JP4544342B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus that performs image processing for selecting an image suitable for use in image processing from a plurality of images.

Conventionally, an image processing apparatus for extracting a similar image from a plurality of images is known (see, for example, Patent Document 1). The image processing apparatus extracts a feature amount from the input image, and searches the image database for an image having a feature amount similar to the key, using the extracted feature amount as a key. A similar image can be automatically extracted.
Japanese Examined Patent Publication No. 7-107683

  By the way, when performing image processing on such a plurality of similar images or a plurality of images acquired by including a common point, an image with the least disturbance is extracted from the plurality of images. However, image processing may be performed on an image with less disturbance. However, although the image processing apparatus disclosed in Patent Document 1 can extract similar images, it includes a common point among a plurality of similar images, for example, under a plurality of different environmental conditions. There has been a problem that it is not possible to select an image that is less affected by disturbances from a plurality of images acquired by imaging.

  Accordingly, an object of the present invention is to provide an image processing apparatus capable of selecting an image with less disturbance from a plurality of images acquired by capturing images including a common point.

  An image processing apparatus according to the present invention that has solved the above-described problems is an image input unit that inputs a plurality of images captured and acquired including a common point under a plurality of different environmental conditions, and stores a plurality of images. An image storage means, an environmental condition addition area extraction means for extracting an environmental condition addition area added according to the environmental conditions in each of the plurality of images, and an additional environmental condition criterion selection image based on the area of the environmental condition addition area And additional environmental condition criterion selection image selection means for selecting from a plurality of images stored in the image storage means.

  In the image processing apparatus according to the present invention, for a plurality of images acquired by capturing images including a common point under a plurality of different environmental conditions, an environmental condition addition is added according to the environmental conditions in each of the plurality of images. The area is extracted. Based on the area of the environmental condition addition region, an additional environmental condition reference selection image is selected from a plurality of images stored in the image storage means. For this reason, an image with little disturbance can be selected from a plurality of images acquired by capturing images including a common point.

  Here, the additional environmental condition criterion selection image selection means selects the environmental load condition small amount image from the plurality of images stored in the image storage means in which the environmental condition addition region has an area smaller than a predetermined area. It can be set as the aspect which is an image selection means. Alternatively, the additional environmental condition criterion selection image selecting unit selects the image having the smallest area of the environmental condition additional region as a minimum environmental load condition image from the plurality of images stored in the image storage unit. It can be set as the aspect which is an image selection means.

  In this way, an image with less disturbance can be selected by using the environmental load condition small amount image ring in which the area of the environmental condition addition region is smaller than the predetermined area or the image having the smallest area of the boundary condition addition region.

  Here, the environmental condition addition region extracting means may be configured to extract the environmental condition addition region by performing statistical processing on a plurality of images.

  Thus, by extracting the environmental condition addition region by statistical processing, the environmental condition addition region can be extracted easily and accurately.

The environmental condition addition region extracting means includes an edge image converting means for converting each of the plurality of images into an edge image, an edge image storing means for storing the plurality of edge images converted by the edge image converting means, Edge appearance frequency measuring means for measuring the number of edge appearances in a predetermined area in an edge image, and an edge for extracting an edge frequent appearance area that is a predetermined area in which the edge appearance frequency is equal to or greater than a predetermined threshold among the predetermined areas The frequent environment extraction means and the edge frequent area storage means for storing the edge frequent area are provided, and the additional environment condition criterion selection image selection means generates a high-frequency edge image based on each edge frequent area in a plurality of edge images. high frequency and edge image generation unit, based on the approximate degree of the high frequency edge image, a plurality of edge images that are stored in the edge image storage means for From among, and a approximate degree reference edge image extraction unit that extracts an approximate degree reference edge image, the additional environmental conditions minimal image selecting means as an additional environmental condition criteria selected image, an image corresponding to an approximate degree reference edge image Can be selected from a plurality of edge images stored in the edge image storage means .

