Variable Bit-Width Data Path for Image Sensor Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor imaging devices suffer from computational errors in their image processing data paths due to fixed bit bus width, leading to loss of fractional data and magnified errors during digital gain amplification, which affects image quality.
Innovation Solution
Implementing an image processing data path with a variable bus bit width that includes an additional M-bit portion to preserve fractional data, allowing for accurate output even with maximum digital gain, specifically defined as N+M-bit wide data paths for column fixed pattern noise correction and digital gain processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a fixed bit bus width is used in the image processing data path, then the device complexity is reduced and manufacturing is simplified, but computational errors occur due to loss of fractional data during digital gain amplification
Solution Approach 1:
The data path is segmented into two distinct portions: an N-bit portion for integer data and an additional M-bit portion for fractional data. This segmentation allows the system to preserve computational precision during digital gain operations while maintaining a manageable structure. The fractional portion is selectively extended only where needed in the data path, avoiding unnecessary complexity throughout the entire system.
Solution Approach 2:
The data path structure is made dynamic by conditionally extending the bit width from N-bit to N+M-bit in specific segments where digital gain amplification occurs. This dynamic adaptation allows the system to increase precision only where computational errors would otherwise occur, rather than uniformly increasing precision throughout the entire data path, thus balancing image quality with device complexity.
2Manufacturing precision
If the bus bit width is increased uniformly throughout the data path, then computational errors are reduced, but the device complexity and area increase
Solution Approach 1:
Instead of uniformly increasing the bus bit width throughout the entire data path, the invention applies local quality by extending the bit width to N+M-bit only in specific local segments where digital gain amplification occurs and computational errors would otherwise be magnified. Other portions of the data path maintain the standard N-bit width, thus reducing the total area required while still achieving the necessary computational accuracy where it is most needed.
3Device complexity
If fractional data is truncated to maintain fixed bit width, then device complexity is reduced, but image quality deteriorates due to magnified errors during amplification
Solution Approach 1:
The invention applies preliminary action by preserving the fractional data in an extended M-bit portion before digital gain amplification occurs. By maintaining this additional precision data in advance through the amplification process, the system prevents the magnification of truncation errors that would otherwise occur if fractional data were truncated beforehand. This preliminary preservation of precision ensures high image quality without requiring complex error correction mechanisms.
Data Source
AI summary
An imager device having a varied bus bit width in the image processing data-path. The imager device has a high accuracy output irrespective of digital gain value. The imager device can be adopted for all sensors that have column fixed pattern noise correction and digital gain in their data-path block.