Road Defect Detection Using Sensor-Image Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for detecting road defects rely on sensors that react to the impact of defects, which is after the vehicle has already hit them, and vision systems require manual labeling and lack precise location information.
Innovation Solution
A system that correlates motion sensor data with time sequential images using a computer vision algorithm and machine learning models to predict road defects before impact, utilizing local and remote databases for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion sensors are used to detect road defects, then the system can detect defects when the vehicle hits them, but the detection occurs after the vehicle has already encountered the defect
Solution Approach 1:
The system performs preliminary action by capturing images of the roadway before the vehicle reaches the defect location. The image capture system continuously monitors the roadway ahead, and when a defect is detected through motion sensor correlation, the corresponding pre-captured image is already available for immediate analysis, eliminating detection delay
Solution Approach 2:
The patent introduces an intermediary correlation system that links motion sensor data with image data. The correlation module acts as a mediator that matches acceleration/impact sensor signals with specific image frames, enabling the system to identify defects in images captured before the vehicle actually hits them
2Loss of time
If vision systems with cameras are used to see road defects before impact, then early detection is possible, but manual labeling is required and precise location information is lacking
Solution Approach 1:
The system implements self-service by using motion sensor data to automatically identify and label road defects without manual intervention. The correlation between sensor-detected defects and image features enables automated defect identification, eliminating the need for manual labeling while maintaining accurate defect classification
Solution Approach 2:
The patent employs feedback mechanisms where motion sensor data provides verification and refinement for vision system detections. The sensor data feedback loop confirms the presence and location of defects, improving the accuracy of automated identification and reducing false positives in the vision system
3Loss of information
If global positioning is used to determine road defect location, then defect location can be recorded, but accurate precise information on lateral location within the roadway is not provided
Solution Approach 1:
The system transitions from one-dimensional GPS coordinates to two-dimensional precise location mapping by incorporating camera image coordinates. The defect location is determined by mapping the defect's position within the image frame to real-world coordinates, adding lateral precision dimension that GPS alone cannot provide
Solution Approach 2:
The patent segments the location information into multiple components: longitudinal position from GPS, lateral position from image coordinates, and vertical depth from stereo vision or focus analysis. This segmentation of location data across different spatial dimensions enables precise three-dimensional defect localization
Data Source
AI summary
A system for identifying road defects within a roadway ahead of a moving vehicle, includes at least one camera in communication with a system controller and adapted to capture a plurality of time sequential images of the roadway ahead of the moving vehicle, and at least one motion sensor in communication with the system controller and adapted to detect when the vehicle encounters a road defect, the system controller adapted to correlate the road defect with at least one of the plurality of time sequential images of the roadway, analyze, using a computer vision algorithm, the at least one of the plurality of time sequential images, identify, with the computer vision algorithm, the road defect within the at least one of the plurality of time sequential images, and label the identified road defect within the at least one of the plurality of time sequential images.