Collating system, collating method, and camera device

1. A collating system comprises:

a camera device configured to capture image data of a monitoring target area; and

a server device communicably connected to the camera, and the server device is configured to

store a plurality of reference images from a plurality of viewpoints;

store an image correction model;

process the reference images using the image correction model to generate a plurality of corresponding sharpened reference images;

store the sharpened reference images;

send, to the camera device, the image correction model,

wherein the camera device is configured to

detect an object appearing in the captured image data of the monitoring target area,

generate a target image of the object from the captured image data,

generate, using the image correction model received from the server device, a sharpened image based on the target image of the object, and

send the sharpened image to the server device, and

wherein the server device is configured to:

receive and store the sharpened image sent from the camera device,

perform collating processing of the object corresponding to the sharpened image based on the sharpened image generated using the image correction model, and the plurality of sharpened reference images generated using the same image correction model used by the camera, and

send a result of the collating processing including identification information of the object to an external device,

wherein the server device is further configured to

store a first image set including an image related to the object,

improve an image quality degradation factor caused by capturing of the object based on the image related to the object and the image correction model, and

add the generated sharpened image in a second image set stored in the server device,

wherein the image quality degradation factor is unevenness occurring in a partial region of the object appearing in the captured image due to sunlight, and

wherein the server device

generates a pseudo unevenness in a partial region of the image related to the object read from the first image set,

generates a sharpened image corresponding to the image having the pseudo unevenness based on the image correction model, and

adds the sharpened image in the second image set stored in the server device.

2. The collating system according to claim 1 ,

wherein the server device is configured to perform the collating processing based on the first image set, the second image set, and the image correction model.

3. The collating system according to claim 1 ,

wherein the first image set includes k number of different images related to the object,

wherein the k is an integer that greater than or equal to 2, and

wherein the second image set includes the k number of different sharpened images corresponding to the respective images related to the object stored in the first image set stored in the server device.

4. The collating system according to claim 3 ,

wherein the server device performs learning processing for generating the k number of different sharpened images corresponding to the respective images related to the object included in the first image set before starting actual operation of the collating system.

5. The collating system according to claim 3 ,

wherein the server device performs learning processing for generating the k number of different sharpened images corresponding to the respective images related to the object included in the first image set during actual operation of the collating system.

6. The collating system according to claim 5 ,

wherein the camera device sends parameters related to capturing processing performed at the time of generating the target image of the object, the sharpened image, and the captured image to the server device in association with identification information of the object, and

wherein the server device stores the parameters sent from the camera device in association with the sharpened image and executes the learning processing based on the parameters.

7. The collating system according to claim 1 ,

wherein the image quality degradation factor is low illuminance noise occurring in the object appearing in the captured image due to night capturing, and

wherein the server device

generates a pseudo low illuminance noise in an entire image related to the object read from the first image set,

generates a sharpened image corresponding to the image having the pseudo low illuminance noise based on the image correction model, and

adds the sharpened image in the second image set stored in the server device.

8. The collating system according to claim 1 ,

wherein the image quality degradation factor is white saturation occurring in the object appearing in the captured image due to IR light emitted from an IR illuminator included in the camera device at night, and

wherein the server device

generates a pseudo white saturation in an entire image related to the object read from the first image set,

generates a sharpened image corresponding to the image having the pseudo white saturation based on the image correction model, and

adds the sharpened image in the second image set stored in the server device.

9. The collating system according to claim 1 ,

wherein the server device has a registration database that stores face images of a plurality of suspicious persons in advance, and

wherein in a case where the server device extracts a face image that corresponds with the sharpened image used in the collating processing, the server device sends, to the external device, the collating processing result in which the extracted face image and the target image of the object are associated with the identification information of the object.

10. The collating system according to claim 1 ,

wherein the camera device sends, to the server device, the target image along with the sharpened image.

11. The collating system according to claim 1 ,

wherein the camera device sends, to the server device, the sharpened image in association with identification information of the object.

12. The collating system according to claim 1 ,

wherein the monitoring target area is located outdoors and subject to environmental factors.

13. The collating system according to claim 1 ,

wherein the image correction model is generated based on existing learning dataset in the server device, and is updated with additional learning dataset provided by the camera device.

