JP7378347B2 - Evaluation device, evaluation method, evaluation program and storage medium - Google Patents

Description

The present invention relates to an evaluation device, an evaluation method, an evaluation program, and a recording medium recording the program for evaluating orthoimages.

Conventionally, images obtained by orthorectically transforming driving images have been used to assist in creating map information, etc. Patent Document 1 discloses, as a method for synthesizing two ortho-images, a technique in which a feature on a road is linearized and then the ortho-images are compared with each other to detect a positional shift.

JP2018-72637A

By the way, it is important to evaluate the composite ortho image generated by composing two or more ortho images in this way, that is, whether it can be used as an ortho image (Is the image blurred, is there a positional shift at the compositing point? However, considering the number of synthetic orthoimages generated as map information, it must be said that manual evaluation is extremely complicated. I don't get it. Furthermore, in the case of visual inspection, there is a problem in that the evaluation of the composite orthoimage varies due to individual differences.

The present invention has been made in view of the above problems, and an object of the present invention is to provide an evaluation device, an evaluation method, an evaluation program, and a storage medium that quantitatively evaluate orthoimages.

In order to solve the above problems, an evaluation device according to one aspect of the present invention provides a plurality of orthoimages obtained by orthogonally transforming each of a plurality of captured images, at a location where the captured content matches another orthoimage. An evaluation device that evaluates a composite ortho image synthesized such that the images are superimposed on each other, an acquisition unit that acquires each of the plurality of ortho images and the composite ortho image, and pixels included in the plurality of ortho images, A calculation unit that calculates a variance value of a pixel at a position to be superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed; , and an evaluation unit that generates evaluation information for evaluating the composite orthoimage.

In order to solve the above problems, an evaluation method according to one aspect of the present invention provides a method for evaluating each of a plurality of orthoimages obtained by orthogonally transforming each of a plurality of captured images, at a location where the captured content matches another orthoimage. This is an evaluation method for evaluating a composite ortho image synthesized such that the images are superimposed on each other. and a calculation step of calculating a variance value of the pixel at a position where it is superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed, and a variance value calculated by the calculation step. an evaluation step of generating evaluation information for evaluating the composite orthoimage based on the

In order to solve the above problems, an evaluation program according to one aspect of the present invention is configured to perform orthogonal transformation on each of a plurality of captured images, and to detect locations where the captured content matches other orthoimages. This is an evaluation program that evaluates a composite ortho image that has been composited so that the images are superimposed on each other. A calculation function that calculates the variance value of a pixel at a position where it is superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed, and a variance value calculated by the calculation function. and an evaluation function that generates evaluation information for evaluating a composite orthoimage based on the above.

The evaluation device may include an output unit that outputs evaluation information obtained by the evaluation unit evaluating the composite orthoimage.

In the above evaluation device, the evaluation section divides the composite orthoimage into small regions, calculates, for each small region, the average value of the variance values calculated by the calculation section of each pixel included in the small region, and calculates The largest value among the calculated average values may be used as the evaluation information of the composite orthoimage.

In the above evaluation device, the evaluation unit may remove images other than the road portion included in the composite orthoimage and evaluate the composite orthoimage of only the road portion.

An evaluation device according to one aspect of the present invention calculates the variance of pixel values of pixels at superimposed locations when superimposing orthoimages that are the basis of a composite orthoimage. By calculating the variance of the pixel values of pixels in multiple different orthoimages, which are estimated to be the same location, it can be estimated that the larger the variance value, the greater the deviation, and the synthesis It can be estimated that the evaluation as an orthoimage is low. Therefore, the evaluation device can quantitatively evaluate the composite orthoimage.

It is a block diagram showing an example of functional composition of an evaluation device. (a) is a diagram showing an example of a captured image. (b) is a diagram showing an example in which a captured image is orthogonally transformed. (a) is a diagram showing an example of a composite orthoimage. (b) is a diagram showing a mode of synthesizing orthoimages. FIG. 6 is a diagram illustrating an example of dividing a composite orthoimage into small regions. 7 is a flowchart illustrating an example of the operation of the evaluation device for evaluating orthoimages. (a) is a diagram showing an example of a non-road area. (b) is a diagram showing an example of dividing the composite orthoimage into small regions after non-road region removal. 12 is a flowchart illustrating an example of the operation of the evaluation device regarding evaluation processing of an orthoimage after non-road area removal. It is a block diagram showing another example of composition of an evaluation device.

