JP2009151396A - Image evaluation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image evaluation device for performing exposure decision without noticing circumstances where an image has been photographed, and for satisfactorily detecting an exposure defective image corresponding to various exposure defective circumstances. <P>SOLUTION: This image evaluation system 12 is configured of an image information storage part 13 for storing an evaluation object image and the attribute information of the evaluation object image by associating those images; and an image evaluation part 14 for selecting an evaluation object image 100 from the image information storage part 13, and for performing exposure decision to the evaluation object image 100. This image evaluation part 14 extracts at least two or more regions from the evaluation object image 100, and generates an edge image for every region, and selects the region whose edge components to be calculated from the edge region are the highest as a decision region, and performs exposure decision based on the edge components of the decision region. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image evaluation apparatus.

After mounting a photometric sensor on an imaging device such as an image centered on a digital camera etc., the exposure state is determined based on the sensor information measured by the photometric sensor, and the illuminance information of the image to be captured is sensed. Various imaging devices have been proposed in which imaging parameters are adjusted so that a captured image can obtain an appropriate exposure and an image is captured. In addition to the exposure determination method using these photometric sensors, image information such as Exif information is read in post-processing on an already captured image, and exposure determination is performed using only this image information. Has also been proposed (see, for example, Patent Documents 1 to 3).
JP 2000-134467 A JP 2006-135587 A JP 2002-247361 A

  However, among such conventional methods, the exposure determination method based on the sensor information of the photometric sensor is useful for all images when considering exposure determination in post-processing for images after shooting. Sensor information (photometric information) is not always recorded, and its adaptive range is limited to cases where exposure determination processing is performed within a camera equipped with a photometric sensor. On the other hand, among the exposure determination methods using only image information, the method described in Patent Document 1 simply determines whether the exposure is good or bad based on the shape of the luminance histogram of the image. As a result, there was a problem that many misjudgments were made.

  In order to reduce such misjudgments, in the technique described in Patent Document 2, an area for exposure failure detection (hereinafter referred to as an exposure determination area) is manually set for an exposure determiner (hereinafter referred to as a user). After taking the selection method, exposure determination is performed based on the shape of the exposure luminance histogram in the exposure determination area. However, with this method, there is a problem that it is inefficient to perform these processes on a large number of images because it takes time and effort for the user to specify the exposure determination area. In the technique described in Patent Document 3, the exposure of an image is determined after determining whether the captured image is backlit or nighttime. Since this is a method that works under backlighting or nighttime conditions, when shooting in clear sky, cloudy weather, sunset scenes, etc., when the overall exposure is under or over, the exposure determination installed in the camera There is a problem that the system may make a false determination, and the exposure determination system does not function effectively.

  The present invention has been made in view of such problems, and when determining whether exposure is good or bad by using only image information, it is possible to perform exposure determination without being aware of the situation in which the image was captured. An object of the present invention is to provide an image evaluation apparatus that can detect a poorly exposed image corresponding to various poorly exposed situations.

  In order to solve the above problems, an image evaluation apparatus according to the present invention includes an image information storage unit that stores an evaluation target image and attribute information of the evaluation target image in association with each other, and an evaluation target image from the image information storage unit. And an image evaluation unit that selects and determines whether the evaluation target image is good or poorly exposed. The image evaluation unit divides the evaluation target image into a plurality of regions, extracts at least two or more regions from the divided regions, generates an edge image for each region, and generates the edge A region having the highest edge component calculated from the image is selected as a determination region, and exposure determination is performed based on the edge component of the determination region.

  In such an image evaluation apparatus according to the present invention, it is preferable that the image evaluation unit is configured to extract at least two or more regions including the vicinity of the center of the evaluation target image among the plurality of divided regions. .

  In the image evaluation device according to the present invention, the image evaluation unit performs face detection processing on the evaluation target image, and includes a region where the face is detected by the face detection processing among the plurality of divided regions. It is preferable that at least two or more regions are extracted.

