JP4715703B2 - Optical flare inspection apparatus and inspection method - Google Patents

Description

  The present invention relates to an optical flare inspection apparatus and an inspection method for evaluating an optical flare that is a problem in a projection image obtained by a projector in an inspection stage of the projector.

  In order to enlarge and project a desired image on a screen, a projector is generally used. For example, as shown in FIG. 1 of Patent Document 1, such a projector includes a projection light source, a color separation optical system, a light modulation device, a cross dichroic prism, and a projection lens. Here, the projection light source is a mercury or halogen lamp or the like, and emits light that is not monochromatic. The color separation optical system separates this light into three color light beams R (Red), G (Green), and B (Blue) by a dichroic mirror. The light modulation device is provided for each of the three separated RGB light beams, and modulates each light beam by, for example, a liquid crystal panel according to the RGB information of the projected image. The cross dichroic prism forms an optical image by combining three types of modulated light beams. Finally, the projection lens enlarges this optical image, enlarges the desired image, and projects it onto the screen.

  In such a projector, the occurrence of optical flare (hereinafter referred to as flare) in the projected image becomes a problem. A flare is an area that appears around the original light emitting area (bright area) in the projected image and is displayed brighter than the original luminance. Flares are generated due to the optical elements of the projector, including the projection lens, and the projected image in which the flare exists does not accurately reproduce the original image. Therefore, in such a projector manufacturing / inspection process, a technique for detecting and evaluating flare in a projected image on a screen is important. Such techniques include those described in Patent Document 1 and Patent Document 2, for example. Patent Document 1 describes a method for inspecting a projection lens when the cause of occurrence of flare is in the projection lens, and Patent Document 2 describes an inspection method for the entire projector. At the time of flare inspection of such a projector, an inspection pattern for inspecting flare is used, and a clear projected image of the inspection pattern is obtained using a projector or a projection lens to be inspected. For this purpose, focus adjustment is performed in which each liquid crystal panel is accurately arranged at the focus position of the projection lens. Also, the pixels of each liquid crystal panel are aligned, the relative positions of the liquid crystal panels are aligned, the optical axis entering the projection lens is aligned, and each liquid crystal panel is joined to an optical device including a dichroic prism. Thereafter, a predetermined inspection pattern is projected onto the screen, and for example, the amount of flare caused by the projection lens is visually measured. For example, when an optical device including three liquid crystal panels and a cross dichroic prism is to be evaluated, first, a light beam is incident on an image forming area of each liquid crystal panel and is incident on a light incident end surface of the cross dichroic prism. The light beam emitted from the projector is projected onto the screen. A projected image of the projected inspection pattern is captured by an imaging device (for example, a CCD camera), the projected actual image is captured by the image capturing device, and the projected image is enlarged and displayed by the image processing unit. Alternatively, an operator visually determines the flare amount on the screen, and performs a flare inspection by visual measurement.

  When a CCD camera is used for imaging, for example, a projection image is captured when a standard projection lens is pre-installed in the inspection apparatus as a master lens, or when a good projector with little flare is generated, and the imaging is performed. Data can be stored as a sample. In this case, the flare can be determined by comparing the sample with a projection image when the projection lens to be evaluated is used. Also in this case, the projected image is enlarged and displayed, and the operator makes a comparative determination by visual observation.

  Thus, in the above flare inspection method, after a complicated adjustment of the apparatus, a human (operator) finally detects the flare visually.

JP 2001-4492 A JP 2004-45809 A

  In the flare detection method visually observed by humans, for example, a judgment that depends on the subjective and sense of the inspection operator is made, such as recognizing flare up to about 50% of the brightness of the light emitting area of the projected image. . Accordingly, misjudgment is likely to occur due to variations among workers, quality variations due to eye fatigue, and differences in the appearance of projected images.

  In addition, for inspection workers with excellent visual skills, it is experienced that bright areas in a projected image can be classified into four types according to luminance, for example, an original light emitting area and three types of (dark, thin, and extremely thin) flare. Known. In this case, it has been difficult to specifically describe these boundaries and numerical standards for each type of flare used as the reference. Further, the absolute value of the brightness and the like changes depending on how the projector is used. For example, if the brightness of the light source fluctuates, the luminance fluctuates, and even with a projector having the same specifications, the luminance and flare shape of the projected image may vary depending on the production lot. Furthermore, the projection magnification also changes depending on the mode of use of the projector, and the size and brightness of the entire projected image vary depending on this magnification. It was difficult to set a robust judgment standard that universally adapts to all of these situations, matches the sense of visual judgment, and identifies the above flare types. For this reason, it is extremely difficult to automate flare inspections, and ultimately evaluation based on human senses was mainly performed.

