Road Surface Recognition Using Local Intensity Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane keeping assistance systems (LKAS) in advanced driver assistance systems (ADAS) struggle to accurately detect road surfaces and markings, leading to potential accidents when drivers take their hands off the steering wheel for non-driving tasks.
Innovation Solution
A method and device for road surface recognition using a vision sensor to generate an original intensity map, local intensity map, and normalized intensity map, enabling accurate detection of road surfaces and markings by normalizing signal intensities based on local neighborhoods and thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional road surface detection methods are used, then the system is simple to implement, but the detection accuracy deteriorates when external factors or driver distractions occur
Solution Approach 1:
The patent divides the road surface detection process into multiple processing stages: obtaining original intensity map, generating local intensity map through neighborhood analysis, and producing normalized intensity map. This segmentation allows each stage to focus on specific aspects of detection, improving overall accuracy while maintaining manageable system complexity through modular processing.
Solution Approach 2:
The patent transitions from analyzing raw sensor data in one dimension to creating intensity maps that represent road surface properties in an additional dimension. By mapping sensor readings to intensity values and processing them through local neighborhood analysis, the system gains a new dimensional perspective on road surface characteristics, enhancing detection capability.
2Measurement precision
If local neighborhood analysis is performed for each element, then the normalization accuracy is improved, but the processing time increases
Solution Approach 1:
The patent performs preliminary organization of data into intensity maps before normalization. By pre-processing the sensor data into structured intensity representations and pre-calculating local neighborhood relationships, the system prepares the data in advance, making the subsequent normalization process more efficient and reducing overall processing time.
Solution Approach 2:
The patent applies local quality by performing neighborhood analysis specifically on elements within the intensity map. Instead of uniformly processing all data with the same complexity, the system focuses computational resources on local regions, calculating local intensity values based on neighboring elements. This localized approach improves normalization accuracy where needed while avoiding unnecessary processing elsewhere.
Data Source
AI summary
A method performed by one or more processors includes: obtaining an original intensity map including original elements having respective signal intensities with respect to a road surface, the signal intensities based on sensing of a vision sensor arranged on a moving object that is moving on a road including the road surface, wherein each of the original elements has a respective local neighborhood of neighboring original elements; generating a local intensity map by determining local values respectively corresponding to the original elements of the original intensity map, wherein the local values are determined based on the original elements in the respectively corresponding local neighborhoods; generating a normalized intensity map by normalizing the original intensity map based on the local intensity map; and recognizing the road surface based on the normalized intensity map.