Single-Camera Dual Video Streams for ADAS Vision Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current camera systems for combined computer vision and human vision in Advanced Driver-Assistance Systems (ADAS) suffer from low resolution and poor color accuracy, with post-processing techniques like sharpening algorithms introducing artifacts that are detrimental to computer vision applications.
Innovation Solution
A video processing system that generates separate computer vision and human vision video streams using a single camera, where the computer vision stream maintains a wide aspect ratio field of view and the human vision stream is cropped for higher resolution with a narrower field of view, optimized for each respective application.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If post-processing techniques like sharpening algorithms and contrast enhancement are applied to improve human vision quality, then color accuracy and visual appeal are improved, but artifacts are introduced that render frames useless to computer vision
Solution Approach 1:
The patent divides the video stream processing into two separate processing paths: one for computer vision and one for human vision. The single video stream from the camera is segmented into two distinct output streams, each optimized for its specific application without interference from the other's processing requirements.
Solution Approach 2:
The patent extracts only the necessary portions of the video stream for each application. For human vision, it extracts and enhances only the relevant visual qualities (color, contrast) without applying computer vision processing. For computer vision, it extracts only the raw or minimally processed data needed for analysis, excluding all post-processing that would introduce artifacts.
2Device complexity
If a single camera is used for both computer vision and human vision, then device complexity is reduced, but the resolution and field of view cannot be optimized for both applications simultaneously
Solution Approach 1:
The patent makes the single camera system universal by enabling it to serve two different functions simultaneously. The same camera hardware produces video streams that are then processed differently to fulfill both computer vision analysis requirements and human vision display requirements, eliminating the need for separate camera systems.
Solution Approach 2:
The patent changes the processing parameters of the video stream based on the intended application. The same input stream is processed with different parameters: one path maintains wide field of view and original resolution for computer vision, while the other path crops and enhances for human vision display, thereby optimizing resolution for each application from a single camera source.
3Manufacturing precision
If the video stream is cropped for human vision to achieve higher resolution, then color accuracy is improved, but the wide aspect ratio field of view necessary for computer vision is lost
Solution Approach 1:
The patent segments the single video stream into two separate output streams with different characteristics. One stream maintains the original wide aspect ratio and full field of view for computer vision applications, while the other stream is cropped and enhanced for human vision display, allowing both requirements to be met simultaneously from a single camera input.
Data Source
AI summary
Embodiments include providing a computer vision video stream and a human vision video stream using a single camera. The camera can comprise a single, high-resolution senor and a wide-angle lens providing high resolution video having a wide aspect ratio field of view. The streams can be generated, wherein the computer vision video stream maintains the wide aspect ratio field of view and the human vision video stream comprises high resolution video. The computer vision video stream can be provided to a computer vision system, wherein the computer vision system uses the computer vision video stream as input for an automated process, and the human vision video stream can be provided to a system that displays the human vision video stream to a user.