Image Processing Device Region-Based Correction Synthesis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing methods, such as those described in NPL 1 and NPL 2, face accuracy degradation in scenes with multiple pure substances and dark pixels or shadows, respectively, leading to suboptimal image quality after correction.

Innovation Solution

An image processing device that employs a correction unit to generate multiple correction images using different methods or parameters, an evaluation unit to assess each area's quality, and a synthesis unit to combine the best correction images based on evaluation results, thereby improving overall image quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a single correction method is used, then the correction process is simple, but the accuracy degrades in specific scenes (scenes with many pure substances or scenes without dark pixels)

Engineering Contradiction:

Improvecorrection accuracyVSAvoidcorrection process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the correction process by dividing the image into multiple regions and selecting different correction methods for each region based on local characteristics. This allows the system to achieve high accuracy in different scene types without applying a complex universal correction method to the entire image.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic selection of correction methods based on scene characteristics. The system automatically determines the appropriate correction method for each region by analyzing local image properties, enabling adaptive correction that maintains accuracy across varying scene conditions without requiring manual intervention.

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If multiple correction methods are applied to all areas, then correction accuracy improves, but processing time and computational load increase

Engineering Contradiction:

Improvecorrection accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies local quality by using different correction methods for different regions of the image based on their specific characteristics. Instead of uniformly applying multiple correction methods to the entire image, the system selects appropriate methods for each region, reducing overall processing time while maintaining accuracy where needed.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements partial action by applying correction methods only to regions where they are needed. The system identifies regions that benefit from specific correction approaches and applies those methods selectively, avoiding unnecessary processing in regions where simple correction suffices, thereby reducing total processing time.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentEP3506195B1Image processing device, image processing method, and computer-readable recording medium

Publication Date: 2023.11.01 NEC CORP

Figure 1~2
Figure 3~4
Figure 5~6

AI summary

To improve image quality of an image after correction. An image processing device 100 includes: a correction unit 110 for generating a plurality of correction images from an image by using a plurality of correction methods; an evaluation unit 120 for evaluating the image or each of the plurality of correction images for each of areas constituting the image or the correction image; and a synthesis unit 130 for synthesizing the plurality of correction images according to an evaluation result by the evaluation unit 120.