  On the other hand, if the flare is determined based on the optical measurement value, it is possible to automate the inspection, but ultimately the appearance of the projected image by human eyes is also important. It is a fact. However, there is no means that specifically associates the visual sensation of each flare type with the optical numerical value indicating the physical characteristics of the projector optical system measured on the screen. For this reason, in flare, for example, it has been difficult to associate an optical value such as illuminance with a sensory value obtained by human visual sense. Therefore, even if the measured optical numerical value is good, it may actually be a projected image that does not look good for humans, and an appropriate evaluation cannot always be made by evaluating the optical numerical value. That is, it has been particularly difficult to perform a flare inspection that is compatible with human senses regardless of human vision.

  As described above, in the conventional flare inspection method, it is difficult to automatically perform evaluation suitable for human senses. For this reason, it has been difficult to perform flare inspection with high reproducibility and high efficiency.

  The present invention has been made in view of the above problems, and an object thereof is to provide an optical flare inspection method and an inspection apparatus for a projector that solve the above-described problems.

In order to solve the above problems, the present invention has the following configurations.
The gist of the invention described in claim 1 is that a projection image obtained by enlarging and projecting an inspection pattern having light emitting pixels by a projector is formed on a screen, and an original light emitting area and flare recognized in the projection image are classified. An optical flare inspection apparatus for inspecting, comprising: an image capturing device that captures an image obtained by capturing the projection image as a projected actual image; and an image processing device that analyzes the projected actual image, the image processing device comprising: Create an appearance frequency distribution of the brightness of each pixel in the projected actual image, and determine whether each pixel is included in the original light emitting area or flare from the brightness value of each pixel in the appearance frequency distribution It exists in the optical flare inspection apparatus characterized by having the flare detection and classification part to determine.
The gist of the invention described in claim 2 is that, when the flare is classified into a plurality of types, the flare detection / classification unit determines whether each flare is based on the luminance value of each pixel in the appearance frequency distribution. The optical flare inspection apparatus according to claim 1, wherein a determination is made as to which of the plurality of types of flare a pixel belongs to.
The gist of the invention described in claim 3 is that, in the luminance frequency distribution, a plurality of luminance ranges are determined by combinations of peaks and valleys, and the luminance of each pixel in the projected actual image is the plurality of luminances. 3. The method according to claim 2, wherein by determining which of the ranges is included, it is determined whether each pixel in the projection image belongs to an original light emitting region or the plurality of types of flares. It exists in an optical flare inspection device.
The gist of the invention of claim 4 is that, when the flare is classified into N types, the image processing apparatus calculates an average luminance value of an area composed of pixels belonging to the i-th flare as I _FL (i) (here Where i is an index corresponding to the type of flare, 1 ≦ i ≦ N), and the average luminance value in the original light emitting region is P ₀ .
4. The optical flare inspection apparatus according to claim 2, wherein the flare standard amount FL (i) obtained by the above is calculated.
The gist of the invention described in claim 5 is that the image processing apparatus uses the luminance I of each pixel in the actual projection image as the luminance average value in the original light emitting area as P ₀ .

