Monocular Road Surface Detection via Normalized Speed Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road surface detection devices require complex configurations, such as binocular or stereo cameras, to accurately detect road surfaces, which complicates their installation and operation in vehicles.
Innovation Solution
A road surface area detection device that uses a single wide-angle fisheye camera to calculate normalized speeds of feature points across the vehicle's width, determining a road surface area by identifying positions with normalized speeds within a predetermined range, allowing for road surface detection without the need for multiple cameras.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a binocular camera or stereo camera is used for road surface detection, then detection accuracy is improved, but device complexity increases
Solution Approach 1:
The patent uses a single camera to capture images and creates a virtual stereoscopic effect through image processing. By generating a V-disparity map from monocular images and extracting line segments, the system replicates the road surface detection capability of stereo cameras without requiring multiple physical cameras, thus reducing device complexity while maintaining detection accuracy
Solution Approach 2:
The patent replaces the mechanical/optical system of multiple cameras with a computational approach using a single camera. Instead of using physical stereo vision hardware, the system uses image processing algorithms to generate depth information and detect road surfaces from monocular images, substituting mechanical complexity with computational processing
2Reliability
If multiple cameras are installed in the vehicle, then road surface detection capability is improved, but installation complexity increases
Solution Approach 1:
The system replicates the functionality of multiple cameras using a single camera through computational methods. By processing images from one camera to generate V-disparity maps and extract road surface line segments, the system achieves reliable road surface detection without the installation complexity of mounting and calibrating multiple cameras in the vehicle
3Device complexity
If a single camera is used for road surface detection, then device complexity is reduced, but detection accuracy deteriorates
Solution Approach 1:
The patent transforms the detection parameters by converting monocular image data into a V-disparity map representation. This parameter transformation allows the system to extract depth information and road surface characteristics from a single camera, maintaining detection accuracy while using a simpler device configuration
Solution Approach 2:
The system substitutes the optical-mechanical depth sensing capability of multiple cameras with computational image processing. By using algorithms to generate V-disparity maps and extract line segments from single-camera images, the patent replaces hardware complexity with software-based measurement techniques that maintain precision
Data Source
AI summary
A road surface area detection device includes a normalized speed calculation portion configured to calculate a normalized speed based on a movement of a feature point in an image captured by a camera that is disposed in a vehicle; a determination range calculation portion configured to calculate a road surface determination range, which is indicated by a magnitude of the normalized speed, based on the normalized speeds of at least two feature points at different positions in a width direction of the vehicle in a predetermined central area where the vehicle is positioned in a center thereof in the width direction perpendicular to a vehicle traveling direction; and a road surface area identification portion configured to identify, as a road surface area on which the vehicle travels, a position in the width direction that includes the feature point whose normalized speed is within the road surface determination range.