Lane Line Recognition via Image Segmentation and Symmetrical Convolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane line detection technologies have low accuracy and efficiency due to the inability to distinguish between left and right lane lines, requiring extensive calculations and being prone to false detection from pedestrians and vehicles, leading to slow calculation speeds and high hardware requirements.
Innovation Solution
A method that divides a road image into left and right sub-images, using symmetrical convolution filter kernels for feature extraction and Hough space transformation to accurately detect and track individual lane lines, reducing computational load and improving detection speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire lane image frame is processed for lane line detection, then comprehensive detection coverage is achieved, but the calculation amount increases significantly leading to slow calculation speed
Solution Approach 1:
The patent divides the lane image frame into multiple sub-regions (first sub-image and second sub-image) corresponding to different lane lines. Each sub-region is processed independently to detect specific lane lines, reducing the calculation amount per region while maintaining comprehensive detection coverage. This segmentation approach directly resolves the contradiction by partitioning the large-scale detection problem into smaller, more manageable sub-problems.
2Device complexity
If lane line detection is performed without distinguishing left and right lane lines, then the detection process is simpler, but the detection accuracy is low
Solution Approach 1:
The patent segments the detection process by creating separate processing paths for left and right lane lines. The first sub-image processing detects the first lane line while the second sub-image processing detects the second lane line. This segmentation enables accurate distinction between different lane lines while maintaining manageable process complexity through systematic organization of the detection workflow.
Solution Approach 2:
The patent applies different processing characteristics to different regions of the image. By dedicating specific sub-images to detect specific lane lines (first sub-image for first lane line, second sub-image for second lane line), the system optimizes detection quality for each local region, thereby improving overall detection accuracy while keeping the local processing complexity low.
3Ease of operation
If the original image is transformed into a bird's-eye view to make lane lines parallel and vertical, then lane line extraction becomes more convenient, but the calculation amount increases and false detection from pedestrians and vehicles occurs
Solution Approach 1:
The patent extracts and removes the harmful elements (pedestrians and vehicles) from the detection process by avoiding the bird's-eye view transformation that causes their distortion. Instead, the system processes the original image directly, extracting only the relevant lane line information from appropriately segmented sub-regions. This extraction approach maintains ease of lane line extraction while eliminating the harmful false detection caused by distorted pedestrian and vehicle images.
4Measurement precision
If the entire image frame is processed for lane line detection, then all lane lines can be detected, but the hardware requirements become high due to large calculation amount
Solution Approach 1:
The patent segments the image processing task into independent sub-tasks for different lane lines. By processing the first sub-image and second sub-image separately to detect their respective lane lines, the system reduces the computational load on hardware while maintaining comprehensive detection coverage. This segmentation enables the use of simpler hardware with lower computational requirements.
Data Source
AI summary
A lane line recognition method, a lane line recognition device and a non-volatile storage medium. The method includes obtaining a first image which is a road image; dividing the first image from a middle of the first image to determine a first sub-image at a left part of the first image and a second sub-image at a right part of the first image; and performing respectively a recognition operation on the first sub-image and the second sub-image to determine a first lane line in the first sub-image and a second lane line in the second sub-image.