Inter-block Prediction Data Compression for Image Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensor technologies face challenges in efficiently compressing data-charge phase data from area analogue-to-digital sensors, leading to suboptimal image quality and increased settling times, especially in high-speed camera applications.
Innovation Solution
The implementation of inter-block prediction techniques for data compression, which involves computing prediction values, quantizing residual data, and entropy encoding, while assigning an additional bit-budget to the first block of image data, improves both subjective and objective performance compared to intra-prediction based methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If intra-prediction based methods are used for data compression, then device complexity is reduced, but image quality and compression efficiency deteriorate
Solution Approach 1:
The patent introduces an inter-block prediction mechanism that uses previously decoded blocks as reference data for predicting current blocks. This intermediary approach leverages spatial redundancy across block boundaries, achieving superior compression efficiency and image quality compared to intra-prediction, while maintaining computational complexity through systematic reference block selection and prediction mode evaluation.
2Manufacturing precision
If additional bit-budget is assigned to the first block, then image quality is improved, but data transmission volume increases
Solution Approach 1:
The patent implements non-uniform bit-budget allocation where the first block receives additional bits for higher precision encoding, while subsequent blocks use standard bit allocation. This local quality enhancement focuses computational resources on blocks that benefit most from higher precision (typically blocks with complex edge information or high-frequency content), improving overall image quality without proportionally increasing total data transmission volume.
Solution Approach 2:
The patent dynamically adjusts the bit-budget parameter based on block position and content characteristics. By changing the precision parameter for the first block versus subsequent blocks, the system optimizes the trade-off between image quality and data transmission volume, allocating more bits where they provide maximum visual improvement while maintaining efficiency elsewhere.
3Productivity
If inter-block prediction is used for data compression, then compression efficiency is improved, but settling time increases
Solution Approach 1:
The patent performs prediction operations in advance during the decoding process, using previously decoded blocks to predict current block values before actual decoding occurs. This preliminary prediction action reduces the computational burden during critical settling phases, allowing the system to achieve high compression efficiency while minimizing the time required for the sensor to stabilize and produce usable image data.
Data Source
AI summary
Implementations generally relate to data-charge phase data compression. In one implementation, a method includes computing prediction values for image data, where the image data is data-charge phase data, where the computing of prediction values is based on inter-block prediction, and where an additional bit-budget is assigned to a first block of the image data. The method also includes computing residual data based on the prediction values. The method also includes quantizing the residual data. The method also includes entropy encoding the quantized residual data. The method also includes refining an inverse quantized residual data based on one or more of the residual data and a number of left-over bit-budget after entropy encoding.