The flare sensation amount THfl obtained in step (1) is obtained, the minimum luminance of the pixels in the plurality of types of flares is MIN (I _FL (i)), and the maximum luminance of the pixels belonging to the plurality of types of flares is MAX (I _FL (i) ) As said THfl
5. The optical flare inspection apparatus according to claim 4, wherein it is determined which of the plurality of types of flares each pixel belongs to by determining whether the pixel is in the range of 5.
The gist of the invention described in claim 6 is that the image processing apparatus normalizes and quantifies the flare width with a width in the projected actual image corresponding to a minimum unit of light emitting pixels in the inspection pattern as a reference unit. 6. The optical flare inspection apparatus according to claim 1, further comprising a pitch calculation unit.
The gist of the invention described in claim 7 is that the image processing device calculates the reference unit from an interval between the light emitting regions in the projected actual image and a number of pixels between the light emitting pixels in the inspection pattern. An optical flare inspection apparatus according to claim 6.
The gist of the invention described in claim 8 is that, when there are a plurality of types of flares, the image processing apparatus calculates an average brightness value in a region composed of pixels belonging to the plurality of types of flares as I _FL (i) (here Where i is an index corresponding to the type of flare, 1 ≦ i ≦ N, N is the number of flare types), and the average luminance value in the original light emitting region is P ₀ .
FL (i) calculated in step 1 is calculated as the first flare standard amount, and the amount obtained by normalizing the sum of the widths of the plurality of types of flare in the reference unit is calculated as the second flare standard amount. 8. The image processing apparatus according to claim 6, wherein the processing device includes an image quality determination unit that determines whether the projection performance of the projector is good or bad from the first flare standard amount and the second flare standard amount. It exists in an optical flare inspection device.
The gist of the invention described in claim 9 is that the projection image is a color image, and the image processing device converts the projection actual image into a monochrome image and creates the appearance frequency distribution. Item 10 is the optical flare inspection apparatus according to any one of Items 1 to 8.
The gist of the invention described in claim 10 is that, in the conversion to the monochrome image, the luminance of each pixel in the projected actual image is captured as I _R , I _G , and I _B for each of the R, G, and B color components, respectively. , The luminance I of each pixel in the projected actual image,
The optical flare inspection apparatus according to claim 9, wherein:
The gist of the invention described in claim 11 is that a projection image obtained by enlarging and projecting an inspection pattern having light emitting pixels by a projector is formed on a screen, and an original light emitting area and flare recognized in the projection image are classified. An optical flare inspection method for inspecting, comprising: a step of capturing an image obtained by capturing the projection image as a projection real image; and a step of analyzing the projection real image. In the step of analyzing the projection real image, Create an appearance frequency distribution of the brightness of each pixel in the projected actual image, and determine whether each pixel is included in the original light emitting area or flare from the brightness value of each pixel in the appearance frequency distribution An optical flare inspection method is characterized by determining.
The gist of the invention described in claim 12 is that when the flare is classified into a plurality of types, in the step of analyzing the projected actual image, from the luminance value of each pixel in the appearance frequency distribution, 12. The optical flare inspection method according to claim 11, wherein it is determined which of the plurality of types of flare each of the pixels belongs to.
According to a thirteenth aspect of the invention, in the luminance frequency distribution, a plurality of luminance ranges are determined by combinations of peaks and valleys, and the luminance of each pixel in the projected actual image is the plurality of luminances. 13. The method according to claim 12, wherein by determining which of the ranges is included, it is determined whether each pixel in the projection image belongs to an original light emitting region or the plurality of types of flares. Lies in optical flare inspection methods.
According to a fourteenth aspect of the present invention, in the step of analyzing the projected actual image, when there are a plurality of types of flares, an average luminance value in a region composed of pixels belonging to the plurality of types of flares is expressed as I _FL ( i) (where i is an index corresponding to the type of flare, 1 ≦ i ≦ N, and N is the number of flare types), and the average luminance value in the original light emitting region is P ₀ .
FL (i) calculated in step 1 is calculated as the first flare standard amount, and the amount of the flare width normalized using the width in the projected actual image corresponding to the minimum unit of the light emitting pixels in the inspection pattern as a reference unit 14 is calculated as a second flare standard amount, and whether the projection performance of the projector is good or bad is determined from the first flare standard amount and the second standard amount. Or the optical flare inspection method according to claim 1.
The gist of the invention described in claim 15 is that the projected image is a color image, and in the step of analyzing the projected actual image, the projected actual image is converted into a monochrome image to create the appearance frequency distribution. The optical flare inspection method according to any one of claims 11 to 14, wherein the optical flare inspection method is characterized.

  According to the present invention, since the correspondence between the flare amount sensuously determined by human eyes and the optical numerical value of the product becomes easy, it is easy to automate the inspection of the optical flare suitable for the human sense. . Thereby, it becomes possible to perform the inspection of the optical flare appropriately and efficiently.

  Hereinafter, an optical flare inspection apparatus and an inspection method for a projector according to embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of an optical flare inspection apparatus for a projector according to the present embodiment. In the optical flare inspection apparatus 1, optical flare (hereinafter referred to as flare) generated when a projector 3 equipped with a projection lens 2 is used is inspected. The video pattern 5 sent from the pattern signal generator 4 is sent to the projector 3. The image pattern 5 is an image of an inspection pattern having light emitting pixels so that flare can be easily evaluated as will be described later. As a result, the inspection pattern is enlarged and projected onto the screen 6 by the projector 3, and a projected image of the inspection pattern is formed on the screen 6. This projection image is color-captured by the imaging device 7, and this imaging data is captured by the image capturing device 8 as a projection actual image 9 separated into R, G, and B components. Further, an image processing apparatus 10 is connected to appropriately perform image processing on the imaged data and inspect the flare. In this inspection apparatus 1, the projection lens 2 in the projector 3 is removable. For example, when the combination of the projection optical system and the color optical system mounted on the projector 3 is the main cause of flare, the projection lens 2 is replaced to perform a more appropriate evaluation of the projector 3. Can do. Further, even when the projection lens 2 is a main cause of flare, it is possible to perform a more appropriate inspection by exchanging it. In the projection image inspected by this inspection apparatus, the original light emitting area and flare on which the light emitting pixels in the inspection pattern are projected are recognized.

  As the imaging device 7, for example, a CCD camera can be used. As shown in FIG. 1, the imaging device is preferably fixed vertically at a position away from the display surface of the screen 6 by a distance d. As a result, it is possible to capture an image without interrupting the inspection pattern in the projected image in the vicinity of the image end of the projected image where flare is easily measured.

