Image Processing Apparatus Texture-AI Hybrid Super-Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image quality improvement techniques, especially those utilizing artificial intelligence, struggle to effectively enhance image quality, particularly in texture areas with high frequency components, while also facing challenges related to computational complexity and battery consumption in devices like TVs and mobile terminals.
Innovation Solution
An image processing apparatus and method that employs machine learning for self-transformation of images, adjusting parameters to minimize differences in visual and structural features between the processed image and a high-quality original image, thereby generating a high-quality image while reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If AI technology is applied to image quality improvement, then image quality can be enhanced, but computational complexity and battery consumption increase significantly
Solution Approach 1:
The image is divided into texture areas and non-texture areas, and different processing methods are applied to each region. Texture areas use traditional interpolation methods while non-texture areas use AI-based super-resolution, reducing overall computational complexity while maintaining image quality
Solution Approach 2:
Different processing qualities are applied to different regions of the image based on their characteristics. Texture-rich regions receive lighter processing while smooth regions receive more intensive AI processing, optimizing the balance between quality improvement and computational load
2Manufacturing precision
If AI technology is applied to image quality improvement, then image quality can be enhanced, but battery consumption increases significantly
Solution Approach 1:
The image processing is segmented into different regions requiring different computational resources. By identifying texture areas and applying appropriate processing methods, the overall energy consumption is reduced while maintaining image quality
Solution Approach 2:
AI-based processing is applied partially only to non-texture areas rather than the entire image. This partial application reduces computational burden and battery consumption while still achieving meaningful image quality improvement in the most beneficial regions
3Device complexity
If traditional image processing methods are used, then computational complexity is low, but texture representation quality is poor
Solution Approach 1:
Different processing approaches are applied to different image regions based on their texture characteristics. Smooth regions use traditional methods while texture-rich regions use AI-based methods, optimizing both quality and computational efficiency
Solution Approach 2:
The patent uses an intermediary approach by combining traditional interpolation methods with AI-based super-resolution. The traditional methods handle the bulk of the processing efficiently while AI enhances specific regions, creating a balanced solution
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
An image processing apparatus is provided. The image processing apparatus according to an exemplary embodiment includes a communicator configured to receive an image, and a processor configured to generate a first image obtained by performing image processing on the received image by using a parameter for image processing, generate a second image obtained by reducing the first image at a predetermined ratio, and extract respective visual features from the first image and the second image, wherein the processor is further configured to adjust the parameter to allow a difference between the visual feature of the first image and the visual feature of the second image to be within a predetermined range.