Pixel Collection Circuit for High-Speed Optical Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional optical flow methods are inefficient in processing high-speed object movement due to redundant data generation and short time delays, requiring a faster analysis mechanism to accurately determine object movement.
Innovation Solution
A pixel collection circuit and optical flow sensor that detects light intensity changes in real-time, triggering initiatively-activated states to transmit pulse signals to adjacent pixels, allowing for the collection of light intensity and time information, thereby extracting optical flow parameters through hardware, reducing parallel computation and enhancing processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional frame-based optical flow calculation is used to process all pixel information, then complete image data is collected, but processing speed decreases and time delay increases for high-speed objects
Solution Approach 1:
The patent extracts only the necessary pixel information (light intensity changes meeting predetermined conditions) rather than processing all pixel data. The pixel collection circuit selectively outputs electric signals only when light intensity changes exceed a threshold, eliminating redundant data processing while maintaining sufficient information for optical flow calculation.
Solution Approach 2:
The patent implements partial action by collecting light intensity information only from pixels that exhibit significant changes (meeting predetermined conditions). This partial sampling approach reduces the volume of data requiring parallel computation while still capturing essential motion information for high-speed object analysis.
2Loss of information
If all pixel information is processed in parallel to acquire gradient values and light intensity changes, then complete optical flow parameters are obtained, but the quantity of parallel computation increases processing time
Solution Approach 1:
The patent performs preliminary filtering at the pixel level by comparing light intensity changes against predetermined conditions before data leaves the sensor array. This preliminary action eliminates obviously redundant data early in the processing chain, reducing the burden on subsequent parallel computation stages while ensuring that meaningful motion information is preserved.
Solution Approach 2:
The patent applies different processing qualities to different spatial locations dynamically. Pixels experiencing significant light intensity changes (indicating motion) are processed with full detail, while pixels in stationary regions are filtered out. This local differentiation optimizes the balance between information completeness and computation time.
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
Enables rapid identification and processing of high-speed objects by collecting optical flow parameters, improving analysis speed and accuracy, and reducing time delays in applications like collision avoidance systems and unmanned vehicles.
Implementation Method 1
a light intensity detector configured to output in real time a first electric signal representing an intensity of a received optical signal
Data Source
Figure 1
Figure 2
Figure 3A~3C
AI summary
The present disclosure provides a pixel collection circuit and an optical flow sensor including the pixel collection circuit. The pixel collection circuit at least includes a light intensity detector, a first state storage module, a second state storage module, a light intensity signal collection and storage module, and a time information storage module.