14. A collating method implemented by a collating system including a camera device and a server device that are communicably connected to each other,

the collating method comprising:

storing, by the server device, a plurality of reference images from a plurality of viewpoints;

storing, by the server device, an image correction model;

processing, by the server device, the reference images using the image correction model to generate a plurality of corresponding sharpened reference images;

storing, by the server device, the sharpened reference images;

capturing, by the camera device, image data of a monitoring target area,

receiving, by the camera device and from the server device, the image correction model stored at the server device,

detecting, by the camera device, an object appearing in the captured image data of the monitoring target area and generating a target image of the object from the captured image data,

generating, by the camera device and using the correction model received from the server device, a sharpened image to be used for collating processing of the object performed by the server device based on the target image of the object,

sending, by the camera device, the sharpened image to the server device,

receiving and storing, by the server device, the sharpened image sent from the camera device,

performing, by the server device, collating processing of the object corresponding to the sharpened image based on the sharpened image generated using the image correction model, and the plurality of sharpened reference images generated using the same image correction model used by the camera,

sending, by the server device, a result of the collating processing including identification information of the object to an external device,

storing, by the server device, a first image set including an image related to the object,

improving, by the server device, an image quality degradation factor caused by capturing of the object based on the image related to the object and the image correction model,

adding, by the server device, the generated sharpened image in a second image set stored in the server device,

wherein the image quality degradation factor is unevenness occurring in a partial region of the object appearing in the captured image due to sunlight,

generating, by the server device, a pseudo unevenness in a partial region of the image related to the object read from the first image set,

generating, by the server device, a sharpened image corresponding to the image having the pseudo unevenness based on the image correction model, and

adding, by the server device, the sharpened image in the second image set stored in the server device.

15. A collating system comprises:

a camera device configured to capture image data of a monitoring target area; and

a server device communicably connected to the camera, and the server device is configured to store a plurality of reference images from a plurality of viewpoints;

store an image correction model;

process the reference images using the image correction model to generate a plurality of corresponding sharpened reference images;

store the sharpened reference images;

send, to the camera device, the image correction model,

wherein the camera device is configured to

detect an object appearing in the captured image data of the monitoring target area,

generate a target image of the object from the captured image data,

generate, using the image correction model received from the server device, a sharpened image based on the target image of the object, and

send the sharpened image to the server device,

wherein the server device is configured to:

receive and store the sharpened image sent from the camera device,

perform collating processing of the object corresponding to the sharpened image based on the sharpened image generated using the image correction model, and the plurality of sharpened reference images generated using the same image correction model used by the camera, and

send a result of the collating processing including identification information of the object to an external device,

wherein the server device is further configured to

store a first image set including an image related to the object,

improve an image quality degradation factor caused by capturing of the object based on the image related to the object and the image correction model, and

add the generated sharpened image in a second image set stored in the server device

wherein the image quality degradation factor is a shadow occurring in a partial region of the object appearing in the captured image due to sunlight, and

wherein the server device

generates a pseudo shadow in a partial region of the image related to the object read from the first image set,

generates a sharpened image corresponding to the image having the pseudo shadow based on the image correction model, and

adds the sharpened image in the second image set stored in the server device.

16. A collating method implemented by a collating system including a camera device and a server device that are communicably connected to each other,

the collating method comprising:

storing, by the server device, a plurality of reference images from a plurality of viewpoints;

storing, by the server device, an image correction model;

processing, by the server device, the reference images using the image correction model to generate a plurality of corresponding sharpened reference images;

storing, by the server device, the sharpened reference images;

capturing, by the camera device, image data of a monitoring target area,

receiving, by the camera device and from the server device, the image correction model stored at the server device,

detecting, by the camera device, an object appearing in the captured image data of the monitoring target area and generating a target image of the object from the captured image data,

generating, by the camera device and using the correction model received from the server device, a sharpened image to be used for collating processing of the object performed by the server device based on the target image of the object,

sending, by the camera device, the sharpened image to the server device,

receiving and storing, by the server device, the sharpened image sent from the camera device,

performing, by the server device, collating processing of the object corresponding to the sharpened image based on the sharpened image generated using the image correction model, and the plurality of sharpened reference images generated using the same image correction model used by the camera, and

sending, by the server device, a result of the collating processing including identification information of the object to an external device,

storing, by the server device, a first image set including an image related to the object,

improving, by the server device, an image quality degradation factor caused by capturing of the object based on the image related to the object and the image correction model, adding, by the server device, the generated sharpened image in a second image set stored in the server device,

wherein the image quality degradation factor is a shadow occurring in a partial region of the object appearing in the captured image due to sunlight,

generating, by the server device, a pseudo shadow in a partial region of the image related to the object read from the first image set,

generating, by the server device, a sharpened image corresponding to the image having the pseudo shadow based on the image correction model, and

adding, by the server device, the sharpened image in the second image set stored in the server device.