  Furthermore, in the image evaluation apparatus according to the present invention, the image evaluation unit determines that the evaluation target image is a good exposure image when the value of the edge component of the determination region selected from the evaluation target image is equal to or greater than a threshold value. When it is less than the threshold value, the evaluation target image is preferably determined to be a poorly exposed image.

  In this case, the image evaluation unit further divides the determination area into small areas when the evaluation target image is determined to be a poorly exposed image, and calculates an edge component calculated from the edge image obtained for each small area. On the basis of the exposure determination for each small area, if the number of small areas determined as the poorly exposed area exceeds the threshold, the evaluation target image is determined as the poorly exposed image, and the small area determined as the poorly exposed area is determined. When the number of regions is equal to or less than the threshold value, it is preferable that the evaluation target image is determined to be a good exposure image.

Furthermore, in this case, when the exposure evaluation for each small region is performed, the image evaluation unit sets the intensity of the red component to R (m, n) in each pixel (m, n) of the small region, and the green component. Is G (m, n), the intensity of the blue component is B (m, n), the largest component value V _max (m, n) and the smallest component value V _min (m, n) in each pixel. n) and its difference S (m, n)
However, m and n are defined by representing the x and y coordinate values of the pixels in the small region, and are based on the V _max (m, n), V _min (m, n), and S (m, n). It is preferable to be configured to perform exposure determination.

Further, when the image evaluation unit performs the exposure determination for each small region, the saturation S_red (m, n) centered on the red component and green in each pixel (m, n) of the small region. Saturation S_green (m, n) centered on the component and S_blue (m, n) saturation on the blue component
Here, m and n are defined by representing the x and y coordinate values of the pixels in the small region, and V _max (m, n), V _min (m, n), and S (m, n), and The exposure determination is preferably performed based on the S_red, S_green, and S_blue.

  In this case, the small region is preferably obtained by dividing the determination region into a plurality of substantially equal parts.

  Furthermore, in this case, the image evaluation unit performs face detection processing for each small region, and based on the result of exposure determination for the small region where the face is detected by the face detection processing, the total number of exposure good regions or exposure When adding the total number of defective areas, it is preferable to add a number obtained by multiplying a set weighting factor for a small area where a face is detected.

  The image evaluation apparatus of the present invention preferably includes a determination result setting unit that sets the result of exposure determination by the image evaluation unit as an exposure determination value in the attribute information of the evaluation target image and stores it in the image information storage unit. .

  When the image evaluation apparatus according to the present invention is configured as described above, when determining whether exposure is good or bad by using only image information, it is possible to perform exposure determination without being aware of the situation in which the image is captured, It is possible to satisfactorily detect a poorly exposed image corresponding to various poorly exposed situations.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. First, the configuration of the image evaluation apparatus (image evaluation system 12) according to the present embodiment will be described with reference to FIGS. In the present embodiment, a case where the image evaluation system 12 is applied to a digital camera 10 having an imaging unit 11 will be described. As shown in FIG. 1, the digital camera 10 is connected to a display 20 that is a display / playback device via a video output line 30. And the digital image 100 image | photographed with the imaging part 11 is displayed on the screen G1 of the display 20. FIG. Moreover, you may comprise on this screen G1 so that the exposure determination result of the digital image 100 obtained with the image evaluation system 12 concerning a present Example may be displayed.

  As illustrated in FIG. 2, the digital camera 10 includes an imaging unit 11, an image evaluation system 12 that performs exposure determination of a digital image captured by the imaging unit 11, and a video output unit 16 that outputs the digital image to a display 20. It consists of. The image evaluation system 12 also includes an image information storage unit 13 in which a digital image (evaluation target image 100) captured by the imaging unit 11 and its attribute information are stored in association with each other, and an evaluation from the image information storage unit 13. From the image evaluation unit 14 that selects the target image 100 and performs exposure determination on the evaluation target image 100 based on the image information, and the determination result setting unit 15 that stores the determination result in the image information storage unit 13 Composed. In the digital camera 10, the image pickup unit 11 includes an optical system and an image pickup device, and detects an image formed on the image pickup surface of the image pickup device by the optical system to obtain a digital image. And attribute information such as the shooting date and time are stored and recorded in the image information storage unit 13.

