V-disparity Road Segmentation for Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle detection systems are inefficient in rapidly and accurately detecting objects on a road surface, particularly under complex backgrounds with varying lighting conditions, leading to potential safety issues during driving.
Innovation Solution
A method and system utilizing a V-disparity algorithm for binocular vision, which simplifies the division of road surfaces and obstacles in a 3D disparity map, employing Hough transforms and point clustering to enhance detection speed and accuracy, and is not sensitive to lighting or shadows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If grouping process is carried out on the whole disparity map to detect objects, then detection accuracy is improved, but processing speed decreases to 2 frames per second
Solution Approach 1:
The patent divides the disparity map processing into two stages: first, road surface regions are identified and segmented using the oblique line in V-disparity image; second, object detection is performed only on the remaining non-road regions. This segmentation reduces the processing area from the whole disparity map to only relevant regions, improving processing speed while maintaining detection accuracy.
Solution Approach 2:
The patent extracts and removes the road surface information from the disparity map by detecting the oblique line in the V-disparity image. By taking out the road surface points that would interfere with object detection, the system can focus computational resources on detecting objects only in the remaining regions, thus improving both speed and accuracy.
2Productivity
If brightness-based clustering is used to detect moving objects, then processing speed is improved, but detection reliability deteriorates when object brightness matches surrounding environment
Solution Approach 1:
The patent transitions from brightness-based detection (2D image intensity) to disparity-based detection (3D depth information). By using the disparity map and V-disparity image which encode depth information, the system can detect objects based on their spatial position relative to the road surface rather than their brightness, making detection reliable even when object brightness matches the surrounding environment.
3Adaptability or versatility
If conventional vehicle detection systems process complex backgrounds with varying lighting, then comprehensive detection coverage is achieved, but detection accuracy decreases due to sensitivity to lighting and shadows
Solution Approach 1:
The patent replaces brightness-based detection mechanisms with disparity-based detection mechanisms. By substituting the reliance on optical intensity (brightness) with reliance on geometric depth information from stereoscopic vision, the system becomes insensitive to lighting variations and shadows, maintaining high detection accuracy across diverse lighting conditions while preserving comprehensive detection coverage.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed are a method and a system for detecting an object on a road. The method comprises a step of simultaneously capturing two depth maps, and then calculating a disparity map; a step of obtaining, based on the disparity map, a V-disparity image by adopting a V-disparity algorithm; a step of detecting an oblique line in the V-disparity image, and then removing points in the disparity map, corresponding to the oblique line so as to acquire a sub-disparity map excluding the road; a step of detecting plural vertical lines in the V-disparity image, and then extracting, for each of the plural vertical lines, points corresponding to this vertical line from the sub-disparity map as an object sub-disparity map corresponding to this vertical line; and a step of merging any two rectangular areas of the object sub-disparity maps approaching each other, into a rectangular object area.