3D Obstacle Detection With Road-Surface Height Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing obstacle detection systems face challenges in accurately estimating road-surface height and correcting the three-dimensional position of objects due to variations in the mounting height of imaging apparatus caused by factors like air suspension and load changes in vehicles, leading to inaccurate obstacle detection.
Innovation Solution
An obstacle detection apparatus that includes a three-dimensional estimating unit to generate a three-dimensional estimation image, an image classifying unit to classify objects into road-surface and other classes, a road-surface point extracting unit to fuse and associate feature points with the road-surface class, a road-surface height estimating unit to accurately estimate height, and a correcting unit to adjust object positions based on this height, ensuring precise obstacle detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the imaging apparatus is mounted on a vehicle with air suspension, then the vehicle can adapt to different road conditions, but the mounting height of the imaging apparatus varies temporarily, leading to inaccurate obstacle detection
Solution Approach 1:
The system uses sonar to detect the height of the road surface in real-time and feeds this information back to correct the three-dimensional position of objects detected by the imaging apparatus. This feedback mechanism compensates for the temporary variations in mounting height caused by air suspension, maintaining accurate obstacle detection while preserving the adaptability benefits of air suspension.
Solution Approach 2:
The patent introduces sonar as an intermediary measurement device to indirectly measure the road surface height. This intermediary measurement allows the system to compensate for mounting height variations without directly measuring the imaging apparatus position, thereby resolving the contradiction between adaptability and measurement precision.
2Device complexity
If the mounting height of the imaging apparatus is not corrected, then the system structure remains simple, but the three-dimensional position of objects is inaccurate due to road surface variations
Solution Approach 1:
The patent uses sonar as an intermediary device to measure road surface height, which then serves as a correction parameter for the imaging apparatus. This approach maintains relatively simple system structure while significantly improving three-dimensional position accuracy by compensating for road surface variations through the intermediary sonar measurements.
3Adaptability or versatility
If the imaging apparatus mounting height varies due to vehicle load, then the vehicle can carry different loads flexibly, but the obstacle detection accuracy deteriorates
Solution Approach 1:
The system continuously monitors road surface height using sonar and applies real-time feedback correction to the three-dimensional position calculations. This feedback mechanism enables the system to maintain accurate obstacle detection despite flexible load carrying capabilities that cause mounting height variations.
Solution Approach 2:
The system uses the sonar device already present on the vehicle for other purposes (such as basic distance measurement) and leverages it to also measure road surface height for correction purposes. This self-service approach allows the system to compensate for load-induced height variations without adding dedicated correction hardware.
Data Source
AI summary
In an obstacle detection apparatus, the captured image of a vicinity of a vehicle from an imaging apparatus is acquired. A three-dimensional estimation image showing a three-dimensional position of a feature point in the captured image is generated, and a three-dimensional position of an object is estimated. An attribute image in which an object in the captured image is classified into one or more classes that include at least a road-surface class is generated. The three-dimensional estimation image and the attribute image are fused, the feature points and the classes are associated, and road-surface points associated with the road-surface class are extracted. Based on the road-surface points, a road-surface height in the vicinity of the vehicle is estimated. Based on the estimated road-surface height, the three-dimensional position of the object is corrected. Based on the three-dimensional position of the object, an obstacle in the vicinity of the vehicle is detected.