  The digital image stored in the image storage unit 13 has a data structure 50 shown in FIG. 3 and includes a header area 51, an attribute information area 52, and an image information area 53. Here, the image information area 53 is a portion in which the digital image itself captured by the imaging unit 11 is stored. On the other hand, the attribute information area 52 is a portion in which attribute information of a digital image is stored, and in this embodiment, an exposure determination result by the image evaluation system 12 is also stored. The attribute information area 52 includes an exposure determination value area 52a in which the result of exposure determination is stored, a shooting date / time area 52b in which shooting date / time is stored, an AF area information area 52c in which AF area information is stored. As such a data structure, Exif (Exchangeable image file format) standardized by JEIDA is known. In the following description, the exposure determination result is expressed as a character string (exposure good: “good”, exposure failure: “bad”) and stored in the exposure determination value area 52a as an example. However, the present invention is not limited to this embodiment. For example, the character string of the determination result may be converted into a numerical value and stored in the exposure determination value area 52a.

(Image exposure judgment and judgment result setting)
Then, an explanatory diagram of an example of use of the digital camera shown in FIG. 1, a block diagram shown in FIG. 2, an explanatory diagram of a data structure shown in FIG. 3, a flowchart shown in FIGS. 4 and 5, and an explanation shown in FIGS. An image exposure determination process and a determination result setting process performed by the image evaluation unit 14 will be described with reference to the drawings. These processes are started, for example, at the timing when the imaging unit 11 captures an image and the digital image is registered in the image information storage unit 13, and the process is executed on the registered digital image. In the following description, this digital image registered by the imaging unit 11 is referred to as “evaluation target image 100”.

  When the exposure determination process is activated, as shown in the flowchart of FIG. 4, the image evaluation unit 14 first reads the evaluation target image 100 from the image information storage unit 13 (step S10). Then, an exposure determination process is performed on the read evaluation target image 100 (step S20). Next, the result of exposure determination by the image evaluation unit 14 is determined by the determination result setting unit 15 (in this embodiment, “good” when exposure is good with respect to the variable “exposure” as described later, In the case of poor exposure, “bad” is set), and is stored in the exposure determination value area 52a of the image information storage unit 13 shown in FIG. 3 (step S30). The exposure determination processing and determination result setting processing (steps S20 and S30) are performed on the registered image when activated when the captured image is registered in the image information storage unit 13. In the case where the process is terminated and activated for an image registered in the image information storage unit 13 by some operation, the process is performed by performing the process on all evaluation target images 100 that have not been evaluated. finish.

(Exposure judgment processing)
Next, the exposure determination process (step S20) will be described in detail with reference to the flowchart of FIG. In this process, exposure determination is performed based on the edge component of the evaluation target image. First, the image evaluation unit 14 creates an edge image set I _edge from the evaluation target image 100 (hereinafter referred to as the evaluation target image I) read in step S10 (step S101). For this purpose, as shown in FIG. 6A, the sobel filters Sh and SV defined by the following equation (1) are applied to the given evaluation target image I, and FIG. The edge image set Iedge shown in FIG.

  In the above formula (1), * represents convolution calculation, and m and n represent x and y coordinate values of each pixel.

