UAV Obstacle Avoidance Using Depth Image Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current obstacle avoidance methods for unmanned aerial vehicles (UAVs) in automatic flight mode require creating and maintaining environment models, which is time-consuming and computationally intensive, necessitating high-cost processors.
Innovation Solution
A method using depth information from images to determine flight direction and distance, employing an image capturing unit and processing unit to generate a depth image, binarize it, and control the UAV to avoid obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If environment model creation and maintenance is used for obstacle avoidance, then obstacle avoidance capability is improved, but time consumption and computational load increase
Solution Approach 1:
The patent extracts only the essential depth information needed for obstacle avoidance from the complete environment model, using depth images to represent obstacles rather than maintaining comprehensive environmental data. This selective extraction reduces time consumption while preserving obstacle avoidance capability.
Solution Approach 2:
Instead of creating an environment model and then extracting obstacle information, the patent inverts the approach by directly capturing depth images that immediately represent obstacles. This reversal eliminates the time-consuming model creation process while maintaining obstacle detection effectiveness.
2Reliability
If environment model creation and maintenance is used for obstacle avoidance, then obstacle avoidance capability is improved, but computational load increases
Solution Approach 1:
The patent extracts only the necessary depth information for obstacle avoidance from what would otherwise be a comprehensive environment model. By using depth images to represent only obstacle-relevant data, computational load is significantly reduced while obstacle avoidance capability is maintained.
Solution Approach 2:
The patent uses disposable depth images captured in real-time rather than maintaining a persistent, computationally expensive environment model. Each depth image is processed independently and discarded after use, reducing overall computational burden while preserving obstacle avoidance functionality.
3Productivity
If high-efficiency processor is equipped for flight path calculation, then flight path calculation speed is improved, but device cost increases
Solution Approach 1:
The patent extracts only the essential depth information needed for flight path calculation, eliminating the need for complex environment models. This reduction in data complexity allows standard processors to achieve sufficient calculation speeds without requiring expensive high-efficiency processors.
Solution Approach 2:
The patent uses simple, disposable depth images instead of complex environment models, enabling flight path calculation with less powerful, more cost-effective processors. The reduced computational requirements eliminate the need for expensive high-efficiency processing units.
Data Source
AI summary
Method for providing obstacle avoidance using depth information of image is provided. The method includes the following steps. Shoot a scene to obtain a depth image of the scene. Determine a flight direction and a flight distance according to the depth image. Then, fly according to the flight direction and the flight distance.