Hereinafter, an evaluation device according to an embodiment of the present invention will be described in detail with reference to the drawings.

<Embodiment 1>
<Configuration of evaluation device>
An evaluation device according to one aspect of the present invention synthesizes each of a plurality of orthoimages obtained by orthogonally transforming each of a plurality of captured images so that locations where the captured content matches that of other orthoimages are superimposed on each other. This is an evaluation device that evaluates synthetic orthoimages. The evaluation device includes an acquisition unit (see 105 in FIG. 1) that acquires each of the plurality of orthoimages and a composite orthoimage, and a pixel included in the plurality of orthoimages at a position superimposed on another orthoimage. For a certain pixel, a calculation unit (see 105 in FIG. 1) calculates a variance value of the pixel based on the pixel value of each pixel of the ortho image to be superimposed, and a combination is performed based on the variance value calculated by the calculation unit. The evaluation unit (see 105 in FIG. 1) generates evaluation information for evaluating the orthoimage. This will be explained in detail below.

FIG. 1 is a block diagram showing an example of the functional configuration of the evaluation device 100. As shown in FIG. 1, the evaluation device 100 includes a reception section 101, a reception section 102, an output section 103, a storage section 104, and a CPU 105.

The evaluation device 100 is realized by a server device, a PC, etc., but is not limited to these, and may be realized by a mobile terminal such as a smartphone. Each functional unit will be explained in detail below.

The receiving unit 101 has a function of receiving information from other devices through communication. The receiving unit 101 may receive, for example, driving history information that is a driving history of a vehicle generated using information such as the position where the car actually traveled and the vehicle speed, so-called probe information (probe traffic information). This travel history information includes captured images taken while the vehicle was traveling. Further, the receiving unit 101 receives, for example, a captured image captured while the vehicle is running from a user's PC, the web, a flash memory, or the like. The receiving unit 101 also receives an orthogonal image obtained by orthogonally transforming a captured image including a roadway. The receiving unit 101 transmits the received captured image or orthoimage to the CPU 105.

The receiving unit 102 has a function of receiving input from the user of the evaluation device 100 and transmitting the input to the CPU 105. The reception unit 102 can be realized by, for example, a hardware key provided in the evaluation device 100 or a soft key such as a touch key. Note that the input to the reception unit 102 may be a voice input.

The output unit 103 has a function of outputting instructed data according to instructions from the CPU 105. The output unit 103 outputs, for example, evaluation information indicating an evaluation value of the composite orthoimage. The output unit 103 may output the evaluation information as image information (or text data) to a monitor connected to the evaluation device 100, or output audio data indicating the evaluation information to a speaker connected to the evaluation device 100. The evaluation information may be output to an external device, or may be transmitted to an external device via communication. That is, the output unit 103 functions as a communication interface that outputs information specified by the CPU 105 to a specified destination.

The storage unit 104 is a recording medium that stores various programs and various data including map information necessary for the operation of the evaluation device 100. The storage unit 104 is realized by, for example, an HDD (Hard Disc Drive), an SSD (Solid State Drive), a flash memory, or the like.

The storage unit 104 may store a captured image 141, an orthoimage 142, and a composite orthoimage 143. The captured image 141 is an image obtained by capturing an image including a roadway. The ortho image 142 is an ortho image obtained by orthogonally transforming the captured image 141. The composite orthoimage 143 is an image obtained by superimposing a plurality of orthoimages 142 so that their contents (positions) overlap with each other. The evaluation device 100 evaluates the composite orthoimage 143.

The CPU 105 is a processor that uses various programs and various data stored in the storage unit 104 to execute processes to be executed by the evaluation device 100.

