Distant Marking Line Detection via Segmented Road Surface Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing marking line detection systems face challenges in accurately detecting distant marking lines due to issues like shadow interference, line hiding by vehicles or bicycles, and wavy lines, leading to incorrect area analysis and reduced detection accuracy.
Innovation Solution
A marking line detection system that includes an imaging device, an information extraction unit, a road surface area identification unit, and a marking line detection unit, which extracts depth distance information and identifies distant road surface areas based on vehicle speed, allowing separate detection of marking lines without estimating the road surface shape from the immediately-preceding area to the distant area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the white line detection area is set as large as possible to detect distant marking lines, then the detection coverage is improved, but the processing speed decreases and the number of incorrect detections increases
Solution Approach 1:
The patent divides the road surface into multiple detection areas (first detection area near the vehicle and second detection area distant from the vehicle) with different measuring point densities. This segmentation allows the system to process distant areas with lower density, improving processing speed while maintaining detection coverage.
Solution Approach 2:
The patent applies different measuring point densities to different regions: higher density in the near area and lower density in the distant area. This local quality approach optimizes processing efficiency by allocating computational resources according to the actual detection needs of each region.
2Measurement precision
If the measuring point density in the distant area is increased to improve detection accuracy, then the detection precision is improved, but the computational load and processing time increase
Solution Approach 1:
The patent segments the road surface into near and distant areas with different measuring point densities, allowing the distant area to be processed with lower density to reduce computation time while maintaining adequate detection accuracy.
Solution Approach 2:
The patent changes the parameter of measuring point density based on distance from the vehicle, using lower density for distant areas to reduce computational load while maintaining detection effectiveness.
3Measurement precision
If the road surface shape is estimated sequentially from near to distant area to improve detection accuracy, then the detection precision is improved, but the system fails when shadows or hidden lines prevent correct estimation
Solution Approach 1:
The patent performs preliminary detection of marking lines in the distant area before using them to estimate road surface shape. This preliminary action provides initial detection results that can be used even when sequential estimation from near to distant fails due to shadows or hidden lines.
Solution Approach 2:
The patent inverts the conventional sequential estimation approach by first detecting distant marking lines and then using them for road surface estimation, rather than estimating from near to distant. This inversion allows detection to proceed even when near-area information is unavailable or incorrect.
4Productivity
If the white line detection area is reduced to increase processing speed, then the processing efficiency is improved, but the detection coverage and ability to detect distant lines decrease
Solution Approach 1:
The patent segments the detection area into near and distant regions with different measuring point densities, maintaining comprehensive detection coverage while improving processing efficiency through optimized resource allocation in each segment.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A marking line detection system (1) includes an imaging device (10), an information extraction unit, a road surface area identification unit and a marking line detection unit. The information extraction unit is configured to extract depth distance information from the imaging device to an imaging object based on image information in an imaging area captured by the imaging device. The road surface area identification unit is configured to identify a distant road surface area based on the depth distance information, the distant road surface area being a road surface area that excludes an immediately-preceding road surface area of the vehicle in the image information and is more distant from the vehicle than the immediately-preceding road surface area. The marking line detection unit is configured to detect a marking line in the distant road surface area based on image information corresponding to a position of the distant road surface area.