Image Processing Device for High-Resolution Object Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In high-resolution images, the amount of processing required for object detection and recognition increases significantly, making it inefficient for real-time processing of small objects.

Innovation Solution

An image processing device that acquires a lower resolution image from a high-resolution image, classifies objects in the lower resolution image, identifies corresponding object areas in the high-resolution image, and performs recognition processing on these areas.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If recognition processing is performed on the entire high-resolution image, then recognition accuracy is improved, but processing time increases

Engineering Contradiction:

Improverecognition accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent divides the high-resolution image into multiple low-resolution images through downscaling. Recognition processing is then performed on these smaller low-resolution images instead of the entire high-resolution image, significantly reducing processing time while maintaining acceptable recognition accuracy for the purpose of identifying regions of interest.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary downscaling of the high-resolution image to create low-resolution images before conducting recognition processing. This preliminary action reduces the computational burden and processing time, allowing for faster identification of regions of interest that can then be analyzed in detail.

Inventive Principle:

Principle #10Preliminary action

2Productivity

If recognition processing is performed on a low-resolution image, then processing time is reduced, but recognition accuracy deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidrecognition accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the image processing task into two stages: first processing low-resolution images for rapid identification of regions of interest, then applying the same processing to the corresponding high-resolution images to achieve accurate recognition. This segmentation allows the system to benefit from both fast preliminary processing and accurate final recognition.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a multi-resolution dimension by creating and processing images at different resolutions (low-resolution and high-resolution). This dimensional approach allows the system to perform rapid preliminary analysis at low resolution and then enhance results at high resolution, effectively resolving the trade-off between speed and accuracy.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Reliability

If the entire high-resolution image is processed, then complete object detection is achieved, but processing resources increase

Engineering Contradiction:

Improvedetection completenessVSAvoidprocessing resources

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent segments the high-resolution image into multiple low-resolution images for initial processing. This segmentation reduces the computational resources required for the first pass of recognition processing, allowing the system to efficiently identify regions of interest before applying more resource-intensive processing only to those specific regions in the high-resolution image.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs partial processing by first analyzing only low-resolution versions of the images to identify regions of interest. This partial action approach allows the system to achieve detection completeness by subsequently processing the corresponding high-resolution regions, while minimizing overall resource consumption by avoiding full high-resolution processing of the entire image.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12236659B2Image processing device to detect an object in a high-resolution image

Publication Date: 2025.02.25 SONY GROUP CORP

AI summary

The present disclosure relates to an image processing device, an image processing method, and a program that can reduce the amount of processing required for a series of processing from detection to recognition of an object in a high-resolution image.An acquisition unit acquires, from a first resolution image, a second resolution image having a lower resolution than the first resolution image, a classification unit classifies an object included in the second resolution image, an identification unit identifies an object area corresponding to the object of a predetermined classification in the first resolution image, and a recognition unit performs recognition processing of the object on the object area identified in the first resolution image. The technology according to the present disclosure can be applied to a camera system of a remote control tower, for example.