Next, a determination area I _{edge that} is an exposure determination target is extracted (step S102). For this _{purpose, after obtaining the} edge image set I _edge as described above, as shown in FIG. 6B, the edge image set I _edge is divided into three equal parts in length and breadth, and a total of nine small area sets {I _sub — n | n = 1, 2,. . . 9}. Thereafter, two regions {I _{sub_2} , I _{sub — 5} , I _{sub — 8} } and {I _{sub — 4} , I _{sub — 5} , I _{sub — 6} } are extracted as regions including the region near the center of the evaluation target image I, and the respective edge images are extracted as Icand_1. , Icand_2 (that is, an edge image is generated for each region). In Icand_1 and Icand_2, the number of edge components (that is, non-black pixels in FIG. 6B) is summed, and the larger of the total values of pixels in Icand_1 and Icand_2 is the edge image set Iedge. The determination region is I _cand (schematic diagram in FIG. 6C). The processing for extracting this I _cand is expressed by the following formula (2). Note that, in FIG. 6B, in order to clarify the edge image, black and white are reversed and displayed, unlike an image used in actual processing.

  However, in the above formula (2), the subscripts n and k represent indexes for specifying each region, sum () is a function for calculating the sum of white pixels of a given binary image, and max () Is a function that returns the maximum value of Icand_k, and arg is a function that returns the index of the image area that is the maximum value of max (Icand_k). Region (I, Icand_n) is a function for cutting out an area corresponding to Icand_n from the evaluation target image I.

  Next, a comparison is made between the sum sum (Icand) of the white pixels of the clipped image area Icand and a predetermined threshold value Tcand (step S103). If sum (Icand) <Tcand (the edge component is less than the threshold value), the process proceeds to the exposure failure determination process (steps S104 to S112) described below. On the other hand, if sum (Icand) ≧ Tcand (the edge component is equal to or greater than the threshold value), the process proceeds to step S112, where the valence target image I is determined to be good exposure, and “good” is set in the parameter “exposure”. The exposure determination process ends.

  In this embodiment, after generating the edge image set Iedge of the evaluation target image I, two regions {Isub_2, Isub_5, Isub_8} and {Isub_4, Isub_5, Isub_6} are extracted from Iedge, and the respective edge images Icand_1 are extracted. , Icand_2 has been obtained. On the other hand, as another different embodiment, after extracting two regions {Isub_2, Isub_5, Isub_8}, {Isub_4, Isub_5, Isub_6} from the evaluation target image I, edge images Icand_1 and Icand_2 only for the respective regions. May be generated.

(Exposure defect judgment processing)
Hereinafter, the exposure failure determination process (steps S104 to S112) will be described. In this process, exposure determination is performed in more detail based on the saturation and lightness of the determination area I _cand . Here, the case where Icand_1 = {Isub_2, Isub_5, Isub_8} is selected as the Icand will be described as an example. First, for each small region {Isub_k, | k = 2, 5, 8} constituting Icand, {Vmax_k (m, n)} defined in the following equation (3) for each pixel (m, n), {Vmin_k (m, n)}, {S _k (m, n)}, and {S_red _k (m, n)}, {S_greenk (m, n)}, {S_bluek} defined in the following equation (4) (M, n)} is obtained (step S104).

In the above equation (3), m and n represent the x and y coordinate values of the pixels in the small area, and k represents the small area index (k = 2, 5, 8). R _{_k} (m, n) represents the intensity of the red component of each pixel of Icand_k (m, n), G _k (m, n) represents the intensity of the green component, B _{_k} (m, n) is blue Represents the strength of the component. The largest component value {V _{max — k} (m, n)}, the smallest component value {V _{min — k} (m, n)} and the difference {S _k (m, n)} are obtained for each pixel.

In the above equation (4), m and n represent the x and y coordinate values of the pixels in the small area, and k represents the small area index (k = 2, 5, 8). Further, Widthk and height _k represent the width and height of each small area respectively. S_red _k (m, n) is the saturation centered on the red component, S_reen _k (m, n) is the saturation centered on the green component, and S_blue _k (m, n) is the saturation centered on the blue component. Represents. In this way, by defining the saturation corresponding to each RGB color and using these saturations for exposure determination, it is possible to perform finer exposure determination. _{_}

Thereafter, S _k obtained in step S104 (m, n), V max_k (m, n), V min_k (m, n), S_red k (m, n), S_green k (m, n) and S_blue _k Using (m, n), the exposure failure determination values expv _{und_k} and expv _{ov_k} are obtained (step S105). To do this, first, preset threshold Ts_und, Ts_ov, with using Tmax, Tmin, Tred, the T _green and T _blue, for each small region satisfies the determination expression defined by the following formula (5) The number of pixels C _{und_k} and C _{ov_k} is obtained. Next, expv _{und_k} and expv _{ov_k} defined in the following formula (5) are obtained based on these numbers.