The CPU 105 controls each part of the evaluation device 100 and evaluates a composite orthoimage obtained by orthographically transforming and combining images.

The CPU 105 functions as an acquisition unit that acquires a plurality of orthoimages and a composite orthoimage generated from these orthoimages. As the plurality of orthoimages, those received by the receiving unit 101 may be used, or those stored in the storage unit 104 may be used. Further, as the composite orthoimage, one received by the receiving unit 101 may be used, or one generated by the CPU 105 from a plurality of orthoimages.

Here, orthoimages and composite orthoimages will be explained. FIG. 2 is a diagram for explaining an orthoimage. FIG. 2A is a diagram showing an example of a captured image 200 captured in the vehicle traveling direction by an on-vehicle camera mounted on a vehicle or the like. The captured image may be a still image or may be one frame of a moving image captured by an on-vehicle camera. FIG. 2(b) is an example of an orthoimage 210 obtained by orthogonally transforming the captured image shown in FIG. 2(a). Orthographic conversion refers to converting an image so that it is displayed in the correct size and position without any tilt, as if the image were viewed from directly above. Generally speaking, it refers to converting an image captured by central projection, such as an aerial photograph or a satellite photograph, into an orthogonal projection mode, but it can also be used when converting an image captured by an on-vehicle camera, etc. can. Note that the orthoimage 210 in FIG. 2(b) shows only the roadway included in FIG. 2(a), but in reality, signboards, buildings, etc. may also be included.

The CPU 105 can create a map by connecting a plurality of orthoimages as shown in FIG. 2(b). Such an image is called a composite orthoimage. For example, assume that FIG. 3A is an example of a composite orthoimage 300. This composite ortho image 300 is actually a superimposed image generated by superimposing a plurality of ortho images 311 to 317 so that the captured contents overlap in each, as shown in FIG. 3(b). Become.

Such a composite orthoimage becomes a blurred image due to a shift in the superimposition of each orthoimage, and in some cases, it may not be possible to use it as data that is the source of a map image. Therefore, the CPU 105 quantitatively evaluates the composite orthoimage.

That is, the CPU 105 functions as an evaluation unit that evaluates the composite orthoimage. To this end, the CPU 105 determines the pixel value of each pixel of the orthoimage to be superimposed on an image that is included in a plurality of orthoimages that is the basis of the composite orthoimage and is located at a position where it is superimposed with other orthoimages. It also functions as a calculation unit that calculates the variance value of the pixel based on . Then, the CPU 105 generates evaluation information for evaluating the composite orthoimage based on the calculated variance value.

This will be explained with reference to FIG. 3(b). The frames extracted from the video captured using the in-vehicle camera are highly continuous, and the ortho images obtained by orthogonally converting these frames are superimposed on each other as shown in Figure 3(b). A composite orthoimage (see FIG. 3(a)) as one map image can be generated. The more orthoimages that are superimposed, the more accurate a composite orthoimage can be generated as a map image on average. At this time, if the superimposition positions of the orthoimages are shifted, the combined orthoimage may become a blurred image. Therefore, the CPU 105 calculates the variance of the pixel values of pixels where the orthoimages are superimposed on each other.

As an example, referring to FIG. 3B, the CPU 105 selects a pixel indicated by a position 320 of the ortho image 311, a pixel indicated by a position 320 of the ortho image 312 superimposed on the ortho image 311, and a pixel indicated by the position 320 of the ortho image 311. The pixel values of the pixel indicated by the position 320 of the ortho image 313 superimposed on the ortho image 312 and the pixel indicated by the position 320 of the ortho image 314 superimposed on the ortho images 311 to 313 are obtained. If the pixel values of pixels corresponding to the same position in multiple orthoimages have a high degree of agreement, it means that the orthoimages have a good degree of superimposition, and a highly accurate composite orthoimage has been obtained. . Therefore, the CPU 105 evaluates the composite orthoimage based on the variance value calculated for each pixel of the composite orthoimage. As an example, the CPU 105 may compile the variance values of each pixel of the composite orthoimage and generate a list of pixel variance values as evaluation information, or may select the highest variance value among the variance values calculated for each pixel. may be generated as the evaluation information, or the average value of the variance values may be generated as the evaluation information.

