
Edge detection module, for using edge detection algorithm, obtaining the edge letter in the gray level image after the diminution
Breath；
Primary election object extraction module, for the primary election in the gray level image after the diminution according to the edge extraction
Target；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for according to primary election target after the amplification in the image to be split or the gray level image
Primary election target corresponding to middle extraction, obtain the primary election target in image or gray level image to be split；
Binarization block, for carrying out binary conversion treatment to the image to be split or the gray level image, obtain two-value
Image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target
Corresponding target.
In one embodiment, in described image target processing unit, in addition to：
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Extraction module, for when the area of the primary election target is less than given threshold, reject the image to be split or
Corresponding primary election target in gray level image.
The processing unit of target in a kind of image, including：
Image collection module, for obtaining image to be split；
Image processing module, for the image to be split to be carried out into diminution processing, the image to be split after being reduced；
Gray processing module, for carrying out gray processing processing to the image to be split after the diminution, obtain gray level image；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for primary election mesh corresponding to being extracted according to primary election target after the amplification in image to be split
Mark, obtains the primary election target in image to be split；
Binarization block, for carrying out binary conversion treatment to the image to be split, obtain bianry image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target
Corresponding target.
The processing unit of target in a kind of image, including：
Acquisition module, for obtaining the gray level image of image to be split；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Module is rejected, for when the area of the primary election target is less than given threshold, it to be right in the gray level image to reject
The primary election target answered；
Binarization block, for carrying out binary conversion treatment to the gray level image, obtain bianry image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target
Corresponding target.
The processing method of target in the image of present embodiment, the gray-scale map after processing by obtaining image to be split
Picture, using edge detection algorithm, the marginal information in gray level image after acquisition processing；After being handled according to edge extraction
Primary election target in gray level image；Predeterminable area where primary election target is extracted in gray level image obtains area image；To area
Area image carries out binary conversion treatment and obtains bianry image；According to the primary election target of bianry image, extracted respectively from primary election target
Obtain the target of each area image.This method utilizes edge detection method by the gray level image of the image to be split to acquisition
Marginal information is obtained, and according to edge extraction to primary election target, to realize the preliminary extraction to target.Further, exist
On the basis of the primary election target tentatively extracted, predeterminable area where extraction primary election target obtains region corresponding to each primary election target
Image, by area image binaryzation, and according to the primary election target of the image after binaryzation, second extraction from primary election target to carry
The target of each area image is taken, to realize accurate extraction target.Due to being on the area image where each primary election target
Target second extraction is carried out respectively, rather than second extraction is carried out on the basis of original image, and substantially reducing need to be to be processed
Dimension of picture, it is possible to increase the processing speed of Objective extraction in image.
Brief description of the drawings
Fig. 1 is a kind of flow chart of the processing method of target in image of embodiment；
Fig. 2 be another embodiment image in target processing method flow chart；
Fig. 3 is a kind of excrement image of embodiment；
Fig. 4 is an area image from the excrement image zooming-out shown in Fig. 3；
Fig. 5 is the bianry image of the area image shown in Fig. 4；
Fig. 6 is that the excrement image shown in Fig. 3 uses the worm's ovum target that the processing method of target in image is extracted；
Fig. 7 is a kind of flow chart of the processing method of target in image of embodiment；
Fig. 8 be another embodiment image in target processing method flow chart；
Fig. 9 again another embodiment image in target processing method flow chart；
Figure 10 is a kind of high-level schematic functional block diagram of the processing unit of target in image of embodiment；
Figure 11 be another embodiment image in target processing unit high-level schematic functional block diagram；
Figure 12 is a kind of high-level schematic functional block diagram of the processing unit of target in image of embodiment；
Figure 13 is a kind of high-level schematic functional block diagram of the processing unit of target in image of embodiment.
Embodiment
In order that the purpose of the present invention, technical scheme and advantage are more clearly understood, below in conjunction with drawings and Examples,
The present invention will be described in further detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention,
Do not limit the present invention.
In one embodiment, there is provided the processing method of target in a kind of image, as shown in Figure 1.The present embodiment is with can
Realize to being illustrated in image exemplified by the terminal of the processing of target, the terminal includes but is not limited to computer.Transported in the terminal
Row has application program, and the processing method of target in the image can be realized by the application program.This method includes following step
Suddenly：
S102：Obtain image to be split.
The mode for obtaining image is the image shot using CCD camera, for example, the sample after the display mirror amplification of shooting obtains
The image arrived.Sample is biological specimen, including blood sample, fecal sample and urine specimen etc..Specific sample type according to
Depending on the project of detection.
S104：Image to be split is subjected to gray processing processing, obtains gray level image.
The gray processing of image refers to original coloured image being converted into gray level image, and specific method is according to original color
R, G, the B component of pixel in image calculate gray value, and the method for specific gray processing can use simple component method, weighted average
Any one of method, maximum value process.Gray level image refers to that each pixel only has the image of monochrome information.This kind of image is generally shown
For from most furvous to most bright white gray scale.
S106：Using edge detection algorithm, the marginal information in gray level image is obtained.
Rim detection refers to detect brightness in gray level image and changes obvious point, and these points are usually marginal point.This implementation
Edge detection algorithm in mode can use Canny operators, Prewitt operators, Lrisch operators and Gauss-Laplacian to calculate
Any one of son.By using edge detection algorithm, the marginal information in gray level image is obtained.Signified side in the present embodiment
Edge information refers to utilize edge detection algorithm, the edge data information of the material present in gray level image.Existing material bag
Include target and other impurity.
S108：Primary election target in edge extraction gray level image.
In one embodiment, using the method for Contour extraction, according to marginal information by sequentially find out marginal point with
Primary election target is extracted to form the profile of closure in track border.
In another embodiment, using algorithm of region growing, according to rim detection marginal point as seed, seed edge
Eight neighborhood or four neighborhoods are extended growth, and the final result of seed growth is primary election target.
