Road Surface Detection Using Vertical Distribution Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques for stereo cameras struggle to accurately measure road surfaces in rainy weather due to noise generated by reflection, leading to inaccurate detection and potential obstacles being misidentified.
Innovation Solution
An image processing apparatus that generates vertical direction distribution data from range images captured by multiple imaging parts, estimates multiple road surfaces, and determines the desired road surface by analyzing parallax image data to improve road surface detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereo camera image processing is used to detect road surfaces, then distance measurement capability is improved, but measurement precision deteriorates in rainy weather due to reflection noise
Solution Approach 1:
The patent segments the vertical direction distribution data into multiple peaks, where each peak corresponds to a different road surface. This segmentation allows the system to identify and separate multiple road surfaces (e.g., actual road surface and reflected road surface) from the distribution data, enabling accurate detection even in rainy conditions with reflection noise.
Solution Approach 2:
The patent introduces vertical direction distribution data as an intermediary representation between the raw stereo camera images and the final road surface detection. This distribution data serves as a mediator that transforms the complex image processing problem into a more manageable form where multiple road surfaces can be estimated and distinguished through peak analysis.
2Productivity
If traditional noise removal techniques are applied to parallax images, then image processing speed is improved, but measurement precision deteriorates due to loss of valid road surface data
Solution Approach 1:
Instead of removing noise through traditional filtering that may eliminate valid data, the patent segments the vertical direction distribution data into multiple peaks. Each peak represents a distinct road surface, allowing the system to preserve all valid road surface data while separating it from reflection noise through peak identification and selection.
Solution Approach 2:
The patent changes the parameter representation from direct parallax image values to vertical direction distribution characteristics (peak positions and shapes). This parameter transformation allows the system to distinguish between actual road surfaces and reflections based on the distribution pattern, maintaining measurement precision without requiring aggressive noise removal.
Data Source
AI summary
An image processing apparatus includes one or more processors; and a memory, the memory storing instructions, which when executed by the one or more processors, cause the one or more processors to generate vertical direction distribution data indicating a frequency distribution of distance values with respect to a vertical direction of a range image, from the range image having distance values according to distance of a road surface in a plurality of captured images captured by a plurality of imaging parts; estimate a plurality of road surfaces, based on the vertical direction distribution data; and determine a desired road surface, based on the estimated plurality of road surfaces.