However, when generating the pixel with the highest variance value in the composite orthoimage as evaluation information, or generating the average value of the pixel values of the entire composite orthoimage as evaluation information, it is necessary to Sometimes it doesn't happen. Therefore, in order to output a more accurate evaluation, the CPU 105 divides the composite orthoimage into a plurality of small regions. FIG. 4 shows an example in which a composite orthoimage is divided into a plurality of small regions. Note that although it is described as dividing, this means that small areas are set for the composite orthoimage 300. FIG. 4 shows an example in which the composite orthoimage 300 is divided into small regions 401 to 415. It is desirable that these small regions have the same area, but they may be divided to have different areas. The small area may have a size of 400×400 pixels, for example, but may have a size other than this. The CPU 105 calculates, for each of the divided small regions, the average value of the variance values of pixels included in each small region. Then, the CPU 105 generates the highest variance value among the calculated average values as evaluation information of the composite orthoimage.

The CPU 105 causes the output unit 103 to output the generated evaluation information.

The above is an example of the configuration of the evaluation device 100.

<Operation of evaluation device>
Next, the operation of the evaluation device 100 will be explained. FIG. 5 is a flowchart showing an example of the operation of the evaluation device 100.

The CPU 105 acquires an image captured by an on-vehicle camera, which is the basis for generating a composite orthoimage. The CPU 105 extracts each frame from the acquired captured video. Then, the CPU 105 orthographically transforms each extracted frame to generate an orthoimage (step S501).

The CPU 105 generates a composite orthoimage by superimposing the same positions of the generated orthoimage as other orthoimages (step S502). The ortho images to be superimposed here are at least two ortho images, and three or more ortho images may be superimposed.

The CPU 105 calculates the variance value of the pixel at the position of the composite orthoimage from the pixel value of the pixel at the position where the plurality of orthoimages are superimposed (step S503). That is, for each pixel of the composite orthoimage, the CPU 105 acquires the pixel value of the pixel existing at the corresponding position in each of the original orthoimages, and calculates a variance value using each of the acquired pixel values. . That is, the CPU 105 specifies how much the distribution of pixel values of pixels at the same location that overlap each other is scattered from the expected value. It can be estimated that the smaller the variance, the more the pixel values of the superimposed pixels are the same and the smaller the deviation. The CPU 105 stores in the storage unit 104 the calculated variance value in association with each pixel (or the position coordinate of the pixel on the image) of the composite orthoimage.

The CPU 105 divides the composite orthoimage into a plurality of small regions (step S504). The example in FIG. 4 shows an example in which the composite orthoimage 300 is divided into 15 small areas 401 to 415.

Then, for each divided small area, the CPU 105 acquires the pixel variance value calculated for the pixels included in the small area from the storage unit 104. Then, the average value of the obtained variance values is calculated (step S505). That is, in the example of FIG. 4, the average value of the variance values of each of the small regions 401 to 415 is calculated. Therefore, the CPU 105 can specify, in units of small regions, how much the pixel values of pixels are shifted in each small region. It can be evaluated that the closer the average value of the variance values is to 0, the smaller the deviation of the pixel values of the pixels, that is, the smaller the deviation in the superimposition of orthoimages in a small area. Conversely, the closer the average value of the variance values is to 1, the greater the deviation in the pixel values of the pixels, that is, the larger the deviation in the superimposition of orthoimages in a small area (or the captured image may differ depending on the driving environment or driving conditions at the time of shooting). It can be evaluated that the image was not suitable for ortho-transforming and compositing.

The CPU 105 identifies the largest value among the average values of the variance values calculated for each of the divided small areas. Then, the CPU 105 causes the output unit 103 to output the identified largest value as the evaluation value of the composite orthoimage (step S506).

The above is a description of the method for evaluating a composite orthoimage using the evaluation device 100.

