Granted Patent B2

US 12,236,669 · App. 17/750,440 · Granted Feb 25, 2025

Inventors: Takanori Nakao (Kawasaki, JP); Tomonori Kubota (Kawasaki, JP)

Assignee: FUJITSU LIMITED

G06V10/771H04N19/124H04N19/136H04N19/176G06V10/82

Abstract

Claims (22)

1. An image processing system comprising:

one or more memories; and

one or more processors coupled to the one or more memories and the one or more processors configured to:

acquire a degree of influence of each region of image data on a recognition result when recognition processing is performed for the image data;

reduce data size of the image data based on the degree of influence, the degree of influence of each region on the recognition result is aggregated for each processing block of when compression processing is performed for the image data;

determine a quantization value of each processing block of when the compression processing is performed for the image data according to a value of an aggregation result of each processing block; and

correct the determined quantization value of the each processing block based on change amounts in position and size of a target object between first image data used when calculating the degree of influence of each region on the recognition result and second image data for which the compression processing is performed applying the determined quantization value of the each processing block.

2. The image processing system according to claim 1 , wherein the one or more processors are further configured to acquire the degree of influence of each region of the image data on the recognition result by using one selected from a BP method, a GBP method, and a selective BP method.

3. The image processing system according to claim 1 , wherein the one or more processors are further configured to:

generate first compressed data by performing compression processing for first difference image data that is a difference between image data that includes a target object and image data that does not include the target object at a compression level according to the degree of influence of each region of the image data that includes the target object on the recognition result, and

generate second compressed data by performing compression processing for second difference image data that is a difference between the image data that includes target object and the first difference image data at a certain compression level.

4. The image processing system according to claim 1 , wherein the one or more processors are further configured to determine a method of processing treatment for each processing block of reducing the data size of the image data according to a value of an aggregation result of each processing block.

5. An image processing method for a computer to execute a process comprising:

acquiring a degree of influence of each region of image data on a recognition result when recognition processing is performed for the image data;

reducing data size of the image data based on the degree of influence, the degree of influence of each region on the recognition result is aggregated for each processing block of when compression processing is performed for the image data;

determining a quantization value of each processing block of when the compression processing is performed for the image data according to a value of an aggregation result of each processing block; and

correcting the determined quantization value of the each processing block based on change amounts in position and size of a target object between first image data used when calculating the degree of influence of each region on the recognition result and second image data for which the compression processing is performed applying the determined quantization value of the each processing block.

6. A non-transitory computer-readable storage medium storing an image processing program that causes at least one computer to execute a process, the process comprising:

acquiring a degree of influence of each region of image data on a recognition result when recognition processing is performed for the image data;

reducing data size of the image data based on the degree of influence, the degree of influence of each region on the recognition result is aggregated for each processing block of when compression processing is performed for the image data;

determining a quantization value of each processing block of when the compression processing is performed for the image data according to a value of an aggregation result of each processing block; and

correcting the determined quantization value of the each processing block based on change amounts in position and size of a target object between first image data used when calculating the degree of influence of each region on the recognition result and second image data for which the compression processing is performed applying the determined quantization value of the each processing block.

Assignments (1)

Continuity (2)

References Cited (16)