  FIG. 3 shows an example of a projected image obtained by projecting the inspection pattern onto the screen 6 by the projector 3 in the inspection apparatus 1. As will be described later, the inspection pattern preferably includes a plurality of light emission patterns including light emission pixels arranged in the vertical or horizontal direction so that flare can be easily detected. The area that appears bright in FIG. 3 is composed of a light emitting area corresponding to the light emitting pixel in the inspection pattern and a flare area formed around the light emitting area. That is, the light emitting area is an area where the light emitting pixels are projected, and the flare is not an area where the light emitting pixels are projected, but is an area that appears bright around the light emitting area. When there are a plurality of light emitting areas in the projected image, it is preferable that the plurality of light emitting patterns in the inspection pattern are spaced apart by a plurality of pixels so that the flare associated with each light emitting area does not overlap. The inspection pattern used here is a pattern in which light emitting pixels are arranged in the vertical direction.

  FIG. 4 is an example of a projection image of an actual green (G) projection pattern captured by the image capturing device 8 as a color image. From the actual projection image 9 of the projection image captured by the image capturing device 8, as described later, a luminance threshold value that becomes a boundary between a plurality of types of flare is obtained from the luminance appearance frequency information of the actual projection image 9. Based on this, the flare inspection is performed in the partial selection image shown by the rectangle in FIG. 4 selected by the operator. If a pattern recognition means using a computer is used, it is easy to automatically select this partial selection image from the projected actual image.

  Next, the configuration of the image processing apparatus 10 and the operation of the flare inspection process will be described. As shown in FIG. 2, the image processing apparatus 10 includes an image preprocessing unit 21, a flare detection / classification unit 22, a flare calculation unit 23, a pitch calculation unit 24, and an image quality determination unit 25.

The image pre-processing unit 21 converts the luminance of each pixel in the projection actual image 9 which is a color image based on the calculation formula I = (28 × I _R + 77 × I _G + 151 × I _B ) / 256 to obtain a color image. Is converted into a monochrome image (hereinafter, this operation is referred to as monochrome conversion). Here, I _R , I _G , and I _B are the luminance values of the color components of the R, G, and B color actual images of each pixel in the color projection actual image 9, and I is the luminance value after monochrome conversion at the same position. Become. When the image captured by the imaging device 7 is a monochrome image, or when the projected image on the screen is a monochrome image, monochrome conversion is not necessary.

  The flare detection / classification unit 22 creates the appearance frequency distribution of the luminance of each pixel in the projection actual image 9, identifies the original light emitting area and flare from the result, and further creates data quantifying these. Here, as the luminance, the luminance after monochrome conversion may be used, or the luminance for each of the R, G, and B components may be used. An example of the appearance frequency distribution is shown in FIG. In FIG. 5, the horizontal axis represents the monochrome-converted luminance I, and the vertical axis represents the appearance frequency of the luminance of each pixel in the projected actual image. Here, the luminance is expressed in 256 gradations from 0 to 255. As can be seen from FIG. 5, when the entire pattern is converted into a monochrome pattern and the inspection pattern region range is narrowed by increasing the threshold value by binarization processing by image processing, the region corresponding to the flare region has an appearance frequency of luminance. It always coincides with the region where the appearance frequency of the luminance is small on the distribution. The region with a low appearance frequency is recognized as a light emission influence region in FIG. 5, and the region with a high appearance frequency on the left side is recognized as a background region. The boundary between the light emission affected area and the background area can be easily calculated by defining, for example, a point where the appearance frequency is 5% of the peak of the appearance frequency. In FIG. 5, this boundary is a point having a luminance of 64. The original light emission area and flare are included in the light emission influence area. The flare detection / classification unit 22 determines whether each pixel is included in the original light emission region or flare from the luminance value of each pixel in the luminance appearance frequency distribution in the light emission influence region. .

  The original light emitting area and flare classification method will be described in detail. It is empirically known that flares are classified into three types of flares, dark flares, thin flares, and ultrathin flares according to their brightness. A peak corresponding to each of these types and each of the original light-emitting areas exists in the luminance appearance frequency distribution in the light emission-affected area. FIG. 6 is an enlarged view (left) of the four types of states in the partial selection region of FIG. 4 and the appearance frequency distribution (right) of the luminance in the light emission affected region. Here, (a) shows the original light emitting region, (b) shows the deep flare, (c) shows the thin flare, and (d) shows the ultrathin flare. In addition, the luminance appearance frequency distribution (right) shows the relationship with the flare boundary determined by the operator through visual inspection. A plurality of peaks and valleys exist in the waveform of the luminance appearance frequency distribution. The boundaries of the various flares that the observer has sensuously determined by visual inspection correspond to the regions of combinations of peaks and valleys of a predetermined order in this waveform. First, in FIG. 6A, a region having a higher luminance (right side) than a left trough adjacent to the highest luminance peak (the rightmost peak) corresponds to the original light emitting region. That is, the original light emitting region that is not flare corresponds to a region including a peak with the highest luminance. In FIG. 6 (b), a region that includes the peak with the second highest luminance (second from the right) and is sandwiched between valleys adjacent to both sides thereof corresponds to a dark flare in the flare. In FIG. 6 (c), a region including two peaks sandwiched between a valley portion adjacent to the right side of the peak having the smallest luminance (the leftmost) and the dark flare region, Corresponds to thin flare. In FIG. 6 (d), an area having a lower luminance (on the left side) than the thin flare corresponds to an ultrathin flare. Thus, the peak corresponding to the light emitting region and the three types of flare is determined. As described above, when the flare is classified into a plurality of types, the flare detection / classification unit 22 determines whether each pixel is a plurality of types of flare based on the luminance value of each pixel in the appearance frequency distribution. Whether it belongs can be determined. Specifically, as described above, a plurality of luminance ranges are determined by combinations of peaks and valleys, and each pixel is determined by determining which of the plurality of luminance ranges the luminance of each pixel is included in. Can be determined to belong to the original light emitting region or a plurality of types of flares.