<Summary of Embodiment 1>
The evaluation device 100 according to the first embodiment calculates the variance of the pixel values of the orthoimages at the same position when superimposing the composite orthoimage and the orthoimages from which the composite orthoimage is generated. Based on the value, it can be quantitatively evaluated. Therefore, the composite orthoimage can be evaluated without the user visually checking the composite orthoimage.

<Embodiment 2>
In the first embodiment described above, the synthesized orthoimage is directly divided into small regions, the average of the variance values of the pixel values of the pixels in each small region is calculated, and the average value calculated for each small region is Among them, the highest value is output as the evaluation value of the composite orthoimage. By the way, when using a composite orthoimage for a map image, etc., what is important is mainly the road. Therefore, it can be considered that the composite orthoimage should be accurate as long as the image of the roadway portion is accurate. Therefore, in the second embodiment, a method for evaluating a composite orthoimage specialized for roadways will be described. Note that in the second embodiment, a different configuration from the first embodiment will be explained.

<Configuration of evaluation device>
In addition to the functions described in the first embodiment, the CPU 105 according to the second embodiment specifies a road area and a non-road area from the composite orthoimage. Then, the identified non-road area is removed (deleted) from the composite orthoimage. Then, the CPU 105 evaluates the composite orthoimage from which the non-road area has been removed.

FIG. 6(a) shows a composite ortho image for removing the non-road area, and FIG. 6(b) shows a composite ortho image after removing the non-road area.

The composite ortho image 300 shown in FIG. 6(a) is similar to the composite ortho image 300 shown in FIG. 3(a), and non-road areas are indicated by diagonal lines. That is, the composite orthoimage 300 includes a non-road area 601 and a non-road area 602. The CPU 105 removes the day's road area 601 and the non-road area 602 from the composite orthoimage 300, and generates a composite orthoimage 350 after removal, as shown in FIG. 6(b).

As an example of a method for detecting a non-road area (or road area) in the composite orthoimage 300, it can be realized by using deep learning semantic segmentation, but the method is not limited to this.

<Operation of evaluation device>
FIG. 7 is a flowchart showing an example of the operation of the evaluation device 100 according to the second embodiment. In FIG. 7, the processes that differ from the flowchart shown in FIG. 5 are steps S701 and S702.

As shown in FIG. 7, the CPU 105 identifies a non-road area in the composite orthoimage 300, and removes the identified non-road area from the composite orthoimage 300 (step S701).

Then, the CPU 105 divides the composite orthoimage 350 from which the non-road area has been removed into a plurality of small areas (step S702). The example in FIG. 6(b) shows an example in which the composite orthoimage 350 is divided into small regions 611 to 626.

This allows the CPU 105 to evaluate the composite orthoimage in which only the road area exists. Note that in this process, the calculation of the pixel variance value in step S503 may be performed after removing the non-road area.

<Summary of Embodiment 2>
According to the evaluation device 100 according to the present embodiment, it is possible to evaluate a composite orthoimage specialized for roads. Therefore, for example, a composite orthoimage suitable for creating map information can be specified.

<Supplement>
It goes without saying that the evaluation device according to the embodiment described above is not limited to the embodiment described above, and may be realized by other methods. Various modifications will be described below.

(1) In the above embodiment, the largest value among the average values of the variance values of pixels in a small area is output as the evaluation information of the composite orthoimage, but this is not limited to this. Other information may be used as long as the composite orthoimage can be quantitatively evaluated. For example, the evaluation device 100 may output the average value of the pixel variance values calculated for each small region as is, or may output a graph showing the distribution of the average values.

Further, the CPU 105 of the evaluation device 100 may normalize the variance value of each pixel to a value of 0 to 255 with respect to the composite orthoimage and output an image drawn in gray scale as the evaluation information. In an image in which the variance value is drawn in gray scale, the higher the brightness value, the higher the variance, and the area where there is a possibility that a shift during composition may have occurred.

