Geometric Intersection Lane Restoration From Front-Camera Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately determine the cause of traffic accidents by clearly distinguishing between the offender and victim, and fail to accurately represent the path between vehicles at intersections, especially when lane markings are unclear.
Innovation Solution
A system and method using geometric modeling to recognize and restore intersection lanes by analyzing sequential images from a front camera, detecting driving lanes and intersections, and expressing them in a top view using a geometric structure defined by lower and upper end points and angles, employing variable top-hat filters and Sobel edge operators for lane detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sequential image analysis is used to restore vehicle paths, then the cause of traffic accidents can be determined, but it is difficult to accurately understand the path between vehicles at intersections when lane markings are unclear
Solution Approach 1:
The system performs preliminary detection of intersection lanes using geometric modeling before accident analysis. By pre-identifying intersection locations and lane configurations through geometric patterns in sequential images, the system prepares path restoration data in advance, enabling accurate vehicle path reconstruction even when lane markings are unclear during the accident event.
2Measurement precision
If geometric modeling is used to recognize intersection lanes, then intersection paths can be accurately represented, but the system complexity increases
Solution Approach 1:
The geometric modeling process is segmented into distinct operational stages: intersection detection phase, lane configuration analysis phase, and path restoration phase. Each segment handles specific computational tasks independently, reducing overall system complexity while maintaining high recognition accuracy through specialized processing at each stage.
Data Source
AI summary
Disclosed are an intersection recognition and restoration system using geometric modeling method and intersection recognition and restoration method using the same. The intersection recognition and restoration system using geometric modeling of the present invention for solving the above problems comprises an image capturing unit that sequentially acquires images from a front camera installed in a subject vehicle; an image analysis unit that detects a driving lane from a sequential image of the video file acquired by the image capturing unit, and recognizes and restores an intersection by a geometric modeling; a memory for storing driving lane and intersection data detected by the image analysis unit; and a display unit that expresses the driving lane and intersection data stored in the memory in the form of a top view.