S110：Predeterminable area where primary election target is extracted from the image of association obtains area image.
The image of association refers to image corresponding with primary election target, and the image of association includes image to be split and gray level image
In one kind.
By taking gray level image as an example, predeterminable area refers on the border of primary election target, spreads default pixel to correspondence direction
The region that point obtains.For example, on the up, down, left and right border of primary election target, four pixels are spread respectively and obtain administrative division map
Picture.Therefore, area image includes the background where primary election target and primary election target.It is understood that the quantity of area image
It is identical with the quantity of primary election target.In the case of having multiple primary election targets in gray level image, extraction obtains the area of respective amount
Area image.
If the image of association is image to be split, the coordinate of primary election target is corresponded in image to be split, treated point
The border of primary election target in image is cut, region that default pixel obtain is spread to correspondence direction and extraction obtains administrative division map
Picture.
S112：Binary conversion treatment is carried out to area image and obtains bianry image.
Binary conversion treatment refers to the gray value of the pixel on image being arranged to 0 or 255, and whole image only has black and white
Visual effect, will area image carry out binary conversion treatment after obtain bianry image.Binaryzation back zone area image only has black and white
Two kinds of colors, primary election target is white, and background is in black.It should be appreciated that for extracting multiple primary election mesh in original image
It is corresponding to obtain each area image, it is necessary to each area image progress binary conversion treatment for the situation of area image corresponding to mark
Bianry image.
S114：According to the primary election target of bianry image, the target of each area image is respectively obtained.
In the present embodiment, according to the primary election target of bianry image, second extraction is carried out to primary election target and obtains each region
The target of image.It should be appreciated that in the present embodiment, the bianry image of the area image where each primary election target is carried out
Second extraction, obtain the target of each area image.
According to the primary election target of bianry image, second extraction obtains mesh calibration method, the method that can use morphology operation.
Morphology operations refer to for image processing method of the bianry image according to the set enumeration tree of mathematical morphology.Morphology operations
Including erosion operator and Expanded Operators, corrosion is a kind of process for eliminating boundary point, making border internally shrink.It can be used for disappearing
Except small and insignificant object, target is separated with other impurity, obtain target, all background dots of expansion object contact merge
Into the object, border the object expansion of determination can be become into big to the process of outside expansion.Therefore, using form student movement
Calculation can finely be extracted from the basis of the primary election target tentatively extracted, by target with other impurity are separated obtains
Target.It should be appreciated that, it is necessary to every in the case of area image corresponding to multiple primary election targets is extracted in original image
Individual area image carries out binary conversion treatment, and the primary election target to each bianry image, is respectively obtained often using morphology operations
The target of individual area image.
Due to being to carry out Target Segmentation respectively on the area image where each primary election target, rather than in original image
On the basis of carry out Target Segmentation, substantially reducing needs dimension of picture to be processed, it is possible to increase the processing speed of Objective extraction.
The processing method of target in the image of present embodiment, edge is utilized by the gray level image to be split to acquisition
Detection method obtains marginal information, and according to edge extraction to primary election target, to realize the preliminary extraction to target.Enter one
Step, on the basis of the primary election target tentatively extracted, predeterminable area where extraction primary election target obtains each primary election target pair
The area image answered, by area image binaryzation, and according to the primary election target of the image after binaryzation, second extraction is with from primary election
The target of each area image is extracted in target, to realize accurate extraction target.Due to being in the area where each primary election target
Target second extraction is carried out on area image respectively, rather than second extraction is carried out on the basis of original image, substantially reducing needs
Dimension of picture to be processed, it is possible to increase the processing speed of Objective extraction in image.
In another embodiment, between step S102 and step S104, in addition to step：Treat the progress of segmentation figure picture
Diminution is handled.Diminution processing refers to downscaled images area to the size of preset area.After image down to be split, corresponding ash
Degree image is also contracted to the image of default size, gray level image size reduction, it is possible to increase the processing speed of subsequent edges detection.
It is corresponding, comprise the following steps (1) and (2) before step S108 and step step S110：
(1) primary election target, is amplified to initial size and is amplified rear primary election target.
(2) primary election target corresponding to, being extracted according to primary election target after amplification in image to be split, obtains image to be split
In primary election target.
Initial size refers to the size of the image to be split before reducing, in the present embodiment, to improve edge inspection
The speed of survey, before rim detection, original image is contracted to default size.To avoid the area image area mistake of extraction
It is small, the subsequently fine segmentation to primary election target is not easy to, it is necessary to extract marginal information, after the coordinate for obtaining primary election target,
The primary election target of extraction is amplified to original size, coordinate of the primary election target in original image to be split is obtained, so as to contract
Primary election target corresponding to being obtained in original image before small, obtains the primary election target in image to be split, so as to avoid subsequently locating
The bianry image of reason is too small, is not easy to become more meticulous and mentions worm's ovum target.
In another embodiment, after step s 104, in addition to：Diminution processing is carried out to gray level image, contracted
Gray level image after small.
Corresponding, step S106 is：Using edge detection algorithm, the marginal information in the gray level image after reducing is obtained.
Step S108 is：Primary election target in gray level image after being reduced according to edge extraction.
In this embodiment, the image of association includes image to be split and gray level image.
It is corresponding, step (1) is included between step S108 and step S110 to step (2)：
Step (1), primary election target is amplified to initial size it is amplified rear primary election target.
Step (2), extracted in image to be split or gray level image according to primary election target after amplification corresponding to primary election target,
Obtain the primary election target in primary election target or the gray level image in image to be split.