Alternatively, the CPU 105 of the evaluation device 100 normalizes the variance value of the pixel value of each pixel of the composite orthoimage by dividing it by the standard deviation of the brightness of the composite orthoimage, and then It is also possible to calculate the value. In the road area, the variance value for each pixel tends to be high mainly near the edges of white lines (compartment lines), and the larger the difference in brightness between the white line and the road surface, the larger the variance value tends to be. Therefore, due to differences in exposure during photographing and the presence or absence of shadows, differences in brightness may occur between white lines and road surfaces, and the dispersion value may also change. Therefore, by performing normalization as described above, it is possible to perform evaluations that suppress the influence of brightness differences.

(2) In the embodiment described above, the CPU 105 may function as a correction unit that corrects (synthesizes) the combined orthoimage based on the evaluation information. Specifically, as an example, evaluation information calculated for the composite orthoimage is compared with a predetermined threshold. This predetermined threshold value is a value for determining whether the synthesized orthoimage is an image that satisfies the reference value.

If the evaluation information satisfies a predetermined threshold (the largest variance value output as evaluation information is less than the predetermined threshold), the synthesized orthoimage will be an image with a high evaluation value (satisfying the criteria). , the CPU 105 does not perform correction. On the other hand, if the evaluation information does not satisfy the predetermined threshold (the variance value is greater than or equal to the predetermined threshold), the CPU 105 determines the superimposition position of the ortho image corresponding to the small area for which the variance value was calculated. Fix it. Then, the variance value as the evaluation value is calculated again for the modified composite orthoimage. If this variance value is smaller than the previous variance value, it can be said that a good correction has been made; if the variance value is larger than the previous one, it can be said that a bad correction has been made.

The CPU 105 may correct the composite orthoimage by repeating corrections in this manner until the variance value satisfies a predetermined criterion.

(3) In the above embodiment, an example was shown in which the small areas are divided so that they do not overlap with each other, but the small areas may be set so that some of the small areas overlap with each other.

(4) Embodiment 2 above shows a configuration in which non-road areas are removed (deleted) from the composite orthoimage, but this does not involve removing non-road areas, but rather repainting them with a single color. You can do it like this. For example, the non-road areas in the composite orthoimage may be painted black or white, and then the composite orthoimage may be evaluated.

(5) In the above embodiment, it is described that the evaluation device 100 may receive input of a composite orthoimage and evaluate the received composite orthoimage. That is, in the embodiment described above, an example is also shown in which the evaluation device 100 orthogonally transforms a captured image, synthesizes the obtained orthoimages to generate a composite orthoimage, and evaluates the generated orthoimage. , for the evaluation device 100, the function of converting a captured image into an ortho image and the function of composing ortho images to generate a composite ortho image are not essential functions, and the evaluation device 100 does not have these functions. Tomoyoshi.

(6) In the above embodiment, the captured image is an image captured by an on-vehicle camera, but this may also be an image captured from the air by a drone, etc., or a satellite photo captured by a satellite, etc. It may be.

(7) In the above embodiment, as a method of evaluating a synthesized orthoimage in the evaluation device, the processor of the evaluation device executes an evaluation program etc. to create the image. It may be realized by a logic circuit (hardware) formed on an IC (Integrated Circuit) chip, an LSI (Large Scale Integration), or a dedicated circuit. Further, these circuits may be realized by one or more integrated circuits, and the functions of the plurality of functional units shown in the above embodiments may be realized by one integrated circuit. LSI is sometimes called VLSI, super LSI, ultra LSI, etc. depending on the degree of integration. That is, as shown in FIG. 8, the evaluation device 100 may include a receiving circuit 101a, a receiving circuit 102a, an output circuit 103a, a storage circuit 104a, and a control circuit 105a, and the receiving section 101 and the receiving section 102, respectively. , the output section 103, the storage section 104, and the CPU 105.

Further, the evaluation program may be recorded on a processor-readable recording medium, and the recording medium may be a "non-temporary tangible medium" such as a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit. etc. can be used. Further, the evaluation program may be supplied to the processor via any transmission medium (communication network, broadcast wave, etc.) that can transmit the evaluation program. That is, for example, the evaluation program may be downloaded from a network and executed using an information processing device such as a smartphone. The present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the evaluation program is embodied by electronic transmission.