  In the image processing apparatus according to the present invention, a high-frequency edge image is generated based on an edge frequent region in a plurality of edge images, and an approximation degree reference edge image is extracted based on the degree of approximation with the high-frequency edge image. Further, an image corresponding to the recent degree reference selection edge image is selected from the plurality of images stored in the image storage means as the additional environment condition reference selection image. For this reason, it is possible to select an image having a high ratio of the edge frequent appearance region as the additional environment condition criterion selection image. Therefore, it is possible to reduce the influence of noise other than the additional environmental conditions in a plurality of images captured and acquired including a common point, and select an image with less disturbance.

  Here, the approximation degree reference edge image extraction unit extracts, as the approximation degree reference selection edge image, a high approximation edge image that is an edge image whose approximation degree with the high-frequency edge image is a predetermined threshold value or more. can do. Alternatively, the approximate degree reference edge image extraction unit can extract an approximate edge image that is an edge image that is closest to the high-frequency edge image as the approximate degree reference selection edge image.

  As described above, the approximation degree reference selection edge image is an edge image that approximates a high-frequency edge image or a high-frequency edge image that is an edge image whose approximation degree with a high-frequency edge image is a predetermined threshold value or more. By using the approximate edge image, an image with less disturbance can be selected.

  Furthermore, the environmental condition addition region extraction unit includes an edge image conversion unit that converts each of the plurality of images into an edge image, and an edge amount calculation unit that calculates each edge amount for a predetermined region in each edge image. The additional environmental condition criterion selection image selection means includes an edge amount occupation degree reference selection edge image selection means for selecting an edge amount occupation degree reference selection edge image based on the edge amount occupation degree of a predetermined area in the edge image, and an addition An aspect comprising: an edge amount occupation degree reference selection image selection means for selecting an image corresponding to the edge amount occupation degree reference selection edge image from a plurality of images stored in the image storage means as the environmental condition reference selection image; can do.

  Thus, based on the edge amount occupation degree of the predetermined region in the edge image, the edge amount occupation degree reference selection edge image is selected, and the image corresponding to the edge amount occupation degree reference selection edge image is stored in the image storage means. By selecting from a plurality of images, it is possible to accurately detect a place that is not subject to fluctuations in the additional environmental conditions. As a result, an image with less disturbance can be selected with high accuracy from a plurality of images captured and acquired including a common point.

  Here, the edge amount occupation degree reference selection edge image is an edge amount occupation degree reference selection edge image in which an edge amount small edge image in which the edge amount of a predetermined region in the edge image is equal to or less than a predetermined occupation degree is selected. can do. Alternatively, the edge amount occupation degree reference selection edge image selection unit may select an edge image with a minimum edge amount that minimizes the edge amount of a predetermined region in the edge image as the edge amount occupation degree reference selection edge image. it can.

  As described above, as the edge amount occupation degree reference selection edge image, the edge amount of the predetermined area in the edge image is equal to or less than the predetermined occupation degree, and the edge amount of the predetermined area in the edge image is minimized. By using the edge image with the smallest edge amount, an image with less disturbance can be selected.

  According to the image processing device of the present invention, an image with less disturbance can be selected from a plurality of images captured and acquired including a common point.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted. For the convenience of illustration, the dimensional ratios in the drawings do not necessarily match those described.

  FIG. 1 is a block diagram of an image processing apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the image processing apparatus 1 according to the present embodiment includes an image database 11, a statistical processing unit 12, and a representative image selection unit 13. Further, the image processing apparatus 1 is connected with an imaging apparatus 10. The imaging device 10 is fixed, for example, at a fixed point outdoors, and performs imaging of the same location. The imaging device 10 transmits the captured image to the image processing device 1.

  The image processing apparatus 1 includes a CPU that performs arithmetic processing, a ROM and a RAM that are storage units, an input signal circuit, an output signal circuit, a power supply circuit, and the like. The image processing device 1 stores the image transmitted from the imaging device 10 in the image database 11. Background images are transmitted from the imaging device 10 at regular intervals, and the image database 11 stores a plurality of background images obtained by imaging the same location. A statistical processing unit 12 is connected to the image database 11.