By taking worm's ovum target in excrement image as an example, worm's ovum content ratio is less in excrement image, generally for detection as far as possible
Go out worm's ovum, wide-field sampling can be used, therefore, the resolution ratio of image is larger, has reached 2400 × 1800.In worm's ovum image
In extraction, worm's ovum target is big, therefore, still more visible after image down, can extract marginal information.Gray level image is contracted to preset
The image of size, it is possible to increase the processing speed of subsequent edges detection.In a particular embodiment, image down degree can be by
User sets according to the size of shooting picture and detection project by rule of thumb, for example, in one embodiment, by image down extremely
The 1/16 of original area, the gray level image after being handled.It is follow-up to use edge detection algorithm, obtain the gray level image after reducing
Marginal information.In the present embodiment, by reducing in advance in the gray level image after image or gray processing to be split, and after diminution
Primary election target is extracted in image, to improve the speed of rim detection.Primary election target is amplified in corresponding associated images again, root
Primary election target corresponding to being extracted according to primary election target after amplification in the image of association, i.e., reduce the image to be split of before processing or
It is primary election target corresponding to extraction in the gray level image before reducing.
In another embodiment, the step of step S112 includes：Two are carried out to area image using adaptive threshold
Value handles to obtain bianry image.
Adaptive threshold is a kind of threshold technology improved, and wherein threshold value is a variable in itself, each pixel
Adaptive threshold T (x, y) is different, by calculating the weighted average of the certain area around pixel, by pixel with calculating
Adaptive threshold be compared, if being more than, pixel is entered as 0, otherwise, is entered as 255.In order to be sealed as far as possible
Image is closed, fixed adjustment adaptive threshold, subtracts a constant to obtain threshold value T (x, y)-A, such adaptive threshold effect ratio
Preferably, Closed Graph picture is obtained as far as possible, can be automatically found the closed outline of image object, consequently facilitating subsequently filling primary election mesh
Mark.
In other embodiments, fixed threshold can also be used to carry out binary conversion treatment to area image and obtains two-value
Image.It is fixed that fixed threshold binaryzation, which uses threshold value, by the gray value of pixel successively compared with fixed threshold,
If being more than, pixel is entered as 0, otherwise, is entered as 255, but is unable to self-adaption binaryzation, therefore image object wheel be present
Exterior feature fracture or over-segmentation risk.
In another embodiment, it is further comprising the steps of before step S110 steps after step S108 steps
(1) to step (3)
(1) primary election target, is filled, and calculates the area of the primary election target after filling.
After the primary election target in edge extraction gray level image, the edge for having substantial amounts of primary election target is fracture
, to calculate the area of primary election target, it is necessary to fill primary election target, make primary election target complete.Specifically, calculated using area filling
Method, for example, simple horizontal and vertical filling algorithm is done to each primary election target obtains complete primary election target, then calculate just
Select the area of target.
(2), judge whether the area of primary election target is less than given threshold.
(3) primary election corresponding to being rejected, when the area of primary election target is less than given threshold, on gray level image after treatment
Target.
In the present embodiment, by calculating the area of primary election target after being filled to primary election target, when primary election target
Area rejected when being less than given threshold, in gray level image after treatment corresponding to primary election target, so as to pass through area
Calculating is tentatively screened to the quantity of target, after screening, reduces the quantity of primary election target, so as to be obtained in step S110
The quantity of area image accordingly reduce, further, be advantageous to improve the segmentation efficiency of target in image.
In the present embodiment, the primary election target that area is less than predetermined threshold value is rejected, suitable for the mesh that target is large area
Target extract, for example, in excrement image worm's ovum target extraction.Due in excrement image worm's ovum target compared to other cells
Area is larger, rejects non-worm's ovum target by areal calculation screening, can reduce the quantity of primary election target, and then reduce primary election mesh
The quantity of target area image, so as to reduce follow-up amount of calculation, improve the splitting speed and efficiency of worm's ovum target.
In another embodiment, in a kind of image target processing method, as shown in Fig. 2 specifically including following steps：
S202：Obtain image to be split.
In a specific embodiment, image to be split is excrement image, as shown in Figure 3.
S204：Image to be split is subjected to gray processing processing, obtains gray level image.
S206：Gray level image is contracted to default size, the gray level image after being reduced.
S208：Using edge detection algorithm, the marginal information in the gray level image after reducing is obtained.
S210：Primary election target in gray level image after being reduced according to edge extraction.
S212：Primary election target is filled, and calculates the area of the primary election target after filling.
S214：Judge whether the area of primary election target is less than given threshold.
If so, then perform step S216：Reject corresponding primary election target in the gray level image after reducing.
S218：Primary election target is amplified to initial size and is amplified rear primary election target.
S220：The primary election target according to corresponding to primary election target after amplification is extracted in gray level image, is obtained in gray level image
Primary election target.
S222：Predeterminable area where primary election target is extracted in gray level image obtains area image.
To the primary election target of the excrement image shown in Fig. 3, obtained area image is as shown in Figure 4.
S224：Binary conversion treatment is carried out to area image and obtains bianry image.
Binaryzation is carried out to area image as shown in Figure 4, obtained bianry image is as shown in Figure 5.After binaryzation, primary election
Target is white, and background is black.In the bianry image, worm's ovum is the maximum target of area, there is impurity around worm's ovum
Bonding.
S226：According to the primary election target of bianry image, extraction obtains the mesh of each area image from primary election target respectively
Mark.
The potential worm's ovum target for splitting to obtain to the excrement image shown in Fig. 3 is as shown in Figure 6.
The processing method of target in the image of the present embodiment, diminution processing is carried out by treating segmentation figure picture, to extraction
Predeterminable area where primary election target is screened and extracted to primary election target obtains area image, to improve target in image
Extraction process speed.
In one embodiment, there is provided the processing method of target in a kind of image, as shown in fig. 7, this method is including following
Step：
S702：Obtain image to be split.
The mode for obtaining image is the image shot using CCD camera, for example, the sample after the display mirror amplification of shooting obtains
The image arrived.Sample is biological specimen, including blood sample, fecal sample and urine specimen etc..Specific sample type according to
Depending on the project of detection.
