Image Content Extraction via Contrast Fusion for Writing Tablets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image content extraction methods using threshold segmentation fail to accurately extract characters and graffiti from writing tablets due to varying pixel values, leading to incomplete and broken content extraction.
Innovation Solution
The method involves acquiring an image, performing high-contrast retention and image fusion, followed by linear light enhancement to enhance pixel value differences between image and non-image content regions, using Gaussian filtering and bilateral filtering to sharpen and highlight boundaries, and employing a writing tablet recognition model for precise extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If threshold segmentation is used to extract image content, then the extraction process is simple and fast, but the extraction completeness and integrity deteriorate due to varying pixel values of image contents
Solution Approach 1:
The image processing is divided into multiple sequential stages: high-contrast retention to extract edge information, image fusion to combine original and enhanced images, and linear light enhancement to further amplify pixel value differences. This multi-stage segmentation approach allows each stage to focus on specific enhancement goals, ultimately achieving both speed and integrity in content extraction
Solution Approach 2:
The high-contrast retention operation selectively enhances only the edge regions and boundaries of image contents while preserving the overall image structure. By applying different processing intensities to different regions (edges vs. interiors), the method highlights content boundaries without uniformly degrading the entire image, thus maintaining extraction integrity
2Manufacturing precision
If high-contrast retention and image fusion are performed to enhance pixel value differences, then the image content integrity is improved, but the processing complexity increases
Solution Approach 1:
A fused image is introduced as an intermediary between the original image and the final enhanced image. The fused image combines the original image information with the high-contrast retained image, serving as a bridge that gradually enhances content distinguishability. This intermediary approach breaks down the complex enhancement process into manageable steps, reducing overall processing complexity while maintaining integrity
Solution Approach 2:
The high-contrast retention operation is performed preliminarily to extract and emphasize edge information before the main enhancement process. By preparing the edge-enhanced image in advance and fusing it with the original image, the subsequent linear light enhancement operates on already-prepared data, reducing the computational burden and complexity of the final enhancement stage
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
An image content extraction method and apparatus (100), and a terminal (11) and a storage medium. The image content extraction method specifically comprises: acquiring an image to be processed (S101); performing high-contrast retention processing on the image to be processed, so as to obtain a high-contrast image of the image to be processed (S102); performing image fusion on the image to be processed and the high-contrast image, so as to obtain a fused image (S103); performing linear light enhancement processing on the fused image, so as to obtain a linear light enhanced image (S104); and using, as image content of the image to be processed, a first pixel point, the pixel value of which is within a preset pixel value range, in the linear light enhanced image (S105). The integrity of image content extraction can be improved.