Image Encoding Block Quantization for Edge Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional DPCM-based predictive encoding schemes suffer from significant image quality degradation at steep edge portions due to large pixel differences, requiring quantization that halves the original pixel data bits, leading to substantial image quality loss.
Innovation Solution
An image processing apparatus that acquires encoding blocks of pixels, decides quantization parameters and encoding schemes to maintain a target code length, selecting between PCM and DPCM encoding schemes and adjusting quantization steps to minimize code length, thereby reducing image quality degradation at steep edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If DPCM-based predictive encoding scheme is used to reduce circuit scale and encoding delay, then encoding speed and circuit complexity are improved, but image quality degradation occurs at edge portions with large pixel variations
Solution Approach 1:
The encoding block is divided into multiple pixel groups, and different encoding schemes (PCM or DPCM) are selectively applied to different pixel groups based on their local characteristics. This segmentation allows the system to maintain high encoding speed while preserving image quality in regions with large pixel variations.
Solution Approach 2:
The encoding scheme is dynamically selected for each pixel group based on the difference value between adjacent pixels. When the difference exceeds a threshold, PCM encoding is used; otherwise, DPCM encoding is used. This dynamic adaptation resolves the contradiction between encoding speed and image quality.
2Quantity of substance
If quantization is applied to halve the original pixel data bits when difference value exceeds threshold, then code length is reduced, but a large amount of image quality degradation is caused
Solution Approach 1:
Different quantization parameters are assigned to different pixel groups based on their local characteristics. Pixel groups with large differences use one quantization parameter while those with small differences use another, allowing efficient code length control without uniform quality degradation across the entire block.
Solution Approach 2:
The system changes the quantization parameter dynamically based on the difference value between adjacent pixels. By selecting from multiple quantization parameters, the system optimizes the balance between code length and image quality for each pixel group rather than applying a fixed quantization level.
3Manufacturing precision
If PCM encoding scheme is used to maintain image quality, then image quality degradation is reduced, but code length increases
Solution Approach 1:
The encoding block is segmented into multiple pixel groups, allowing PCM encoding to be applied selectively only to groups that require it (those with large pixel differences), while other groups use more efficient DPCM encoding. This reduces the overall code length while maintaining image quality where necessary.
Solution Approach 2:
Instead of applying PCM encoding to the entire block (excessive action), the system applies it only to the necessary pixel groups (partial action), optimizing the balance between image quality and code length.
Data Source
AI summary
An image processing apparatus comprising an acquiring unit configured to acquire an encoding target block having a plurality of groups each including a predetermined number of pixels, a deciding unit configured to decide, for each group, a quantization parameter used to quantize image data of the group and an encoding scheme so that a code length of the encoding target block does not exceed a predetermined value corresponding to a target code amount of the encoding target block, and an encoding unit configured to generate encoded data by encoding image data of the encoding target block in accordance with the quantization parameters and the encoding schemes decided for the respective groups by the deciding unit.