  The statistical processing unit 12 reads a plurality of background images stored in the image database 11. Statistical processing is performed on the plurality of background images. The statistical processing unit 12 outputs statistical processing information related to the statistical processing result to the representative image selection unit 13. The representative image selection unit 13 selects a representative image from a plurality of background images stored in the image database 11 based on the statistical processing result output from the statistical processing unit 12.

  Next, a processing procedure in the image processing apparatus according to the present embodiment will be described. In the image processing apparatus 1 according to the present embodiment, an operation of selecting an image suitable for subsequent image processing from a plurality of images captured and acquired including a common spot under a plurality of different environmental conditions. I do. As an image suitable for subsequent image processing, it is considered that the image has the smallest influence of environmental conditions in the image.

  Now, the imaging device 10 is imaging the same point. In the images taken at the same point, the background is the same, but the environmental conditions such as time zone, weather, and lighting are different. Therefore, a plurality of images captured under different environmental conditions are transmitted to the image processing apparatus 1. The image processing apparatus 1 performs processing for selecting an image suitable for subsequent image processing from a plurality of images captured under the different environmental conditions.

  FIG. 2 is a flowchart showing a processing procedure in the image processing apparatus according to the present embodiment. As shown in FIG. 2, in the image processing apparatus 1, first, a background image transmitted from the imaging apparatus 10 is acquired (S1) and stored in the image database 11 (S2). Here, the image database 11 also stores background images transmitted from the imaging device 10 in the past. For this reason, the image database 11 stores a plurality of background images acquired by capturing images including a common spot under a plurality of different environmental conditions.

  Next, the statistical processing unit 12 reads a plurality of images stored in the image database 11 (S3). Here, for example, 100 images are read out by the statistical processing unit 12. Of course, an appropriate number of images can be read out.

  When the statistical processing unit 12 reads a plurality of images, the statistical processing unit 12 performs statistical processing on the read plurality of images (S4). The statistical processing here is performed by the following procedure, for example. First, for each pixel in the background image, the luminance of that pixel is obtained. Next, the average luminance of the pixels in the background image is obtained. Subsequently, the pixels are sorted depending on whether or not the luminance at each pixel is a predetermined value or more away from the average luminance.

  Then, a set of pixels whose luminance is a predetermined value or more away from the average luminance is extracted from all the pixels. A set of pixels whose luminance is equal to or higher than a predetermined value is an environmental condition addition region. Subsequently, with respect to the set of extracted pixels, the variance in all the pixels is obtained. Thereafter, the variance among all the pixels of the set of pixels whose luminance is separated from the average luminance by a predetermined value or more is obtained for each of the plurality of background images.

  Then, a background image having the smallest obtained variance is obtained from the plurality of background images. It is considered that the background image having the smallest variance among all the pixels of the set of pixels whose luminance is separated from the average luminance by a predetermined value or more is the image having the smallest environmental condition in the image. The statistical processing unit 12 outputs the statistical processing result to the representative image selection unit 13.

  The representative image selection unit 13 selects a representative image from the background images stored in the image database 11 based on the statistical processing result output from the statistical processing unit 12 (S5). The representative image selection unit 13 selects, as a representative image, a background image having the smallest variance among all the pixels in the set of pixels whose luminance is equal to or greater than a predetermined value as a result of the statistical processing output from the statistical processing unit 12. Thus, the image processing ends.

  As described above, in the image processing apparatus 1 according to the present embodiment, a background image having the smallest variance among all the pixels in a set of pixels whose luminance is a predetermined value or more away from the average luminance is selected from a plurality of background images. Selected. A background image having the smallest variance among all the pixels in a set of pixels whose luminance is more than a predetermined value away from the average luminance is considered to be a background image having the least influence of environmental conditions. For this reason, an image with the least disturbance can be selected from a plurality of images obtained by imaging the same spot. In the image processing apparatus 1 according to the present embodiment, when selecting a background image, the background image is subjected to statistical processing to extract an environmental condition addition region. For this reason, it is possible to easily and accurately extract the environmental condition addition region.