In the above equation (5), m and n represent the x and y coordinate values of the pixels in the small area, k represents an index (k = 2, 5, 8) for designating each small area, and count ( ) Represents a function that means counting up the number of pixels that satisfy the determination formula in the function. Also, widthk and height _k represent the width and height of each small region _{Isub_k} , respectively.

After expv _{und_k} and expv _{ov_k} are obtained in step S105, processing for determining good exposure and poor exposure for the evaluation target image I is performed using preset threshold _values T _und and T _ov (steps described below) S106 to S112). First, in step _S106, expv und_k and _expv ov_k for a small area Isub_k it it is determined whether or not satisfy the determination expression defined by the following formula (6). When the determination formula is satisfied, that is, when at least one of the underexposed and overexposed small regions is present, the small region Isub_k is determined as the “exposure poor region” (step S107). When the determination formula is not satisfied, that is, when neither the under-exposed nor the over-exposed small region exists, the small region Isub_k is determined as a “good exposure region” (step S108).

It is determined whether all the small areas I _{sub_k} have been determined to be good or defective in exposure (step S109). If there is still a small area to be determined, the processes in steps S106 to S108 are repeated.

When the above exposure determination is performed for all the small regions I _{sub_k} , the set of I _{sub_k} determined as the “exposure poor region” in the above step S107 corresponds to which of the determination regions I _cand It is determined whether it occupies a certain area (step S110). For this, in the determination area I _cand , when it is determined that exposure is poor in an area exceeding a certain range (exceeding a threshold), the evaluation target image I is determined as “exposure image” and “exposure” is set to “ "bad" is set (step S111). In this embodiment, the condition for determining that the evaluation target image I is an “exposure poor image” is that, in I _cand , a small area exceeding a majority is determined as an “exposure poor area”. Therefore, if it is determined that “exposed poor region” in two or more small regions among {I _{sub —} k} (k = 2, 5, 8), the evaluation target image I is set as “unexposed image”.

  In cases other than the above, that is, when the small area determined to be poorly exposed is below a certain range (in this embodiment, the small area determined to be “exposed poor area” is 1 or less) (below the threshold value) The image I is determined to be “good exposure image”, and “good” is set to the parameter exposure (step S112).

  In this way, when performing exposure determination of the evaluation target image, by using the edge component, saturation, brightness, etc., by appropriately setting the threshold value of each conditional expression that defines these, underexposure, overexposure, etc. Exposure determination corresponding to various exposure situations can be performed, and an image with poor exposure can be detected well. In the prior art, it is necessary to use information at the time of photographing such as photometric data. In this embodiment, the image stored in the image information storage unit 13 is not conscious of the situation where the image is photographed. Exposure determination can be performed using only information.

In the above embodiment, the two regions Icand_1 and Icand_2 in the height direction and the width direction including the region near the center of the evaluation target image I are extracted, and the determination region I _cand is selected from these. As a different embodiment, for example, three or more regions including the central region such as the height direction, the width direction, and the diagonal direction may be extracted, and the determination region I _cand may be selected from these. Further, a process using a conventionally known face detection function is performed on the evaluation target image I, and two or more areas are extracted around the area where the face is detected by such face detection process. The determination area I _cand may be selected from the _list . A method using the face detection function will be described in detail below. Further, in the above embodiment, the case where the edge image set is divided into 9 parts as a whole by dividing the edge image set into 3 parts has been described. However, the present invention is not limited to this example, and the number of divisions may be larger. However, the number of divisions may be less than that.

