Depth Obstacle Mapping for UAV Collision Avoidance in Dynamic Scenes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing unmanned aerial vehicles (UAVs) face challenges in efficiently navigating dynamic environments with moving obstacles, particularly in swarm-based inspection tasks, where collisions can occur due to limited sensor coverage and the inability to effectively update obstacle maps in real-time.
Innovation Solution
Implementing a depth imaging system with selective forget functionality for obstacle maps, such as voxel maps, that allows UAVs to generate and update three-dimensional obstacle maps dynamically, incorporating probability values to account for moving obstacles and ensure robust collision avoidance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional obstacle detection systems are used in UAVs, then basic collision avoidance is achieved, but the system cannot effectively handle dynamic environments with moving obstacles due to limited sensor coverage and inability to update obstacle maps in real-time
Solution Approach 1:
The obstacle map is implemented as a dynamic data structure that is continuously updated in real-time as the UAV moves through the environment. The map automatically adjusts to reflect current obstacle positions, allowing the system to adapt to moving obstacles while maintaining reliable collision avoidance through continuous re-evaluation of the flight path against the updated map
Solution Approach 2:
The system employs feedback mechanisms where depth sensor data is continuously fed back to update the obstacle map, which in turn feeds back to the flight controller for real-time path adjustment. This closed-loop feedback ensures the UAV maintains reliable collision avoidance while adapting to changing environmental conditions
2Measurement precision
If comprehensive obstacle detection is implemented to improve collision avoidance, then detection accuracy increases, but system complexity and computational requirements increase
Solution Approach 1:
The three-dimensional space is segmented into a voxel grid map where each voxel represents a discrete volume element. This segmentation allows the system to achieve comprehensive obstacle detection by processing space in manageable units, reducing computational complexity while maintaining high detection accuracy through systematic evaluation of each voxel
Solution Approach 2:
The system transitions from two-dimensional image processing to three-dimensional voxel-based obstacle mapping. By adding the depth dimension and creating a volumetric representation of the environment, the system achieves superior obstacle detection accuracy while managing complexity through structured three-dimensional data organization
3Productivity
If real-time obstacle map updates are performed to improve navigation in dynamic environments, then collision avoidance improves, but processing time and computational load increase
Solution Approach 1:
The obstacle map is updated at periodic intervals synchronized with the UAV's movement and sensor refresh rate. This periodic updating strategy maintains high navigation efficiency by ensuring the map is current without requiring continuous processing, thereby reducing computational load while still providing timely collision avoidance
Solution Approach 2:
The system dynamically adjusts the update frequency of the obstacle map based on the UAV's speed, acceleration, and environmental complexity. During high-speed flight or in simple environments, update frequency is reduced to minimize processing time, while during maneuvers or in complex environments, updates are intensified to maintain navigation efficiency and safety
Data Source
Figure 1
Figure 2
Figure 3
AI summary
According to various aspects, an obstacle map generator is provided, including: one or more processors configured to receive one or more depth images from a depth imaging system, determine, for each depth image of the one or more received depth images, a first set of pixels and a second set of pixels, each pixel of the first set of pixels has a depth value assigned thereto and each pixel of the second set of pixels has no depth value assigned thereto or has a depth value outside a predefined depth value range assigned thereto, assign a pre-defined depth value to one or more pixels of the second set of pixels, and generate an obstacle map based on the determined first set of pixels and the one or more pixels of the second set of pixels having the pre-defined depth value assigned thereto.