S704：Image to be split is subjected to gray processing processing, obtains gray level image.
The gray processing of image refers to original coloured image being converted into gray level image, and specific method is according to original color
R, G, the B component of pixel in image calculate gray value, and the method for specific gray processing can use simple component method, weighted average
Any one of method, maximum value process.
S706：Diminution processing is carried out to gray level image, the gray level image after being reduced.
Diminution processing refers to downscaled images area to the size of preset area.Gray level image is contracted to the figure of default size
Picture, it is possible to increase the processing speed of subsequent edges detection.
By taking worm's ovum target in excrement image as an example, worm's ovum content ratio is less in excrement image, generally for detection as far as possible
Go out worm's ovum, wide-field sampling can be used, therefore, the resolution ratio of image is larger, has reached 2400 × 1800.Due in worm's ovum
It is still more visible after image down because worm's ovum target is big in image zooming-out, marginal information can be extracted.Gray level image is contracted to
The image of default size, it is possible to increase the processing speed of subsequent edges detection.
S708：Using edge detection algorithm, the marginal information in the gray level image after reducing is obtained.
Rim detection refers to that brightness changes obvious point in the gray level image after detection process, and these points are usually edge
Point.Edge detection algorithm in present embodiment can use Canny operators, Prewitt operators, Lrisch operators and Gauss-
Any one of Laplacian operators.By using edge detection algorithm, the edge letter in the gray level image after reducing is obtained
Breath.
S710：Primary election target in gray level image after being reduced according to edge extraction.
In one embodiment, using the method for Contour extraction, according to marginal information by sequentially find out marginal point with
Primary election target is extracted to form the profile of closure in track border.
In another embodiment, using algorithm of region growing, according to rim detection marginal point as seed, seed edge
Eight neighborhood or four neighborhoods are extended growth, and the final result of seed growth is primary election target.
S712：Primary election target is amplified to initial size and is amplified rear primary election target.
S714：Primary election target corresponding to being extracted according to primary election target after amplification in image to be split or gray level image, is obtained
To the primary election target in image to be split or gray level image.
Initial size refers to the size of the image to be split before reducing, in the present embodiment, to improve edge inspection
The speed of survey, before rim detection, gray level image is contracted to default size.To avoid the area image area mistake of extraction
It is small, the subsequently fine segmentation to primary election target is not easy to, it is necessary to extract marginal information, after the coordinate for obtaining primary election target,
The primary election target of extraction is amplified to original size, obtains gray-scale map of the primary election target before original image to be split or diminution
The coordinate of picture, so as to primary election target corresponding to acquisition in the image before diminution.
S716：Treat segmentation figure picture or gray level image carries out binary conversion treatment, obtain bianry image.
Binary conversion treatment refers to the gray value of the pixel on image being arranged to 0 or 255, and whole image only has black and white
Visual effect.Binaryzation back zone area image only has black and white two kinds of colors, and primary election target is white, and background is in black.
S718：According to the primary election target of bianry image, the target corresponding to extraction from primary election target respectively.
In the present embodiment, according to the primary election target of bianry image, second extraction is carried out to primary election target and obtains target.
According to the primary election target of bianry image, second extraction obtains mesh calibration method, the method that can use morphology operation.
Morphology operations refer to for image processing method of the bianry image according to the set enumeration tree of mathematical morphology.Morphology operations
Including erosion operator and Expanded Operators, corrosion is a kind of process for eliminating boundary point, making border internally shrink.It can be used for disappearing
Except small and insignificant object, target is separated with other impurity, obtain target, all background dots of expansion object contact merge
Into the object, border the object expansion of determination can be become into big to the process of outside expansion.Therefore, using form student movement
Calculation can finely be extracted from the basis of the primary election target tentatively extracted, by target with other impurity are separated obtains
Target.
In another embodiment, as shown in figure 8, in image target processing method, comprise the following steps
S802：Obtain image to be split.
S804：Image to be split is subjected to diminution processing, the image to be split after being reduced.
S806：Gray processing processing is carried out to the image to be split after diminution, obtains gray level image.
S808：Using edge detection algorithm, the marginal information in gray level image is obtained.
S810：Primary election target in edge extraction gray level image.
S812：Primary election target is amplified to initial size and is amplified rear primary election target.
S814：Primary election target corresponding to being extracted according to primary election target after amplification in image to be split, obtains figure to be split
Primary election target as in.
S816：Treat segmentation figure picture and carry out binary conversion treatment, obtain bianry image.
S818：According to the primary election target of bianry image, the target corresponding to extraction from primary election target respectively.
After the difference of a upper embodiment is that treating segmentation figure picture first reduces in the present embodiment, then carry out gray processing.
The processing method of target in the image of present embodiment, diminution processing is carried out by treating segmentation figure picture, to reducing
Imagery exploitation edge detection method afterwards obtains marginal information, and according to edge extraction to primary election target.Again by primary election mesh
Mark is amplified to initial size, according to primary election target corresponding to extraction in image of the primary election target before diminution after amplification, is contracted
The primary election target of image before small.Further, on the basis of the primary election target tentatively extracted, by the carry out two of correspondence image
Value is handled, and according to the primary election target of the image after binaryzation, second extraction from primary election target to extract target, to realize
Accurate extraction target.By extracting primary election target in the gray level image of diminution in advance, the speed of rim detection is improved, further
Improve image in target dividing processing speed.
In one embodiment, there is provided the processing method of target in a kind of image, as shown in figure 9, this method is including following
Step：
S902：Obtain the gray level image of image to be split.
Gray level image refers to that each pixel only has the image of monochrome information.This kind of image be typically shown as from most furvous to
Most bright white gray scale.In one embodiment, after coloured image to be split first being carried out into gray processing processing, then profit
Target Segmentation is carried out with the processing method of target in the image in the present embodiment.
