LiDAR-Triggered Camera Obstacle Detection for Self-Driving Farm Machines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing agricultural machines face challenges in accurately detecting obstacles while minimizing the processing load for self-driving operations.
Innovation Solution
An obstacle detection system for agricultural machines that utilizes a LiDAR sensor and a camera, where the camera is triggered by the LiDAR to capture images of potential obstacles, determining the need for a change in travel status based on these images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If obstacle detection is performed continuously using both LiDAR and camera, then obstacle detection accuracy is improved, but processing load increases
Solution Approach 1:
The system performs obstacle detection periodically by triggering the camera only when LiDAR detects an obstacle candidate, rather than continuous operation. The controller activates the camera upon receiving obstacle candidate information from the LiDAR sensor, creating a periodic detection rhythm that reduces processing load while maintaining detection accuracy when needed.
Solution Approach 2:
The LiDAR sensor performs preliminary scanning to identify obstacle candidates before the camera captures detailed images. This preliminary detection by LiDAR filters out non-obstacle areas, so the camera only processes regions where obstacles may exist, significantly reducing the camera's processing load while maintaining high detection accuracy.
2Measurement precision
If camera continuously captures images for obstacle detection, then obstacle detection accuracy is improved, but energy consumption increases
Solution Approach 1:
The camera operates periodically rather than continuously, activating only when LiDAR detects an obstacle candidate. This periodic operation dramatically reduces energy consumption while maintaining detection accuracy, as the camera remains inactive during periods when no obstacles are present.
Solution Approach 2:
The LiDAR sensor acts as an intermediary that filters and directs the camera's operation. Instead of the camera continuously scanning all areas, the LiDAR identifies potential obstacle regions and triggers the camera only for those specific areas, reducing overall energy consumption while maintaining detection accuracy.
3Measurement precision
If LiDAR and camera both operate at full capacity, then obstacle detection coverage is improved, but system complexity increases
Solution Approach 1:
The detection system is segmented into two functional parts: LiDAR for preliminary obstacle candidate identification and camera for detailed verification. This segmentation allows each component to operate at optimized capacity levels, reducing overall system complexity while maintaining comprehensive detection coverage.
Solution Approach 2:
The LiDAR performs preliminary detection to identify obstacle candidates, which then triggers the camera for detailed imaging. This preliminary action by LiDAR reduces the camera's workload and simplifies the overall system operation, as the camera only needs to process pre-identified candidate regions rather than performing full-scene analysis.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively detects obstacles with high accuracy while reducing the processing load, enabling precise self-driving operations.
Implementation Method 1
sensing a surrounding environment with a LiDAR sensor
Implementation Method 2
LiDAR sensor
Implementation Method 3
acquire an image of the obstacle candidate
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An obstacle detection system for an agricultural machine to perform self-driving while sensing a surrounding environment with a LiDAR sensor and a camera includes a controller causing the camera upon detecting an obstacle candidate based on data that is output from the LiDAR sensor, as a trigger, to acquire an image of the obstacle candidate, and determining whether or not to change a traveling status of the agricultural machine based on the image of the obstacle candidate acquired with the camera.