CLAHE-Enhanced Lane Detection for Small Markings in Poor Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane detection systems in advanced driver assistance systems (ADAS) and automatic driving systems face significant challenges under complex imaging conditions such as bad weather and complicated lighting, leading to poor image quality and small lane marking targets, which degrade detection performance.
Innovation Solution
A lane detection method integrating image enhancement using the CLAHE algorithm and a deep convolutional neural network, comprising an input module, spatial attention module, feature extraction module, and detection module, to enhance contrast and accurately identify lane marking blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional edge-detection based methods or deep-learning based methods are used under bad weather and complicated imaging conditions, then the detection system can operate, but the detection performance degrades due to poor image quality
Solution Approach 1:
The CLAHE algorithm is applied to enhance image contrast before the deep convolutional neural network performs lane detection. This preliminary image enhancement action prepares the input data by improving visibility of lane markings in challenging conditions, allowing the detection system to maintain high performance even when captured images have poor quality due to rain, fog, backlight, or low illumination
Solution Approach 2:
The patent introduces an intermediate image enhancement step using CLAHE as a mediator between the raw captured image and the deep learning detection model. This intermediary processing transforms the degraded input image into an enhanced version with improved contrast, bridging the gap between poor image quality and the requirements of the detection algorithm
2Measurement precision
If the input image contains small lane marking targets under complicated imaging conditions, then the detection task becomes more difficult, but accurate detection is required for reliable lane identification
Solution Approach 1:
The CLAHE algorithm applies contrast enhancement locally to different regions of the image, with each block processed independently to enhance the visibility of lane markings in that specific area. This local processing approach adapts to varying lighting and weather conditions across different parts of the road scene, improving the detection of small lane marking targets in challenging conditions
Solution Approach 2:
The deep convolutional neural network is pre-trained and configured with spatial attention mechanisms to focus on small lane marking targets before the actual detection task. The network architecture includes preprocessing steps that prepare the enhanced images to highlight small features, making the subsequent detection of small lane markings more accurate and less difficult
Data Source
AI summary
A lane detection method integratedly using image enhancement and a deep convolutional neural network. On the assumption that lanes have similar widths in a local region of an image and a lane can be segmented into several image blocks, each of which contains lane marking in the center, a method based on a deep convolutional neural network is provided to detect lane marking blocks in the image. Input to the model includes road images captured by a camera as well as a set of enhanced images generated by the contrast limited adaptive histogram equalization (CLAHE) algorithm. The method according to the present disclosure can effectively overcome difficulties of lane detection under complex imaging conditions, such as poor image quality, and small lane marking targets, so as to achieve better robustness.