[Judgment area selection procedure using face detection processing]
Here, in the embodiment using the face detection function, the procedure for extracting the determination area I _{cand in} step S102 will be described with reference to FIG. The face detection function here is a function for detecting a human face from the evaluation target image, and the output information includes the number of faces present in the image, the position of the face in the image, and the size of the face. Has already been put to practical use in the fields of digital cameras and security. Also in this embodiment, it is assumed that a face detection function capable of obtaining similar output information is used.

First, as a premise for selecting the determination area I _cand for exposure determination, it is assumed that the number of faces and the position and size of each face are obtained as information by the face detection function as described above as preprocessing. . Next, as in the above-described embodiment, after the evaluation target image I is divided into small areas I _{sub_k} , three small areas are selected in descending order of the proportion of the face in each small area. _{Let cand} = { _{Isub_k} }. In the explanatory diagram of FIG. 7, I _cand = {Isub_5, Isub_6, Isub_9} is selected. Thereafter, an edge component of I _cand is _obtained for each extracted small area of I _cand based on the edge image set Iedge created in step S101. Based on the edge component, exposure determination is performed in the same manner as the method used in the above-described embodiment, and exposure evaluation of the evaluation target image I is performed in accordance with the processing from step S103. If only three small areas cannot be extracted from the result of face detection only (that is, if there are only one or two small areas in which a face exists), the area occupied by the face is the largest. In the adjacent small area, the above-described edge component is obtained, and a small area having a large number of edge components is selected and included in the determination area I _cand .

[Small area exposure using face detection processing]
As _still another embodiment, in the exposure failure determination process (steps S104 to _S112 ), the face detection process using the face detection function as described above is _performed on each small area I _{sub_k} of the determination area I _cand selected above. And weighting may be performed so that exposure determination is focused on a small region in which a face is detected by the face detection process. And according to this weighting, the image evaluation part 14 is comprised so that the exposure determination of the evaluation object image I may be performed. For example, a weighting factor is set for a small region in which a face is detected (for example, weighting factor = 2), and the number of small regions determined to be poorly exposed in step S107 is multiplied by this weighting factor. To do. Then, even if the number of areas determined to be poorly exposed is only 1, the multiplication result is 2, so that it is determined to be “2 or more” in step S110. It is determined as “image”. Moreover, you may weight about elements other than a face. These weightings can be configured such that a weighting factor for each element is stored in the system, and the image evaluation unit 14 performs exposure determination according to the weighting factor. Although not shown, a weighting factor setting unit may be provided so that the user can arbitrarily set and correct the value of the weighting factor so that the user's preference can be reflected in the exposure determination.

  When the image evaluation system 12 as described above is provided in the digital camera 10, a poorly exposed image corresponding to a variety of poorly exposed situations such as underexposure and overexposure with respect to the digital image stored in the image information storage unit 13. Can be detected satisfactorily. Therefore, when organizing digital images and the like, it is possible to efficiently perform sorting operations such as images with good exposure and images with poor exposure and organizing operations such as deletion, printing, display, and storage.

  In the above embodiment, the case where the image evaluation system 12 is provided in the digital camera 10 has been described. However, the present invention is not limited to this embodiment. The image evaluation system 12 is mounted on a server connected to a network when configured as an image display device, and a digital image is transmitted from a terminal (digital camera, mobile phone, PDA or personal computer) connected to the network to the server. The same effect can be obtained even if the exposure determination is performed by transferring.