  Next, a second embodiment of the present invention will be described. FIG. 3 is a block diagram of an image processing apparatus according to the second embodiment. As shown in FIG. 3, the image processing apparatus 2 according to the present embodiment includes an image database 21, an edge image conversion unit 22, an edge image storage unit 23, and an edge appearance number counting unit 24. In addition, the image processing apparatus 2 includes an edge frequent area extraction unit 25, a high-frequency edge image generation unit 26, an approximate edge image selection unit 27, and a representative image selection unit 28. Furthermore, the imaging device 10 is connected to the image processing device 2.

  The image processing device 2 stores a plurality of background images transmitted from the imaging device 10 in the image database 21. The edge image conversion unit 22 performs edge processing on the background image stored in the image database 21 to generate an edge image. The edge image conversion unit 22 stores the generated edge image in the edge image storage unit 23. The edge image storage unit 23 stores a plurality of edge images generated by the edge image conversion unit 22.

  The edge appearance number counting unit 24 measures the number of edge appearances of the target pixel for a plurality of edge images stored in the edge image storage unit 23. The edge appearance count counting unit 24 outputs the measured edge appearance count to the edge frequent appearance region extraction unit 25.

  The edge frequent appearance area extraction unit 25 extracts an area where edges frequently appear in the edge image based on the number of edge appearances output from the edge appearance number counting part 24. The frequent edge region extraction unit 25 outputs the extracted frequent edge region to the high frequency edge image generation unit 26.

  The high frequency edge image generation unit 26 generates a high frequency edge image based on the frequent edge region output from the frequent edge region extraction unit 25. The high-frequency edge image generation unit 26 outputs the generated high-frequency edge image to the approximate edge image selection unit 27.

  Based on the high-frequency edge image output from the high-frequency edge image generation unit 26, the approximate edge image selection unit 27 most closely approximates the high-frequency edge image from the edge images stored in the edge image storage unit 23. Select an approximate edge image. The approximate edge image selection unit 27 outputs the selected approximate edge image to the representative image selection unit 28.

  Based on the approximate edge image output from the approximate edge image selection unit 27, the representative image selection unit 28 selects a background image corresponding to the approximate edge image from the background images stored in the image database 21 as the representative image. To do.

  Next, a processing procedure in the image processing apparatus according to the present embodiment will be described. In the image processing apparatus 2 according to the present embodiment, similar to the image processing apparatus 1 according to the first embodiment, a plurality of images captured and acquired including a common spot under a plurality of different environmental conditions. An operation for selecting an image suitable for later image processing is performed.

  FIG. 4 is a flowchart showing a processing procedure in the image processing apparatus according to the present embodiment. As shown in FIG. 4, in the image processing apparatus 2, first, a background image transmitted from the imaging apparatus 10 is acquired (S11) and stored in the image database 21 (S12). Next, the edge image conversion unit 22 reads a plurality of background images stored in the image database 21 (S13), and converts each background image into an edge image (S14).

  When converting a background image into an edge image, a known edge image generation means can be used. Specifically, the edge strength is calculated for each pixel using a sobel filter. Here, a threshold value for determining the presence / absence of an edge is set in advance, and for a pixel whose edge strength exceeds this threshold value, the pixel value is 1, and the edge strength is equal to or less than the threshold value. For, an edge image with a pixel value of 0 is generated.

  When the edge image is generated, the edge appearance number counting unit 24 counts the number of edge appearances (S15). In counting the number of edge appearances, it is determined whether the pixel value of a pixel corresponding to the same location in a plurality of edge images is 1 or 0. If the pixel value is 1, an edge appears as if an edge appeared. Count as a count. Thus, in a plurality of edge images, the number of edge appearances at pixels located at the same location is counted.

