Image Enhancement Fusion for Local Detail Without Ghosting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image quality enhancement processing in computer vision applications is not effective in capturing and enhancing sufficient local detail information, often resulting in inconsistent pixel contents and ghost images.
Innovation Solution
An image enhancement method that adjusts pixel values of a to-be-processed image to generate multiple images with different pixel values, extracts local and global features, and performs image enhancement processing based on these features to achieve consistent content and improved detail capture.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image quality enhancement processing is used, then the processing speed is maintained, but the ability to capture and enhance local detail information is insufficient
Solution Approach 1:
The patent segments the image enhancement task into multiple processing streams: original image processing, blurred image generation, and super-resolution processing. By dividing the enhancement process into distinct functional modules (blurring unit, super-resolution unit, fusion unit), the system can capture local details at different scales without overwhelming computational complexity
Solution Approach 2:
The patent introduces a new dimension by creating multiple versions of the image with different blur levels and resolutions. Instead of processing a single image, the system operates in a multi-dimensional space where images with varying degrees of blurring are processed simultaneously to extract details at different scales, then fuses them to achieve enhanced local detail capture
2Measurement precision
If multiple images with different pixel values are generated, then more local detail information is obtained, but pixel offsets and ghost images may occur
Solution Approach 1:
The patent introduces a fusion unit that acts as an intermediary between the processed original image and the generated blurred images. This fusion unit carefully combines the images with different blur levels, ensuring proper alignment and weighting to prevent pixel offsets and ghost images while preserving the beneficial local detail information from multiple sources
Solution Approach 2:
The patent systematically varies the blur parameter (kernel size) to generate images with different levels of blurring. By controlling and adjusting this parameter across multiple processing streams, the system obtains complementary local detail information at different scales while maintaining consistent pixel content through the fusion process
3Manufacturing precision
If conventional enhancement methods are used, then the processing is simple, but the enhancement effect is not perfect
Solution Approach 1:
The patent divides the enhancement system into three main functional segments: a blurring unit that generates blurred versions of the input image, a super-resolution unit that processes both original and blurred images, and a fusion unit that combines the results. This segmentation allows each module to specialize in specific tasks, improving overall enhancement quality while keeping individual module complexity manageable
Solution Approach 2:
The patent implements dynamic processing where the blur kernel size and super-resolution parameters can be adjusted based on input image characteristics. The system dynamically adapts the processing strength and parameters to match the specific enhancement needs of different images, achieving higher quality results across diverse input conditions
Data Source
AI summary
This application relates to an image enhancement technology in the field of computer vision in the field of artificial intelligence, and provides an image enhancement method and apparatus. This application relates to the field of artificial intelligence, and specifically, to the field of computer vision. The method includes: adjusting a pixel value of a to-be-processed image, to obtain K images, where pixel values of the K images are different, and K is a positive integer greater than 1; extracting local features of the K images; extracting a global feature of the to-be-processed image; and performing image enhancement processing on the to-be-processed image based on the global feature and the local features, to obtain an image-enhanced output image. This method helps to improve the effect of image quality enhancement processing.