  For example, if only one peak on the rightmost side is recognized in the light emission influence area, only the original light emission area exists and flare cannot be recognized. That is, as described above, the presence / absence of flare can be determined by analyzing the luminance frequency distribution, and if flare exists, it can be classified.

  When the projection lens performance, the combination of projector optical devices, the intensity of the light source, the distance between the projection lens and the screen, and the like vary, the brightness and size of the projected image change. However, the combination of peaks and valleys in the areas corresponding to the various flares described above is uniquely determined for each projection lens type. Therefore, by searching for the peaks and valleys corresponding to various flares for each type of projection lens having the same optical characteristics, and the combination position thereof, it is possible to always automatically classify and extract the flare boundary region that matches the visual judgment. Become. Furthermore, as shown in FIG. 7, even when there is a product variation in the characteristics of the same lens, the waveforms of the appearance frequency distribution of the luminance are similar to each other. For this reason, even if the absolute value of the luminance and the size of each flare in the luminance appearance frequency distribution change greatly, the order of each peak and each valley and the combination thereof do not change. For this reason, it is possible to implement flare automatic classification robustly. In this way, the flare detection / classification unit 22 stably and automatically extracts and classifies flare detection and flare types using the peak / valley rank in the luminance frequency distribution and the combination information thereof. can do.

  The amount of quantitative evaluation of the flare detected by the flare detection / classification unit 22 includes the first flare standard FL (i) indicating the average luminance of the various flares and the second flare indicating the size of the flare. There is a standard FLQ. As described above, the absolute luminance of the real image and the size (magnification) of the projected image vary depending on the usage mode of the projector, but these amounts need to be universal amounts that do not vary depending on the usage mode. is there.

The flare calculation unit 23 calculates FL (i) that is the first flare standard. For this reason, first, the average luminance value I _FL (i) in the area for each type of flare that is the combination position of the peak and valley of the appearance frequency data of the luminance that forms the boundary of each detected flare type is calculated. Next, the first flare standard FL (i) is calculated by the following _{equation 7} from I _FL (i) and the average luminance P ₀ of the light emitting pixel region in the inspection pattern. Here, i is an index indicating a flare type (including a light emitting area and a non-light emitting area), and 0: light emitting area, 1: dark flare, 2: thin flare, 3: very thin flare, 4: no light emitting area. Correspond. The non-light emitting area is equivalent to the background area. Here, although three types of flare are used, when N types of flare are generally recognized due to the characteristics of the projection lens or the like, the light emission region is when i = 0, and the flare is when 1 ≦ i ≦ N. , I = N + 1 corresponds to a non-light emitting region.

  FIGS. 8A, 8B, and 8C are examples of projected images obtained by projecting red (R) test patterns having different brightnesses. FIG. 9 is an example in which the actual projection image in these cases is converted into a monochrome image and quantified using the first flare standard FL (i) calculated in Equation 7. This first flare standard FL (i) is a contrast index for various flares when the average luminance of the light emitting pixel region is 100% and the average luminance of the non-light emitting region is 0%. In FIG. 9, the horizontal axis represents the flare type (i), and the vertical axis represents the first flare standard FL (i). In FIG. 8, a (when bright) is indicated by a circle, b (intermediate) is indicated by an x, and c (when dark) is indicated by a Δ. A value at the center position of each flare type is a representative value of the first flare standard FL (i) of each flare, white display indicates an upper limit and a lower limit, and black display indicates an average value. Thus, even if the absolute values of the brightness are greatly different as shown in FIGS. 8A to 8C, the first flare standard FL (i) according to Equation 7 is a substantially constant value for each flare type. Take. Therefore, the first flare standard FL (i) does not depend on the apparent brightness, and is a value specific to each flare type. In other words, it is a universal amount that does not depend on the mode of use of the projector.

Similarly to FL (i), the luminance I of each pixel in the equi-photorealistic image can be converted into the flare sensation amount THfl by Equation 8.