Note that the evaluation program can be implemented using, for example, a script language such as ActionScript or JavaScript (registered trademark), or an object-oriented programming language such as Objective-C, Java (registered trademark), C++, Python, or R.

(8) The various examples shown in the above embodiment mode and the various examples shown in <Supplement> may be combined as appropriate. Furthermore, the operations shown in each flowchart may be rearranged in the order of execution or executed in parallel if there is no contradiction as a result.

100 Evaluation device 101 Receiving unit 102 Accepting unit 103 Output unit 104 Storage unit 105 CPU (acquisition unit, calculation unit, evaluation unit)

Claims (13)

An evaluation device that evaluates a composite orthoimage obtained by orthogonally converting each of a plurality of captured images so that locations where the captured contents match those of other orthoimages are superimposed on each other. ,
an acquisition unit that acquires each of the plurality of orthoimages and the composite orthoimage;
Calculation of calculating the variance value of a pixel included in the plurality of orthoimages and located at a position where it is superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed. Department and
An evaluation device comprising: an evaluation section that generates evaluation information for evaluating the composite orthoimage based on the variance value calculated by the calculation section. The evaluation device according to claim 1, further comprising an output unit that outputs the evaluation information obtained by the evaluation unit evaluating the composite orthoimage. The evaluation unit divides the composite orthoimage into small regions, calculates, for each small region, the average value of the variance values calculated by the calculation unit of each pixel included in the small region, and The evaluation device according to claim 1 or 2, wherein the largest value among the average values is used as evaluation information of the composite orthoimage. The evaluation according to any one of claims 1 to 3, wherein the evaluation unit removes images other than the road portion included in the composite orthoimage and evaluates the composite orthoimage of only the road portion. Device. An evaluation method for evaluating a composite orthoimage obtained by orthogonally converting each of a plurality of captured images so that locations where the captured content matches those of other orthoimages are superimposed on each other, ,
The computer is
an acquisition step of acquiring each of the plurality of orthoimages and the composite orthoimage;
Calculation of calculating the variance value of a pixel included in the plurality of orthoimages and located at a position where it is superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed. step and
and an evaluation step of generating evaluation information for evaluating the composite orthoimage based on the variance value calculated in the calculation step. The evaluation method according to claim 5, wherein the evaluation step includes an output step of outputting the evaluation information obtained by evaluating the composite orthoimage. In the evaluation step, the synthesized orthoimage is divided into small regions, and for each small region, the average value of the variance values calculated in the calculation step of each pixel included in the small region is calculated. The evaluation method according to claim 5 or 6, wherein the largest value among the average values is used as evaluation information of the composite orthoimage. The evaluation according to any one of claims 5 to 7, characterized in that in the evaluation step, images other than the road portion included in the composite orthoimage are removed and the composite orthoimage of only the road portion is evaluated. Method. An evaluation program that evaluates a composite orthoimage obtained by orthogonally converting each of a plurality of captured images so that locations where the captured content matches those of other orthoimages are superimposed on each other. ,
to the computer,
an acquisition function that acquires each of the plurality of orthoimages and the composite orthoimage;
Calculation of calculating the variance value of a pixel included in the plurality of orthoimages and located at a position where it is superimposed with another orthoimage, based on the pixel value of each pixel of the orthoimage to be superimposed. function and
An evaluation program that realizes an evaluation function that generates evaluation information for evaluating the composite orthoimage based on the variance value calculated by the calculation function. to the computer;
The evaluation program according to claim 9, wherein the evaluation function realizes an output function of outputting the evaluation information obtained by evaluating the composite orthoimage. The evaluation function divides the composite orthoimage into small regions, and calculates, for each small region, the average value of the variance values calculated by the calculation function of each pixel included in the small region. The evaluation program according to claim 9 or 10, wherein the largest value among the average values is set as evaluation information of the composite orthoimage. The evaluation according to any one of claims 9 to 11, wherein the evaluation function removes images other than the road portion included in the composite orthoimage and evaluates the composite orthoimage of only the road portion. program. A recording medium recording the evaluation program according to any one of claims 9 to 12.

Images