It is explanatory drawing explaining the usage method of the digital camera provided with the image evaluation system which concerns on this invention. It is a block diagram which shows the structure of the said digital camera. It is explanatory drawing for demonstrating the data structure of a digital image. It is a flowchart which shows the process of the exposure determination of an image, and the setting of a determination result. It is a flowchart which shows an exposure determination process. It is explanatory drawing for _{demonstrating} the procedure which extracts the determination area _| region I _cand from the evaluation object image I. FIG. It is explanatory drawing for _{demonstrating} the procedure which extracts the determination area _| region I _cand from the evaluation object image I using a face detection process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Digital camera 11 Imaging part 12 Image evaluation system (image evaluation apparatus)
13 Image information storage unit 14 Image evaluation unit 15 Determination result setting unit 52 Attribute information (attribute information region) 52a Exposure determination region 100 Evaluation target image

Claims (10)

An image information storage unit that stores an evaluation target image and attribute information of the evaluation target image in association with each other;
An image evaluation unit that selects the evaluation target image from the image information storage unit and performs exposure determination of good exposure or poor exposure on the evaluation target image;
The image evaluation unit divides the evaluation target image into a plurality of regions, extracts at least two or more regions from the divided regions, generates an edge image for each region, and generates the edge An image evaluation apparatus configured to select the region having the highest edge component calculated from an image as a determination region and perform the exposure determination based on the edge component of the determination region.   The image evaluation device according to claim 1, wherein the image evaluation unit is configured to extract at least two or more regions including a vicinity of a center of the evaluation target image from the plurality of divided regions.   The image evaluation unit performs face detection processing on the evaluation target image, and extracts at least two or more regions including a region where a face is detected by the face detection processing from the plurality of divided regions. The image evaluation apparatus according to claim 1 configured as described above.   The image evaluation unit determines that the evaluation target image is a good exposure image when the value of the edge component of the determination region selected from the evaluation target image is equal to or greater than a threshold value. The image evaluation apparatus according to claim 1, wherein the image to be evaluated is determined to be a poorly exposed image. The image evaluation unit
When the evaluation target image is determined to be a poorly exposed image, the determination area is further divided into small areas, and each small area is calculated based on an edge component calculated from the edge image obtained for each small area. If the number of small areas determined to be poorly exposed areas exceeds a threshold value, the evaluation target image is determined to be a poorly exposed image, and the number of small areas determined to be poorly exposed areas is equal to or less than the threshold value. In this case, the image evaluation device according to claim 4, wherein the image to be evaluated is determined to be a good exposure image. The image evaluation unit
When performing exposure determination for each small area, the intensity of the red component is R (m, n) and the intensity of the green component is G (m, n) in each pixel (m, n) of the small area. , The intensity of the blue component is B (m, n), the largest component value V _max (m, n), the smallest component value V _min (m, n) and the difference S (m, n) in each pixel. )
However, m and n are defined by representing the x and y coordinate values of the pixels in the small region, and are based on the V _max (m, n), V _min (m, n), and S (m, n). The image evaluation apparatus according to claim 5, wherein the image evaluation apparatus is configured to perform exposure determination. The image evaluation unit
When performing exposure determination for each of the small regions, the saturation S_red (m, n) centered on the red component and the saturation centered on the green component in each pixel (m, n) of the small region. S_green (m, n), S_blue (m, n) is the saturation of the blue component.
However, m and n are defined by representing the x and y coordinate values of the pixels in the small region, and V _max (m, n), V _min (m, n), and S (m, n), and The image evaluation apparatus according to claim 6, configured to perform exposure determination based on the S_red, S_green, and S_blue.   The image evaluation apparatus according to claim 5, wherein the small area is obtained by dividing the determination area into a plurality of substantially equal parts. The image evaluation unit
When performing face detection processing on each of the small areas and adding the total number of good exposure areas or the total number of poor exposure areas based on the result of exposure determination for the small areas in which the face is detected by the face detection processing The image evaluation device according to claim 5, wherein the image evaluation device is configured to add a number obtained by multiplying a set weighting coefficient for the small region in which a face is detected.   The determination result setting part which sets the result of the exposure determination by the said image evaluation part as an exposure determination value to the said attribute information of the said evaluation object image, and memorize | stores it in an image information storage part. The image evaluation apparatus described.