S904：Using edge detection algorithm, the marginal information in gray level image is obtained.
Rim detection refers to detect brightness in gray level image and changes obvious point, and these points are usually marginal point.This implementation
Edge detection algorithm in mode can use Canny operators, Prewitt operators, Lrisch operators and Gauss-Laplacian to calculate
Any one of son.
S906：Primary election target in edge extraction gray level image.
In one embodiment, using the method for Contour extraction, according to marginal information by sequentially find out marginal point with
Primary election target is extracted to form the profile of closure in track border.
In another embodiment, using algorithm of region growing, according to rim detection marginal point as seed, seed edge
Eight neighborhood or four neighborhoods are extended growth, and the final result of seed growth is primary election target.
S908：Primary election target is filled, and calculates the area of the primary election target after filling.
After the primary election target in edge extraction gray level image, the edge for having substantial amounts of primary election target is fracture
, to calculate the area of primary election target, it is necessary to fill primary election target, make primary election target complete.Specifically, calculated using area filling
Method, for example, simple horizontal and vertical filling algorithm is done to each primary election target obtains complete primary election target, then calculate just
Select the area of target.
S910：When the area of primary election target is less than given threshold, corresponding primary election target in gray level image is rejected.
After area is less than the primary election target of given threshold in gray level image after rejecting processing, reduce primary election in gray level image
The quantity of target.
S912：Binary conversion treatment is carried out to gray level image, obtains bianry image.
S914：According to the primary election target of bianry image, the target corresponding to extraction from primary election target respectively.
The processing method of target in the image of present embodiment, utilized by the gray level image of the image to be split to acquisition
Edge detection method obtains marginal information, and according to edge extraction to primary election target, to realize the preliminary extraction to target.
By the primary election target to tentatively extracting, primary election target of the area less than given threshold is rejected according to area, realized to primary election mesh
Target is screened.Further, on the basis of primary election target, gray level image is subjected to binary conversion treatment, and according to binaryzation after
Image primary election target, second extraction from primary election target to extract target, to realize accurate extraction target.Pass through area pair
The primary election target tentatively extracted is screened, and can be rejected the primary election target for the condition of being unsatisfactory for, be reduced primary election in gray level image
The data of target, because the quantity of primary election target reduces, therefore, subsequently need the primary election destination number of second extraction to reduce, have
The processing speed of target is obtained beneficial to secondary splitting is improved, further, improves the segmentation efficiency of target in image.
In one embodiment, there is provided the processing unit of target in a kind of image, as shown in Figure 10, including：
Acquisition module 902, for obtaining image to be split.
Gray processing module 904, for image to be split to be carried out into gray processing processing, obtain gray level image.
Edge detection module 906, for using edge detection algorithm, obtaining the marginal information in gray level image.
Primary election object extraction module 908, for the primary election target in edge extraction gray level image.
Area image processing module 910, obtained for the predeterminable area where primary election target is extracted from the image of association
Area image, the image of association include one kind in image to be split and gray level image.
Binarization block 912, bianry image is obtained for carrying out binary conversion treatment to area image.
Split module 914, for the primary election target according to bianry image, extraction obtains each area from primary election target respectively
The target of area image.
The processing unit of target in the image of present embodiment, edge is utilized by the gray level image to be split to acquisition
Detection method obtains marginal information, and according to edge extraction to primary election target, to realize the preliminary extraction to target.Enter one
Step, on the basis of the primary election target tentatively extracted, predeterminable area where extraction primary election target obtains each primary election target pair
The area image answered, by area image binaryzation, and according to the primary election target of the image after binaryzation, second extraction is with from primary election
The target of each area image is extracted in target, to realize accurate extraction target.Due to being in the area where each primary election target
Target second extraction is carried out on area image respectively, rather than second extraction is carried out on the basis of original image, substantially reducing needs
Dimension of picture to be processed, it is possible to increase the processing speed of Objective extraction in image.
In another embodiment, as shown in figure 11, in image target processing unit, the image of association is figure to be split
Picture.
The processing unit of target in image, in addition to：
Image processing module 916, diminution processing is carried out for treating segmentation figure picture；
Amplification module 918, rear primary election target is amplified for primary election target to be amplified into initial size；
Object extraction module 920, for primary election mesh corresponding to being extracted according to primary election target after amplification in image to be split
Mark, obtains the primary election target in image to be split.
In one embodiment, the processing unit of target also includes in image：
Module 922 is filled, for filling primary election target, and calculates the area of the primary election target after filling.
Module 924 is rejected, it is corresponding first in gray level image for when the area of primary election target is less than given threshold, rejecting
Select target.
In one embodiment, as shown in figure 12, in a kind of image target processing unit, including：
Image collection module 702, for obtaining image to be split.
Gray processing module 704, gray processing processing is carried out for treating segmentation figure picture, obtains gray level image.
Image processing module 706, for carrying out diminution processing to gray level image, the gray level image after being reduced.
Edge detection module 708, for using edge detection algorithm, obtaining the edge letter in the gray level image after reducing
Breath.
Primary election object extraction module 710, for the primary election target in the gray level image after being reduced according to edge extraction.
Amplification module 712, rear primary election target is amplified for primary election target to be amplified into initial size.
Object extraction module 714, for being extracted according to primary election target after amplification in image to be split or gray level image pair
The primary election target answered, obtain the primary election target in image or gray level image to be split.
Binarization block 716, binary conversion treatment is carried out for treating segmentation figure picture or gray level image, obtains bianry image.
Split module 718, for the primary election target according to bianry image, the mesh corresponding to extraction from primary election target respectively
Mark.