According to this, the flare sense amount THfl is 100 in the light emitting area and 0 in the non-light emitting area, and takes an intermediate value in each flare area. For this reason, THfl does not change even if the absolute value of the luminance changes due to changes in the usage of the projector, and is an index for identifying the light emitting area and the flare area. Unlike luminance, the non-light emitting area unrelated to flare is set to 0, and the light emitting area is set to 100, which is an index close to human sense. That is, by performing a flare inspection using THfl, it is possible to adapt to human senses and universally cope with fluctuations in luminance and the like. From the THfl value of each pixel, it can be determined by Equation 9 whether the pixel belongs to the light emitting region or the three types of flare. That is, when THfl is in the range of Formula 9, it belongs to one of the light emitting area corresponding to i, the three types of flare, and the non-light emitting area. Here, it is assumed that the minimum luminance of the pixel in the index i region is MIN (I _FL (i)), and the maximum luminance of the pixel in the index i region is MAX (I _FL (i)).
In FIG. 9, the upper limit value, lower limit value, and average value for each flare type indicate the upper limit boundary value, lower limit boundary value, and flare width center position value of the flare amount of each flare type. For each projection lens characteristic, the flare type is discriminated / classified by the flare detection / classification unit 22, and the flare calculation unit 23 stores the existence range of various flares as numerical information of the upper limit boundary value and the lower limit boundary value of the flare amount. .

  The pitch calculation unit 24 calculates the second flare standard FLQ. The second flare standard FLQ is calculated using, for example, an inspection pattern having a plurality of patterns in which light emitting pixels are arranged in the vertical direction. FIG. 10 is a conceptual diagram for illustrating a pitch calculation method performed by the pitch calculation unit 24 in this case. FIG. 10 shows an example of a projected real image when such an inspection pattern is used, and light emitting areas exist at both left and right ends corresponding to the light emitting pixels in the inspection pattern. In this case, for example, it is preferable that the intervals between the light emitting pixels in the inspection pattern are sufficiently separated so that the flare generated due to the light emitting region at one end of the left and right does not overlap with the light emitting region at the other end. The arrangement of the light emitting pixels may be formed as a horizontal line instead of a vertical line. The number of pixels corresponding to the interval between the light emitting pixels in the inspection pattern is preset as a design value (M). In an image obtained by converting the actual projection image formed by such an inspection pattern into a monochrome image, as described above, an area where the luminance is equal to or higher than the lowest luminance value of the light emitting area is identified as the light emitting area. In the example of FIG. 10, the center of gravity in the horizontal direction is obtained in the left and right light emitting areas, and this barycentric position is used as the center of the light emitting area, the coordinates of the left light emitting area center are X1 and the coordinates of the right light emitting area center are X2.

The coordinate values X1 and X2 of the center of gravity of the two light emitting areas are obtained, and further, based on the number of pixels M between the light emitting pixels, the reference unit ΔPd that is the inspection pattern pixel pitch on the projected actual image is calculated based on Expression 10. Calculated. ΔPd is a unit in the actual projection image corresponding to the pixel in the inspection pattern. When the projection magnification of the projector is different in the vertical and horizontal directions of the inspection pattern, different values may be set in the respective directions.

  Next, the pitch calculation unit 24 identifies various flare areas in the same manner as the above-described method of identifying the light emitting area in the monochrome image of the actual projection image. In this way, as shown in FIG. 11, the contour line that forms the boundary line of each flare type can be obtained.

The sum of the average flare widths calculated from the position of the contour line is calculated as the relative flare length from the light emitting region end. In FIG. 11, the region of width L1 is a dark flare, the region of width L2 is a thin flare, and the region of width L3 is an ultrathin flare. The second flare standard FLQ based on the sum of the various flare widths Li (i = 1, 2, 3) is calculated based on Equation 11.

  FLQ is the size of the entire flare area with the reference unit ΔPd as a basic unit. Even when the magnification (size) and resolution of the projected actual image change depending on the mode of use of the projector, FLQ is a universal amount because it is normalized by ΔPd. Similarly, if the size of each type of flare area is expressed in units of ΔPd (L1 / ΔPd, L2 / ΔPd, L3 / ΔPd), the amount does not depend on these distances. FIG. 12 is an example of a projected actual image when the reference unit ΔPd on the projected actual image is obtained by the above method. FIG. 13 shows an example in which a dimensionless quantity obtained by normalizing L1, L1 + L2, and L1 + L2 + L3 by ΔPd, which are distances from the light emission region of the three types of flare ends, is shown in order. In FIG. 13, the rightmost data is the FLQ.

  Subsequently, the set image quality determination unit 25 stores in advance FL (i), FLQ, and the like in the projection image by the projector, which is a reference for evaluation, and sets it as a threshold value for inspection. By measuring these amounts in the projection image by the projector to be inspected and comparing them with these threshold values, the quality determination of the projector to be inspected is automatically executed. It is also possible for a human to make a final determination without using the image quality determination unit 25. Even in this case, since the worker can use the values of FL (i) and FLQ, which are appropriately standardized values, as a criterion for determination, appropriate evaluation can be performed with high efficiency.

In this optical flare inspection apparatus, for example, the correspondence between the first flare standard and the value obtained by measuring the projection image obtained on the screen with an illuminometer can be taken. For example, in FIG. 14, the brightness (gradation) of each of the R, G, and B test patterns is sequentially changed and projected onto a screen by a projector, and an ND (neutral density) filter is attached to the lens of the image capturing device 8. a screen projected image is the result of examining the average brightness P ₀ of the optically measured emission region when imaged. Here, for convenience P _0, and again in terms of the transmittance of the more convenient ND filter as the optical design criteria indicates that the transmittance is larger average luminance higher. From the average luminance P ₀ of the light emitting area stored in the flare calculation unit 23, the transmittance is converted for each gradation of the inspection pattern, and when there is no data, the flare calculation unit 23 interpolates by the least square method or the like. Further, the luminance data P ₀ and the illuminance V obtained by directly measuring the light emitting area in the projected image of the illuminometer with the illuminometer are in a proportional relationship. For this reason, the first flare standard FL (i) expressed by the formula 7 expressed using luminance can be expressed by the illuminance V and the transmittance. FIG. 15 is an example of actual measurement values obtained by calculating the relationship between FL (i) and the transmittance for each of R, G, and B colors. In this way, the flare standard can be expressed by converting it into a convenient transmittance as an optical design guideline. This makes it possible to clarify the design guidelines of the projector and facilitate quality control.

It is a block diagram of the optical flare inspection apparatus which is embodiment of this invention. It is a block diagram of the image processing apparatus in the optical flare inspection apparatus which is embodiment of this invention. It is an example of the projection image which projected the test | inspection pattern on the screen. It is an example which selected the part used as a test | inspection area | region in a projection image. It is a figure which shows an example of the appearance frequency distribution of the brightness | luminance of each pixel in an equality real image. It is a figure of an example which shows the relationship between the light emission area | region, the range of three types of flare, and the appearance frequency distribution of a brightness | luminance. It is a figure which shows typically the relationship of the appearance frequency distribution of the brightness | luminance by the product dispersion | variation of the projection lens which has the same lens characteristic. It is an example of the projection image at the time of projecting the test pattern of red (R) color and changing the brightness of appearance in three main steps. It is a figure of an example which computed the 1st flare standard with respect to each flare type and the range of each flare type with the red test | inspection pattern which changed the brightness. It is a figure which shows typically the method of calculating | requiring reference | standard unit (DELTA) Pd. It is a figure which shows the positional relationship of the flare width of each flare kind, and a light emission area | region. It is a figure which shows the Example which calculated | required reference | standard unit (DELTA) Pd. It is an example which calculated the quantity which normalized the position of each flare end in the standard unit, and the 2nd flare standard FLQ. It is a figure which shows an example which converted the brightness | luminance of the light emission area | region obtained by changing the brightness | luminance of R, G, and B in the test | inspection pattern into the transmittance | permeability of ND filter. It is a figure which shows the relationship between the 1st flare standard and the light transmittance of a ND filter at the time of calculating using the brightness | luminance information of each R, G, B in the result of FIG. 14 for the 1st flare inspection standard calculation.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Optical flare inspection apparatus 2 Projection lens 3 Projector 4 Pattern signal generator 5 Image pattern 6 Screen 7 Imaging device 8 Image capture device 9 Projected real image 10 Image processing apparatus 21 Image pre-processing part 22 Flare detection and classification part 23 Flare calculation part 24 Pitch calculation unit 25 Image quality determination unit

Claims (15)

An optical flare inspection apparatus for forming a projection image obtained by enlarging and projecting an inspection pattern having a light emitting pixel by a projector on a screen, and classifying and inspecting an original light emitting area and flare recognized in the projection image,
An image capturing device that captures an image obtained by capturing the projected image as a projected actual image; and an image processing device that analyzes the projected actual image;
The image processing apparatus creates an appearance frequency distribution of the luminance of each pixel in the projected actual image, and the pixels are compared with the original light emitting area and the flare from the luminance value of each pixel in the appearance frequency distribution. An optical flare inspection apparatus having a flare detection / classification unit for determining which of the two is included.   When the flare is classified into a plurality of types, the flare detection / classification unit determines whether each pixel is one of the plurality of types of flare based on the luminance value of each pixel in the appearance frequency distribution. The optical flare inspection apparatus according to claim 1, wherein the optical flare inspection apparatus determines whether the optical flare belongs.   In the luminance frequency distribution, a plurality of luminance ranges are determined by combinations of peaks and valleys, and it is determined which of the plurality of luminance ranges the luminance of each pixel in the projected actual image is included in. The optical flare inspection apparatus according to claim 2, wherein each of the pixels in the projection image is determined to belong to an original light emitting area or the plurality of types of flare. The image processing apparatus includes:
When the flare is classified into N types, the luminance average value of an area composed of pixels belonging to the i-th flare is expressed as I _FL (i) (where i is an index corresponding to the type of flare and 1 ≦ i ≦ N), and the average luminance value in the original light emitting region is P ₀ ,
The optical flare inspection apparatus according to claim 2, wherein the flare standard amount FL (i) obtained by the calculation is calculated. From the luminance I of each pixel in the projected actual image, the image processing apparatus sets the average luminance value in the original light emitting area as P ₀ ,

The flare sensation amount THfl obtained in step (1) is obtained, the minimum luminance of the pixels in the plurality of types of flares is MIN (I _FL (i)), and the maximum luminance of the pixels belonging to the plurality of types of flares is MAX (I _FL (i) ) As said THfl
The optical flare inspection apparatus according to claim 4, wherein it is determined which of the plurality of types of flares each pixel belongs to by determining whether or not the pixel is in the range of 5.   The image processing apparatus includes a pitch calculation unit that normalizes and quantifies the flare width with a width in the projected actual image corresponding to a minimum unit of light emitting pixels in the inspection pattern as a reference unit. The optical flare inspection apparatus according to any one of claims 1 to 5.   The optical flare according to claim 6, wherein the image processing device calculates the reference unit from an interval between the light emitting regions in the projection actual image and a number of pixels between the light emitting pixels in the inspection pattern. Inspection device. The image processing apparatus includes:
When there are a plurality of types of flares, the average luminance value in a region composed of pixels belonging to the plurality of types of flares is expressed as I _FL (i) (where i is an index corresponding to the type of flare and 1 ≦ i ≦ N, N is the number of flare types), and the average luminance value in the original light emitting area is P ₀ .
FL (i) calculated in step 1 is calculated as the first flare standard amount,
Furthermore, the amount obtained by normalizing the sum of the widths of the plurality of types of flare in the reference unit is calculated as a second flare standard amount,
8. The image processing apparatus according to claim 6, further comprising an image quality determination unit configured to determine whether the projection performance of the projector is good or not based on the first flare standard amount and the second flare standard amount. The optical flare inspection apparatus as described.   The said projection image is a color image, The said image processing apparatus converts the said projection actual image into a monochrome image, and produces the said appearance frequency distribution, The any one of Claim 1 thru | or 8 characterized by the above-mentioned. The optical flare inspection apparatus as described. In the conversion to the monochrome image, the luminance of each pixel in the actual projection image is captured as I _R , I _G , and I _B for each color component of R, G, and B, and each pixel in the actual projection image is captured. Luminance I
The optical flare inspection apparatus according to claim 9, wherein An optical flare inspection method in which a projection image obtained by enlarging and projecting an inspection pattern having a light emitting pixel by a projector is formed on a screen, and an original light emitting area and flare recognized in the projection image are classified and inspected.
Including a step of capturing an image obtained by capturing the projection image as a projection real image, and a step of analyzing the projection real image,
In the step of analyzing the actual projection image, an appearance frequency distribution of the luminance of each pixel in the actual projection image is created, and from the luminance value of each pixel in the appearance frequency distribution, each pixel is originally An optical flare inspection method characterized by determining whether a light emitting area or a flare is included.   When the flare is classified into a plurality of types, in the step of analyzing the projection actual image, each pixel is determined by the plurality of types of flare from the luminance value of each pixel in the appearance frequency distribution. The optical flare inspection method according to claim 11, wherein the optical flare inspection method is determined.   In the luminance frequency distribution, a plurality of luminance ranges are determined by combinations of peaks and valleys, and it is determined which of the plurality of luminance ranges the luminance of each pixel in the projected actual image is included in. 13. The optical flare inspection method according to claim 12, wherein each pixel in the projection image is determined to belong to an original light emitting area or the plurality of types of flare. In the step of analyzing the projected real image,
When there are a plurality of types of flares, the average luminance value in a region composed of pixels belonging to the plurality of types of flares is expressed as I _FL (i) (where i is an index corresponding to the type of flare and 1 ≦ i ≦ N, N is the number of flare types), and the average luminance value in the original light emitting area is P ₀ .
FL (i) obtained in step 1 is calculated as the first flare standard amount,
An amount obtained by standardizing the width of the flare with the width in the projected actual image corresponding to the minimum unit of the light emitting pixels in the inspection pattern as a reference unit is calculated as a second flare standard amount,
14. The optical flare inspection method according to claim 11, wherein whether the projection performance of the projector is good or bad is determined from the first flare standard amount and the second standard amount.   The projection image is a color image, and in the step of analyzing the projection actual image, the projection actual image is converted into a monochrome image to create the appearance frequency distribution. The optical flare inspection method according to claim 1.