Differential Sub-Pixel Layout for Low-Data Image Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image sensing and processing systems face challenges with high data redundancy leading to increased data transmission and processing delays and power consumption, particularly in applications like intelligent visual recognition and edge computing, where efficient image processing is needed without increasing pixel area or circuit complexity.
Innovation Solution
A pixel unit comprising N-type and P-type main pixels and a sub-pixel, where the sub-pixel generates and outputs a signal difference between the two main pixels based on exposure intensity, utilizing an N-type MOS transistor structure, buried oxide layer, and shallow trench isolation to efficiently process image signals directly, reducing output data without adding complexity to the circuit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If pre-processing circuits such as comparators and convolution calculation circuits are added in or around the pixels, then image processing efficiency is improved and redundant data is filtered out, but the circuit complexity increases and pixel area expands
Solution Approach 1:
The patent merges the pre-processing function into the pixel unit itself by integrating a computing unit within the pixel structure. This allows the pixel to perform local computation (such as gradient calculation) directly on the captured light signal, combining the functions of light reception and signal processing in a single integrated unit, thereby improving processing efficiency without requiring separate external pre-processing circuits
Solution Approach 2:
The pixel unit is designed with multi-functionality, serving both as a light-receiving element and a computation unit. The computing unit can perform various operations including gradient calculation, edge detection, and other pre-processing tasks, making the pixel structure universal and adaptable to different image processing requirements without needing dedicated specialized circuits for each function
2Productivity
If pre-processing circuits such as comparators and convolution calculation circuits are added in or around the pixels, then image processing efficiency is improved and redundant data is filtered out, but the pixel area increases
Solution Approach 1:
The computing unit is nested within the pixel unit structure, with the computation elements integrated into the same physical footprint as the light-receiving element. This nested arrangement allows the processing functionality to be contained within the existing pixel area boundaries, avoiding the need for additional external circuitry that would increase the overall pixel area
3Reliability
If complete image data is transmitted to an image signal processor for processing, then imaging quality is maintained, but data transmission amount increases leading to tremendous pressure, delay and power consumption
Solution Approach 1:
The patent extracts and processes only the essential feature information (such as gradient values, edge information, or other salient features) directly at the pixel level, separating this extracted feature data from the complete raw image data. Only the extracted features are transmitted to the image signal processor, while the full-resolution raw data remains at the sensor level or is discarded, significantly reducing the transmission data volume while preserving the information needed for high-quality image processing
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for efficient image processing by generating calculation results simultaneously with original signals, reducing data transmission and processing delays, and maintaining the pixel area and imaging quality without increasing circuit complexity.
Implementation Method 1
both the N-type main pixel and the P-type main pixel are configured to output a current corresponding to an exposure intensity to the sub-pixel according to exposure
Data Source
AI summary
The present application discloses a pixel unit and a signal processing method for a pixel unit. The pixel unit includes at least one pixel, and the pixel includes: an N-type main pixel, a P-type main pixel, and a sub-pixel; and the sub-pixel is located between the N-type main pixel and the P-type main pixel; or the pixel includes at least a first pixel and a second pixel that are adjacent to each other; the first pixel includes an N-type main pixel, and the second pixel includes a P-type main pixel; the first pixel and the second pixel share one sub-pixel; the sub-pixel is configured to generate and output a signal difference between the N-type main pixel and the P-type main pixel according to the current. By adding a sub-pixel between two main pixels, the sub-pixel generates and outputs the signal difference between the N-type main pixel and the P-type main pixel according to the current sent by the two main pixels, so that the received signal can be efficiently processed directly to reduce the amount of output data. Since there is no need to increase a circuit, the pixel area will not increase due to a complicated circuit.