The processing unit of target in the image of present embodiment, diminution processing is carried out by treating segmentation figure picture, to reducing
Imagery exploitation edge detection method afterwards obtains marginal information, and according to edge extraction to primary election target.Again by primary election mesh
Mark is amplified to initial size, according to primary election target corresponding to extraction in image of the primary election target before diminution after amplification, obtains
Primary election target in image before diminution.Further, on the basis of the primary election target tentatively extracted, corresponding image is entered
Row binary conversion treatment, and according to the primary election target of the image after binaryzation, second extraction to extract target from primary election target, with
Realize accurate extraction target.By extracting primary election target in the gray level image of diminution in advance, the speed of rim detection is improved, is entered
The dividing processing speed for improving target in image of one step.
In one embodiment, as shown in figure 13, in a kind of image target processing unit, including：
Acquisition module 802, for obtaining the gray level image of image to be split.
Edge detection module 804, for using edge detection algorithm, obtaining the marginal information in gray level image.
Primary election object extraction module 806, for the primary election target in edge extraction gray level image.
Module 808 is filled, for filling primary election target, and calculates the area of the primary election target after filling.
Module 810 is rejected, it is corresponding first in gray level image for when the area of primary election target is less than given threshold, rejecting
Select target.
Binarization block 812, for carrying out binary conversion treatment to gray level image, obtain bianry image.
Split module 814, for the primary election target according to bianry image, the mesh corresponding to extraction from primary election target respectively
Mark.
The processing unit of target in the image of present embodiment, utilized by the gray level image of the image to be split to acquisition
Edge detection method obtains marginal information, and according to edge extraction to primary election target, to realize the preliminary extraction to target.
By the primary election target to tentatively extracting, primary election target of the area less than given threshold is rejected according to area, realized to primary election mesh
Target is screened.Further, on the basis of primary election target, gray level image is subjected to binary conversion treatment, and according to binaryzation after
Image primary election target, second extraction from primary election target to extract target, to realize accurate extraction target.Pass through area pair
The primary election target tentatively extracted is screened, and can be rejected the primary election target for the condition of being unsatisfactory for, be reduced primary election in gray level image
The data of target, because the quantity of primary election target reduces, therefore, subsequently need the primary election destination number of second extraction to reduce, have
The processing speed of target is obtained beneficial to secondary splitting is improved, further, improves the segmentation efficiency of target in image.
Each technical characteristic of above example can be combined arbitrarily, to make description succinct, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, lance is not present in the combination of these technical characteristics
Shield, all it is considered to be the scope of this specification record.
Embodiment described above only expresses the several embodiments of the present invention, and its description is more specific and detailed, but simultaneously
Can not therefore it be construed as limiting the scope of the patent.It should be pointed out that come for one of ordinary skill in the art
Say, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to the protection of the present invention
Scope.Therefore, the protection domain of patent of the present invention should be determined by the appended claims.
Claims (14)
1. the processing method of target in a kind of image, including：
Obtain image to be split；
The image to be split is subjected to gray processing processing, obtains gray level image；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
Predeterminable area where the primary election target is extracted from the image of association obtains area image, the image bag of the association
Include one kind in image to be split and the gray level image；
Binary conversion treatment is carried out to the area image and obtains bianry image；
According to the primary election target of the bianry image, extraction obtains each area image from the primary election target respectively
Target.
2. the processing method of target in image according to claim 1, it is characterised in that：The image of the association is treats point
Cut image；
The image to be split is subjected to gray processing processing described, before the step of obtaining gray level image, in addition to：To described
Image to be split carries out diminution processing；
After the step of primary election target in the gray level image according to the edge extraction, described from association
Image in extract the primary election target where predeterminable area obtain area image step before, in addition to：
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election target corresponding to being extracted according to primary election target after the amplification in image to be split, is obtained in image to be split
Primary election target.
3. the processing method of target in image according to claim 1 or 2, it is characterised in that described according to the side
After the step of primary election target in gray level image described in edge information extraction, the primary election is extracted in the image from association
Before the step of predeterminable area where target obtains area image, in addition to：
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, corresponding primary election target in the gray level image is rejected.
4. the processing method of target in a kind of image, including：
Obtain image to be split；
Gray processing processing is carried out to the image to be split, obtains gray level image；
Diminution processing is carried out to the gray level image, the gray level image after being reduced；
Using edge detection algorithm, the marginal information in the gray level image after the diminution is obtained；
The primary election target in gray level image after the diminution according to the edge extraction；
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election target corresponding to being extracted according to primary election target after the amplification in the image to be split or the gray level image,
Obtain the primary election target in image or gray level image to be split；
Binary conversion treatment is carried out to the image to be split or the gray level image, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
5. the processing method of target in image according to claim 4, it is characterised in that believed described according to the edge
After breath extracts the step of primary election target in the gray level image after the diminution, the primary election target is amplified to just described
Beginning size was amplified before the step of rear primary election target, in addition to：
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, reject corresponding first in the image to be split or gray level image
Select target.
6. the processing method of target in a kind of image, including：
Obtain image to be split；
The image to be split is subjected to diminution processing, the image to be split after being reduced；
Gray processing processing is carried out to the image to be split after the diminution, obtains gray level image；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
The primary election target is amplified to initial size and is amplified rear primary election target；
Primary election target corresponding to being extracted according to primary election target after the amplification in the image to be split, obtains image to be split
In primary election target；
Binary conversion treatment is carried out to the image to be split, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
7. the processing method of target in a kind of image, including：
Obtain the gray level image of image to be split；
Using edge detection algorithm, the marginal information in the gray level image is obtained；
Primary election target in gray level image described in the edge extraction；
The primary election target is filled, and calculates the area of the primary election target after filling；
When the area of the primary election target is less than given threshold, corresponding primary election target in the gray level image is rejected；
Binary conversion treatment is carried out to the gray level image, obtains bianry image；
According to the primary election target of the bianry image, the target corresponding to extraction from the primary election target respectively.
