Stereo Range Image Processing for Fast Road And Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques using stereo cameras for detecting objects, such as pedestrians and vehicles, are slow due to the time required for processing range images, which can hinder timely collision avoidance in automotive systems.
Innovation Solution
The implementation of a moving object device control system that includes an image processing apparatus capable of fast image processing by generating frequency distribution images of parallax values, allowing for real-time detection of road surfaces and objects through efficient data extraction and interpolation methods, enabling rapid identification of objects and road profiles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image processing techniques are used to generate V-disparity and U-disparity images for object detection, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent divides the parallax image into multiple extraction regions (first extraction region, second extraction region, third extraction region) and processes each region separately to generate corresponding V-disparity images and U-disparity images. This segmentation allows parallel processing and reduces overall processing time while maintaining detection precision for different spatial zones
Solution Approach 2:
The patent performs preliminary extraction of parallax pixel data from the original parallax image before generating disparity images. By pre-processing and organizing the pixel data into structured formats with predetermined coordinate systems, the system reduces the computational burden during the actual object detection phase, thereby decreasing processing time without compromising measurement precision
2Productivity
If fast image processing is implemented by extracting only necessary pixel data, then processing speed is improved, but measurement precision may deteriorate
Solution Approach 1:
The patent applies different extraction rates to different extraction regions based on their specific requirements. The first extraction region uses a first extraction rate, the second extraction region uses a second extraction rate, and the third extraction region uses a third extraction rate. This local quality approach ensures that critical regions are processed with higher precision while less critical regions use faster processing, maintaining overall measurement precision while improving processing speed
Solution Approach 2:
The patent extracts parallax pixel data at different rates for different regions, using partial extraction (lower rates) for some regions and more comprehensive extraction (higher rates) for others. This selective approach processes only the necessary amount of data for each region's detection needs, achieving fast processing speed while preserving measurement precision where required
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image processing apparatus includes an extracting unit configured to obtain a range image having pixel values corresponding to a distance of an object in a plurality of captured images obtained by capturing the object on a road surface so as to extract pixel values from the range image, a first generating unit configured to generate vertical direction distribution data indicating a frequency distribution of distance values in a vertical direction of the range image based on the extracted pixel values, a calculating unit configured to calculate a height of the road surface in the range image before generation of the vertical direction distribution data is completed, and a second generating unit configured to generate horizontal direction distribution data indicating a frequency distribution of distance values in a horizontal direction of the range image based on the extracted pixel values and the calculated height of the road surface.