Lane Keeping Assist Using Reliability-Based Lane Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Lane keeping assist systems (LKAS) face performance variability due to weather, illumination, and road conditions, leading to inaccurate lane detection and potential loss of lane keeping function, especially at intersections or road construction sites.
Innovation Solution
A lane keeping assist system that includes a camera, a lane information generator, an image storage, and a neural network learning device, which generates and selects reliable lane information using image reliability assessment, and a steering controller that adjusts based on the reliability of the image information, employing convolution neural networks (CNNs) and long short-term memory (LSTM) for predicting lane information and generating steering commands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the LKAS uses camera-based lane detection, then the system can provide active intervention for lane keeping, but the detection performance varies with weather, illumination, and road conditions
Solution Approach 1:
The system stores image information for each predetermined time in advance before it is needed for lane detection. This preliminary storage of historical image data allows the neural network to reference multiple time points when current image quality is poor, ensuring continuous reliable lane detection regardless of temporary adverse conditions
Solution Approach 2:
The neural network learning device acts as an intermediary between the camera input and the lane keeping control. It processes current and historical image information to generate reliable lane information, filtering out the effects of poor weather, illumination, or road conditions before the data reaches the control system
2Speed
If the LKAS relies on current image information for lane detection, then the system responds quickly to lane changes, but the lane keeping function may be released when lane information is temporarily inaccurately recognized
Solution Approach 1:
Image information is stored for each predetermined time in advance, creating a temporal buffer of historical data. When current lane detection is inaccurate, the system can immediately reference previously stored images to maintain lane keeping function without interruption or release
Solution Approach 2:
The system uses feedback from both current and historical image information through the neural network to continuously assess lane position. This feedback mechanism allows the system to distinguish between temporary detection errors and actual lane changes, maintaining stability while responding to genuine lane transitions
3Reliability
If the system activates neural network learning for all image conditions, then accurate lane information can be generated from unreliable images, but system complexity and processing load increase
Solution Approach 1:
The neural network learning device is dynamically activated or deactivated based on image reliability assessment. When images are determined to be unreliable, the neural network is activated to process historical image data; when images are reliable, the system uses current images directly, reducing processing complexity while maintaining accuracy
Solution Approach 2:
The system changes the operational parameters of the neural network based on image quality conditions. The activation state of the neural network learning device is adjusted according to the reliability of current image information, optimizing the balance between processing accuracy and system complexity
Data Source
AI summary
A lane keeping assist system includes: a camera configured to provide an image around a vehicle as image information; a lane information generator configured to generate image reliability information and first lane information, based on the image information; an image storage configured to store the image information for each predetermined time among predetermined times; a neural network learning device configured to generate second lane information based on the image reliability information and the image information stored for each predetermined time; and a steering controller configured to select either one of the first lane information and the second lane information as lane information, based on the image reliability information, and generate steering information based on the selected lane information.