A kind of 8. processing unit of target in image, it is characterised in that including：
Acquisition module, for obtaining image to be split；
Gray processing module, for the image to be split to be carried out into gray processing processing, obtain gray level image；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Area image processing module, region is obtained for the predeterminable area where the primary election target is extracted from the image of association
Image, the image of the association include one kind in image to be split and the gray level image；
Binarization block, bianry image is obtained for carrying out binary conversion treatment to the area image；
Split module, for the primary election target according to the bianry image, extract obtain from the primary election target respectively
The target of each area image.
9. the processing unit of target in image according to claim 8, it is characterised in that：The image of the association is treats point
Cut image；The processing unit of target also includes in described image：
Image processing module, for carrying out diminution processing to the image to be split；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for primary election target corresponding to being extracted according to primary election target after the amplification in image to be split,
Obtain the primary election target in image to be split.
10. the processing unit of target in image according to claim 8 or claim 9, it is characterised in that also include：
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Module is rejected, corresponding to when the area of the primary election target is less than given threshold, rejecting in the gray level image
Primary election target.
11. the processing unit of target in a kind of image, including：
Image collection module, for obtaining image to be split；
Gray processing module, for carrying out gray processing processing to the image to be split, obtain gray level image；
Image processing module, for carrying out diminution processing to the gray level image, the gray level image after being reduced；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image after the diminution；
Primary election object extraction module, for the primary election mesh in the gray level image after the diminution according to the edge extraction
Mark；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for being carried according to primary election target after the amplification in the image to be split or the gray level image
Primary election target, obtains the primary election target in image or gray level image to be split corresponding to taking；
Binarization block, for carrying out binary conversion treatment to the image to be split or the gray level image, obtain bianry image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target corresponding
Target.
12. the processing unit of target in image according to claim 11, it is characterised in that also include：
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Extraction module, for when the area of the primary election target is less than given threshold, rejecting the image to be split or gray scale
Corresponding primary election target in image.
13. the processing unit of target in a kind of image, including：
Image collection module, for obtaining image to be split；
Image processing module, for the image to be split to be carried out into diminution processing, the image to be split after being reduced；
Gray processing module, for carrying out gray processing processing to the image to be split after the diminution, obtain gray level image；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Amplification module, rear primary election target is amplified for the primary election target to be amplified into initial size；
Object extraction module, for primary election target corresponding to being extracted according to primary election target after the amplification in image to be split,
Obtain the primary election target in image to be split；
Binarization block, for carrying out binary conversion treatment to the image to be split, obtain bianry image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target corresponding
Target.
14. the processing unit of target in a kind of image, including：
Acquisition module, for obtaining the gray level image of image to be split；
Edge detection module, for using edge detection algorithm, obtaining the marginal information in the gray level image；
Primary election object extraction module, for the primary election target in the gray level image according to the edge extraction；
Module is filled, for filling the primary election target, and calculates the area of the primary election target after filling；
Module is rejected, corresponding to when the area of the primary election target is less than given threshold, rejecting in the gray level image
Primary election target；
Binarization block, for carrying out binary conversion treatment to the gray level image, obtain bianry image；
Split module, for the primary election target according to the bianry image, extracted respectively from the primary election target corresponding
Target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610463501.2A CN107545572A (en) | 2016-06-23 | 2016-06-23 | The treating method and apparatus of target in image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610463501.2A CN107545572A (en) | 2016-06-23 | 2016-06-23 | The treating method and apparatus of target in image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107545572A true CN107545572A (en) | 2018-01-05 |
Family
ID=60960825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610463501.2A Pending CN107545572A (en) | 2016-06-23 | 2016-06-23 | The treating method and apparatus of target in image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107545572A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108304910A (en) * | 2018-02-27 | 2018-07-20 | 华北理工大学 | A kind of slip sheet of air conditioner compressor method of counting based on machine vision |
| CN108765516A (en) * | 2018-03-26 | 2018-11-06 | 阳光暖果（北京）科技发展有限公司 | A kind of face region colouring method based on thumbnail |
| CN109907535A (en) * | 2018-07-26 | 2019-06-21 | 永康市柴迪贸易有限公司 | Corner bookcase based on worm analysis |
| WO2020140826A1 (en) * | 2019-01-02 | 2020-07-09 | 平高集团有限公司 | Infrared temperature sensing method employing contour extraction for target object, device, and storage medium |
| CN112215856A (en) * | 2020-10-20 | 2021-01-12 | 歌尔光学科技有限公司 | Image segmentation threshold determination method and device, terminal and storage medium |
| CN114299494A (en) * | 2022-01-20 | 2022-04-08 | 广东省农业科学院动物科学研究所 | Method and system for detecting worm oval characteristics of aquatic product image |
