Working Vehicle Obstacle Detection by Direction-Based Camera Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge is to detect obstacles with high accuracy and in a short amount of time using captured images of a vehicle's surroundings, given the limitations of processing power in onboard processors.
Innovation Solution
A working vehicle equipped with multiple image capture devices, a direction detection unit, an image selection unit, and a detection unit that selectively uses the surrounding area image captured by the image capture device facing the front advancement direction as the detection image for obstacle analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all surrounding area images from multiple cameras are used for obstacle detection, then detection coverage and reliability are improved, but data processing time and computational load increase excessively
Solution Approach 1:
The patent extracts only the necessary detection image from the multiple surrounding area images captured by different cameras. The image selection unit identifies and selects the most relevant image based on the advancement direction, extracting only the essential data needed for obstacle detection while discarding redundant images, thus reducing processing time while maintaining detection reliability
Solution Approach 2:
The patent segments the obstacle detection task into two distinct stages: first, the direction detection unit determines the advancement direction to identify which camera's image is most relevant; second, the detection unit analyzes only the selected detection image. This segmentation allows the system to process minimal data while maintaining comprehensive detection coverage
2Measurement precision
If a processor with higher processing power is installed to analyze large amounts of captured image data with high accuracy, then obstacle detection accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts only the essential detection image needed for accurate obstacle detection, rather than processing all images from multiple cameras. This extraction approach maintains high detection accuracy by focusing computational resources on the most relevant data, thereby avoiding the need for excessively powerful and complex processors
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of the captured images (the detection image in the advancement direction) rather than all surrounding area images. This partial processing achieves sufficient detection accuracy without requiring the full processing power that would be needed to analyze all images comprehensively
3Area of stationary object
If multiple surrounding area images are combined to generate a comprehensive surroundings image, then detection coverage is improved, but data amount and processing complexity increase
Solution Approach 1:
The patent extracts only the detection image corresponding to the advancement direction from the multiple surrounding area images, rather than combining all images into a comprehensive surroundings image. This extraction maintains adequate detection coverage for obstacle avoidance while significantly reducing the complexity of image processing and data management
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
Provided is a working vehicle capable of operation and driving that includes a plurality of image capture devices 4 that capture an image of surroundings of a vehicle body; a direction detection unit 51 that detects an advancement direction of the operation and driving; an image selection unit 52 that acquires, as a detection image, a surrounding area image captured by one image capture device 4 of the plurality of image capture devices 4 capturing a front advancement direction detected by the direction detection unit 51; and a detection unit 53 that analyzes the detection image acquired by the image selection unit 52 and detects an obstacle.