  When the edge appearance count counting unit 24 counts the number of edge appearances of the pixel at the same location, the edge frequent appearance region extraction unit 25 extracts the edge frequent appearance region (S16). The frequent edge region extraction unit 25 stores a threshold value for determining the number of edge appearances in each pixel as a frequent edge region. The edge frequent appearance area extraction unit 25 compares the number of edge appearances of each pixel output from the edge appearance number counting unit 24 with the stored threshold value, and determines whether the edge appearance frequency exceeds the threshold value. It detects as a pixel. The frequent edge region extraction unit 25 extracts a set of detected edge pixels as a frequent edge region.

  Subsequently, the high-frequency edge image generation unit 26 generates a high-frequency edge image (S17). The high-frequency edge image generation unit 26 generates a high-frequency edge image based on a set of edge pixels output from the frequent edge region extraction unit 25.

  When the high-frequency edge image is generated, the approximate edge image selection unit 27 selects an approximate edge image that most closely approximates the high-frequency edge image from the plurality of edge images stored in the edge image storage unit 23 (S18). . When selecting an approximate edge image, each of the plurality of edge images stored in the edge image storage unit 23 is compared with the high-frequency edge image output from the high-frequency edge image generation unit 26. In this comparison, the number of pixels corresponding to non-edge pixels other than edge pixels in the high-frequency edge image is counted. Then, the edge image having the smallest number of edge pixels corresponding to the non-edge pixels in the high-frequency edge image is selected as the approximate edge image having the highest degree of coincidence with the high-frequency edge image.

  When the approximate edge image is selected, the representative image selection unit 28 selects a representative image (S19). When selecting the representative image, the background image used when generating the approximate edge image among the plurality of background images stored in the image database 21 is selected as the representative image. Thus, the processing in the image processing apparatus 2 is finished.

  The processing procedure of the image processing apparatus 2 according to the present embodiment will be briefly described with reference to FIG. 5 by taking as an example the case where there are two background images. As shown in FIG. Edge processing is performed on each of the first background image G1 and the second background image G2 to generate a first edge image E1 and a second edge image E2 shown in FIG. A high-frequency edge image EM shown in FIG. 5C is generated from the edge images E1 and E2. The high-frequency edge image EM and the edge images E1 and E2 are respectively compared to select an approximate edge image. In this example, the number of pixels corresponding to non-edge pixels in the high-frequency edge image is smaller in the second edge image E2 than in the first edge image E1. Therefore, the second edge image E2 is selected as the approximate edge image. Then, as shown in FIG. 5D, the second background image G2 corresponding to the second edge image E2 is selected as a representative image.

  As described above, in the image processing apparatus 2 according to the present embodiment, in addition to the effects shown in the first embodiment, an edge that is a pixel having a large number of edge appearances in a plurality of edge images generated from the background image. The background image corresponding to the edge image having the lowest edge appearance rate in the non-edge pixels other than the pixels is selected. For this reason, since an image having a high ratio of edge pixels can be selected, the influence of noise other than the additional environmental conditions can be reduced, and an image with the least disturbance is selected from a plurality of background images. be able to.

Subsequently, a third embodiment of the present invention will be described. FIG. 6 is a block diagram of an image processing apparatus according to the third embodiment. As shown in FIG. 6, the image processing apparatus 3 according to the present embodiment includes an image database 31, an edge detection region setting unit 32, an edge image conversion unit 33, an edge image storage unit 34, and an in-region edge amount acquisition unit 35. ing. Further, the image processing device 3 includes a minimum edge amount partial image selection unit 36 and a representative image selection unit 37. Further, an imaging device 10 is connected to the image processing device 3.

  The image processing device 3 stores a plurality of background images transmitted from the imaging device 10 in the image database 31. The edge detection area setting unit 32 sets a predetermined partial area in each of a plurality of background images stored in the image database 31. The partial region set here can be determined as appropriate, and can be around a specific position, or can be the entire background image.

  The edge image conversion unit 33 performs edge processing on the partial region set by the edge detection region setting unit 32 in the background image stored in the image database 21 to generate a partial edge image. The edge image conversion unit 33 stores the generated partial edge image in the edge image storage unit 34. The edge image storage unit 34 stores a plurality of partial edge images generated by the edge image conversion unit 33.

  The in-region edge amount acquisition unit 35 determines whether a pixel in the partial edge image generated from the partial region in the background image and stored in the edge image storage unit 34 is an edge pixel or a non-edge pixel. Get the amount of edge pixels. The in-region edge amount acquisition unit 35 outputs the edge amount in each acquired partial edge image to the minimum edge amount partial image selection unit 36.

  The minimum edge amount partial image selection unit 36 selects a minimum edge amount image based on the edge amount in the partial edge image output from the in-region edge amount acquisition unit 35. Here, the partial edge image having the smallest edge amount in the partial edge image is selected as the minimum edge amount image. The minimum edge amount partial image selection unit 36 outputs the selected minimum edge amount image to the representative image selection unit 37.

  Based on the minimum edge amount image output from the minimum edge amount partial image selection unit 36, the representative image selection unit 37 outputs from the minimum edge amount partial image selection unit 36 from the background images stored in the image database 31. A background image including a partial region corresponding to the minimum edge amount image is selected as a representative image.

  Next, a processing procedure in the image processing apparatus according to the present embodiment will be described. In the image processing apparatus 3 according to the present embodiment, similar to the image processing apparatus 1 according to the first embodiment, a plurality of images captured and acquired including a common point under a plurality of different environmental conditions. An operation for selecting an image suitable for later image processing is performed.

  FIG. 7 is a flowchart showing a processing procedure in the image processing apparatus according to the present embodiment. As shown in FIG. 7, in the image processing device 3, first, a background image transmitted from the imaging device 10 is acquired (S21) and stored in the image database 31 (S22). Next, the edge detection area setting unit 32 reads a plurality of background images stored in the image database 31 (S23), and sets a partial area for edge detection (S24). The partial detection area here may be a preset area or may be set by any other appropriate method. In this embodiment, the edge detection area is a partial detection made up of a part of the background image, but the entire background image can also be used as the edge detection area.

  Once the partial area is set, edge processing is performed on the partial area in each background image to generate a partial edge image (S25). When converting a partial region in the background image into an edge image, a known edge image generation means can be used. Specifically, the edge strength is calculated for each pixel using a sobel filter. Here, a threshold value for determining the presence / absence of an edge is set in advance, and for a pixel whose edge strength exceeds this threshold value, the pixel value is 1, and the edge strength is equal to or less than the threshold value. For, a partial edge image with a pixel value of 0 is generated.

  When the partial edge image is generated, the generated partial edge image is stored in the edge image storage unit 34 (S26). Since the partial edge image is generated for each background image, a plurality of partial edge images are stored in the edge image storage unit 34. Subsequently, the edge amount in the partial edge image is acquired by the in-region edge amount acquisition unit 35 (S27). Here, the pixel value in each partial edge image is referred to, and the number of pixels having a pixel value of 1 is acquired.

  Then, the minimum edge amount partial image selection unit 36 selects the minimum edge amount partial image (S28). Here, the number of edge pixels in each partial edge image is compared, and the partial edge image having the smallest number of edge pixels is selected as the minimum edge amount partial image.

  Thereafter, the representative image selection unit 37 selects a background image including a partial region corresponding to the minimum edge amount partial image from the background images stored in the image database 31 (S29). Thus, the processing in the image processing device 3 is finished.

  The processing procedure of the image processing apparatus 3 according to the present embodiment will be briefly described with reference to FIG. 8 by taking as an example the case where there are two background images. As shown in FIG. The first partial area image B1 and the second partial area image B2 shown in FIG. 8B are set for the first background image G1 and the second background image G2. By performing edge processing on the first partial region image B1 and the second partial region image B2, a first partial edge image E1 and a second partial edge image E2 shown in FIG. 8C are generated. Then, the number of edge pixels in each of the first partial edge image E1 and the second partial edge image E2 is determined, and the minimum edge amount partial image is selected. In this example, since the number of edge pixels is smaller in the second partial edge image E2 than in the first partial edge image E1, as shown in FIG. 8D, the second partial region image B2 has a minimum edge amount portion. It becomes an image. Thereafter, the second background image G2 including the second partial region image B2 is selected as the representative image from the background images stored in the image database 31.

  As described above, in the image processing device 3 according to the present embodiment, in addition to the operational effects shown in the first embodiment, a plurality of partial edge images obtained by performing edge processing on the partial region set in the background image, A background image including a partial area corresponding to a partial edge image with the smallest edge amount, which is a partial edge image having the smallest number of edge pixels, is selected as a representative image. For this reason, since the place which has not received the fluctuation | variation of additional environmental conditions, such as illumination, can be detected accurately, the image with few disturbances can be selected from several background images.

  The preferred embodiment of the present invention has been described above, but the present invention is not limited to the above embodiment. For example, in each of the above-described embodiments, the imaging device 10 is fixed and the same point is imaged. However, for example, it is possible to target an aspect in which a predetermined position where the vehicle passes everyday is captured at different dates and times. Therefore, as a “common point”, in addition to a point that is completely the same, a point that is close and includes at least an overlapping point that is an imaging range.

  In the second and third embodiments, the edge image or partial edge image obtained by performing edge processing on the background image or a part of the background image is generated, but the texture in the image is obtained, It is also possible to select an image with little change in the additional environmental conditions using the texture density. In this case, when an edge in the image appears or does not appear, the image can be selected stably.

  Furthermore, in the first embodiment, the background image having the smallest variance among all the pixels in the set of pixels whose luminance is separated from the average luminance by a predetermined value or more is used as the representative image. It is possible to adopt another aspect in which the representative image is determined based on the dispersion among all the pixels in the set of pixels that are separated from the luminance by a predetermined value or more. For example, an image in which the variance among all the pixels in a set of pixels whose luminance is separated from the average luminance by a predetermined value or more is a predetermined threshold value (reference variance value) or less can be used as the representative image.

  In the second embodiment, the background image corresponding to the approximate edge image that most closely approximates the high-frequency edge image is used as the representative image. In addition to this aspect, the edge image is based on the degree of approximation with the high-frequency edge image. Another mode in which an image is determined and a background image corresponding to the edge image is determined as a representative image may be employed. For example, a background image corresponding to an edge image whose degree of approximation with a high-frequency edge image is a predetermined threshold value (reference degree of approximation) or more can be used as a representative image.

  Further, in the third embodiment, the background image including the partial region corresponding to the minimum edge amount partial image is selected as the representative image. In addition to this aspect, the background image is based on the degree of occupation of the edge amount in the partial edge image. It is also possible to select another edge amount partial image and use a background image including a partial region corresponding to the edge amount partial image as a representative image. For example, a background image including a partial region corresponding to an edge amount partial image in which the degree of occupation of the edge amount in the partial edge image is greater than or equal to a predetermined threshold (reference occupation degree) can be used as the representative image.

1 is a block configuration diagram of an image processing apparatus according to a first embodiment. 3 is a flowchart illustrating a processing procedure in the image processing apparatus according to the first embodiment. It is a block block diagram of the image processing apparatus which concerns on 2nd Embodiment. 10 is a flowchart illustrating a processing procedure in the image processing apparatus according to the second embodiment. It is explanatory drawing which illustrates roughly the procedure of the process in the image processing apparatus which concerns on 2nd Embodiment using the example of an image. It is a block block diagram of the image processing apparatus which concerns on 3rd Embodiment. 10 is a flowchart illustrating a processing procedure in an image processing apparatus according to a third embodiment. It is explanatory drawing which illustrates roughly the procedure of the process in the image processing apparatus which concerns on 3rd Embodiment using the example of an image.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1-3 ... Image processing apparatus, 10 ... Imaging apparatus, 11 ... Image database, 12 ... Statistical processing part, 13 ... Representative image selection part, 21 ... Image database, 22 ... Edge image conversion part, 23 ... Edge image storage part, 24 ... Edge appearance count section, 25 ... Edge frequent area extraction section, 26 ... High-frequency edge image generation section, 27 ... Approximate edge image selection section, 28 ... Representative image selection section, 31 ... Image database, 32 ... Edge detection area Setting unit, 33... Edge image conversion unit, 34... Edge image storage unit, 35... Within-region edge amount acquisition unit, 36... Minimum edge amount partial image selection unit, 37.

Claims (10)

Image input means for inputting a plurality of images captured and acquired including a common point under a plurality of different environmental conditions;
Image storage means for storing the plurality of images;
Environmental condition addition region extracting means for extracting an environmental condition addition region added by the environmental condition in each of the plurality of images;
Based on the area of the environmental condition addition region, an additional environmental condition criterion selection image selection unit that selects an additional environmental condition criterion selection image from a plurality of images stored in the image storage unit;
An image processing apparatus comprising:   The additional environmental condition criterion selection image selection means selects an environmental load condition small amount image from a plurality of images stored in the image storage means, wherein the environmental load condition small area is smaller than a predetermined area. The image processing apparatus according to claim 1, which is an image selection unit.   The image processing apparatus according to claim 1, wherein the environmental condition addition region extraction unit extracts the environmental condition addition region by performing statistical processing on the plurality of images.   The additional environmental condition criterion selection image selection unit selects an image having the smallest area of the environmental condition addition region as a minimum environmental load condition image from a plurality of images stored in the image storage unit. The image processing apparatus according to claim 1, wherein the image processing apparatus is a minimum image selection unit. The environmental condition addition region extraction means includes:
Edge image conversion means for converting each of the plurality of images into an edge image;
Edge image storage means for storing a plurality of the edge images converted by the edge image conversion means;
Edge appearance count measuring means for measuring the number of edge appearances in a predetermined region in the plurality of edge images;
Of the predetermined areas, edge frequent appearance area extracting means for extracting an edge frequent area that is a predetermined area whose number of edge appearances is equal to or greater than a predetermined threshold;
An edge frequent area storage means for storing the edge frequent area;
The additional environmental condition criteria selection image selection means,
A high-frequency edge image generating unit that generates a high-frequency edge image based on each edge frequent region in the plurality of edge images;
Based on the approximate degree of the high frequency edge image, from among a plurality of edge images stored in the edge image storage means, and a approximate degree reference edge image extraction unit that extracts an approximate degree reference edge image,
Claim wherein the additional environmental conditions minimal image selection means, as said additional environmental condition criteria select the image to select an image corresponding to the approximate degree reference edge image from a plurality of edge images stored in the edge image storage means 5. The image processing apparatus according to 4 .   6. The approximation degree reference edge image extraction unit extracts a high approximation edge image that is an edge image whose approximation degree with the high-frequency edge image is a predetermined threshold value or more as an approximation degree reference selection edge image. An image processing apparatus according to 1.   The image processing apparatus according to claim 5, wherein the approximate degree reference edge image extraction unit extracts an approximate edge image that is an edge image that is closest to the high-frequency edge image as an approximate degree reference selection edge image. The environmental condition addition region extraction means includes:
Edge image conversion means for converting each of the plurality of images into an edge image;
An edge amount calculating means for calculating an edge amount for each predetermined region in each edge image,
The additional environmental condition criteria selection image selection means,
Edge amount occupation degree reference selection edge image selection means for selecting an edge amount occupation degree reference selection edge image based on the edge amount occupation degree of the predetermined area in the edge image;
Edge amount occupation degree reference selection image selection means for selecting an image corresponding to the edge amount occupation degree reference selection edge image as the additional environmental condition reference selection image from a plurality of images stored in the image storage means;
An image processing apparatus according to claim 1.   The edge amount occupancy degree reference selection edge image selects, as the edge amount occupancy degree reference selection edge image, an edge amount small edge image in which the edge amount of the predetermined region in the edge image is equal to or less than a predetermined occupancy degree. Item 9. The image processing apparatus according to Item 8.   The edge amount occupation degree reference selection edge image selection means selects an edge amount minimum edge image that minimizes the edge amount of the predetermined region in the edge image as the edge amount occupation degree reference selection edge image. An image processing apparatus according to 1.