| WO2022120714A1 (en) * | 2020-12-10 | 2022-06-16 | 西安大医集团股份有限公司 | Image segmentation method and apparatus, image guidance system, and radiotherapy system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654902A (en) * | 2012-01-16 | 2012-09-05 | 江南大学 | Contour vector feature-based embedded real-time image matching method |
| CN102799898A (en) * | 2012-06-28 | 2012-11-28 | 浙江大学 | Efficient target identification processing method of high-resolution remote sensing image in context |
| CN103793679A (en) * | 2014-03-05 | 2014-05-14 | 江苏物联网研究发展中心 | One-dimensional bar code recognition method based on DaVinci technology |
| CN104700112A (en) * | 2015-02-02 | 2015-06-10 | 电子科技大学 | Method for detecting parasite eggs in excrement based on morphological characteristics |
| CN105005813A (en) * | 2015-06-26 | 2015-10-28 | 广州铁路职业技术学院 | Insect pest analyzing and counting method and system |
| CN105023272A (en) * | 2015-06-26 | 2015-11-04 | 广州铁路职业技术学院 | Crop leaf insect pest detection method and system |
| CN105488492A (en) * | 2015-12-25 | 2016-04-13 | 北京大学深圳研究生院 | Color image preprocessing method, road identification method and related device |
| JP2016071423A (en) * | 2014-09-26 | 2016-05-09 | 株式会社Ｓｃｒｅｅｎホールディングス | Imaging method, imaging device, substrate processing device and inspection device |
-
2016
- 2016-06-23 CN CN201610463501.2A patent/CN107545572A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654902A (en) * | 2012-01-16 | 2012-09-05 | 江南大学 | Contour vector feature-based embedded real-time image matching method |
| CN102799898A (en) * | 2012-06-28 | 2012-11-28 | 浙江大学 | Efficient target identification processing method of high-resolution remote sensing image in context |
| CN103793679A (en) * | 2014-03-05 | 2014-05-14 | 江苏物联网研究发展中心 | One-dimensional bar code recognition method based on DaVinci technology |
| JP2016071423A (en) * | 2014-09-26 | 2016-05-09 | 株式会社Ｓｃｒｅｅｎホールディングス | Imaging method, imaging device, substrate processing device and inspection device |
| CN104700112A (en) * | 2015-02-02 | 2015-06-10 | 电子科技大学 | Method for detecting parasite eggs in excrement based on morphological characteristics |
| CN105005813A (en) * | 2015-06-26 | 2015-10-28 | 广州铁路职业技术学院 | Insect pest analyzing and counting method and system |
| CN105023272A (en) * | 2015-06-26 | 2015-11-04 | 广州铁路职业技术学院 | Crop leaf insect pest detection method and system |
| CN105488492A (en) * | 2015-12-25 | 2016-04-13 | 北京大学深圳研究生院 | Color image preprocessing method, road identification method and related device |
Non-Patent Citations (1)
| Title |
|---|
| 丁莹，杨华民，范静涛: "《复杂环境运动目标检测技术及应用》", 31 January 2014, 北京：国防工业出版社 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108304910A (en) * | 2018-02-27 | 2018-07-20 | 华北理工大学 | A kind of slip sheet of air conditioner compressor method of counting based on machine vision |
| CN108765516A (en) * | 2018-03-26 | 2018-11-06 | 阳光暖果（北京）科技发展有限公司 | A kind of face region colouring method based on thumbnail |
| CN109907535A (en) * | 2018-07-26 | 2019-06-21 | 永康市柴迪贸易有限公司 | Corner bookcase based on worm analysis |
| WO2020140826A1 (en) * | 2019-01-02 | 2020-07-09 | 平高集团有限公司 | Infrared temperature sensing method employing contour extraction for target object, device, and storage medium |
| CN112215856A (en) * | 2020-10-20 | 2021-01-12 | 歌尔光学科技有限公司 | Image segmentation threshold determination method and device, terminal and storage medium |
| CN112215856B (en) * | 2020-10-20 | 2023-08-08 | 歌尔光学科技有限公司 | Image segmentation threshold determining method, device, terminal and storage medium |
| WO2022120714A1 (en) * | 2020-12-10 | 2022-06-16 | 西安大医集团股份有限公司 | Image segmentation method and apparatus, image guidance system, and radiotherapy system |
| CN116635888A (en) * | 2020-12-10 | 2023-08-22 | 西安大医集团股份有限公司 | Image segmentation method and device, image guidance system and radiotherapy system |
| CN114299494A (en) * | 2022-01-20 | 2022-04-08 | 广东省农业科学院动物科学研究所 | Method and system for detecting worm oval characteristics of aquatic product image |
| CN114299494B (en) * | 2022-01-20 | 2022-07-22 | 广东省农业科学院动物科学研究所 | Method and system for detecting worm-egg-shaped characteristics of aquatic product image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107545572A (en) | The treating method and apparatus of target in image | |
| CN106651872B (en) | Pavement crack identification method and system based on Prewitt operator | |
| JP7198577B2 (en) | Image analysis method, device, program, and method for manufacturing trained deep learning algorithm | |
| JP7076698B2 (en) | Image analysis method, image analysis device, program, learned deep learning algorithm manufacturing method and learned deep learning algorithm | |
| CN107507173B (en) | No-reference definition evaluation method and system for full-slice image | |
| Ranjan et al. | Detection and classification of leaf disease using artificial neural network | |
| JP2024038396A (en) | Image analysis method, image analysis device, and program | |
| KR20080016847A (en) | Methods of chromogen separation-based image analysis | |
| CN108181316B (en) | Bamboo strip defect detection method based on machine vision | |
| CN108109157B (en) | Rock mass evaluation analysis method based on digital panoramic borehole image | |
| CN104794502A (en) | Image processing and mode recognition technology-based rice blast spore microscopic image recognition method | |
| CN107545557A (en) | Egg detecting method and device in excrement image | |
| CN110502972A (en) | A kind of pavement crack segmentation and recognition methods based on deep learning | |
| CN111077150A (en) | Intelligent excrement analysis method based on computer vision and neural network | |
| CN113327255A (en) | Power transmission line inspection image processing method based on YOLOv3 detection, positioning and cutting and fine-tune | |
| US10482605B2 (en) | Cell area determination method, cell imaging system, and cell image processing apparatus | |
| CN111368669B (en) | Nonlinear optical image recognition method based on deep learning and feature enhancement | |
| CN109146888A (en) | A kind of soil cracking gap fractal dimension calculation method based on difference coefficient analysis | |
| CN106558044A (en) | The resolution measuring method of image module | |
| CN110969583A (en) | Image background processing method and system | |
| CN112763466B (en) | Method for identifying phase state distribution characteristics of polymer modified asphalt | |
| Raof et al. | Image segmentation of Ziehl-Neelsen sputum slide images for tubercle bacilli detection | |
| CN107239761B (en) | Fruit tree branch pulling effect evaluation method based on skeleton angular point detection | |
| Salih et al. | Adaptive local exposure based region determination for non-uniform illumination and low contrast images | |
| CN108830834A (en) | A kind of cable-climbing robot video artefacts information automation extraction method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |