Stereo Camera Road Surface Detection Beyond 35 m Using Neural Disparity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing obstacle detection systems, such as LiDAR and stereo camera systems, face challenges in accurately determining the road surface and obstacles at distances greater than 35 meters due to laser beam reflection and high signal-to-noise ratios, respectively.
Innovation Solution
A method using a stereo camera system with cameras spaced more than 0.8 m apart, integrated into the vehicle, and an artificial neural network to process image information, determining disparity and distance information, and compensate for calibration inaccuracies, allowing for accurate road surface detection and obstacle identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR systems are used for distance determination, then distance resolution is improved, but road surface detection fails at distances greater than 35m due to total reflection
Solution Approach 1:
The patent introduces an artificial neural network as an intermediary processing layer that takes disparity information from stereo camera systems and converts it into accurate distance information. This neural network mediator enables the stereo camera system to achieve LIDAR-level distance precision without suffering from the total reflection problem at long ranges.
2Measurement precision
If stereo camera systems are used for obstacle detection, then lateral resolution is improved, but signal-to-noise ratio deteriorates making road surface determination inadequate
Solution Approach 1:
The patent replaces traditional mechanical signal processing methods with an artificial neural network that processes image information from stereo cameras. This substitution enables the system to extract meaningful distance information and determine road surface level even in low signal-to-noise conditions, overcoming the limitations of conventional stereo vision systems.
3Device complexity
If traditional stereo camera systems are used, then device complexity is reduced, but distance determination accuracy deteriorates at distances greater than 35m
Solution Approach 1:
The patent changes the processing parameters by introducing an artificial neural network that transforms raw disparity information into accurate distance measurements. This parameter transformation enables the system to maintain high distance determination accuracy at ranges exceeding 35m while keeping the sensor hardware relatively simple.
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 reliable and precise determination of the road surface and obstacles even at distances greater than 35 meters, improving obstacle detectability and enabling timely evasive maneuvers.
Implementation Method 1
stereo camera system for capturing stereo images of the surrounding area of the vehicle
Implementation Method 2
The disparity information specifically indicates the distance between corresponding pixels in the image information from the two cameras. This distance results from the different viewing angles of the scene area represented by this pixel and the resulting parallax.
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method for determining the road surface in the vicinity of a vehicle, wherein the vehicle has a stereo camera system (2) for capturing stereo images of the vehicle's surroundings and an artificial neural network (3) for processing the image information provided by the stereo camera system (2), wherein the neural network (3) determines disparity information of the vehicle's surroundings, wherein distance information is calculated based on the disparity information, which contains information regarding the distance of the objects depicted in the image information to the stereo camera system (2) or the vehicle, wherein road surface distance information is extracted from the distance information, and wherein the road surface is determined based on the road surface distance information.