UAV Waypoint Correction Using Visual Feature Point Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unmanned aerial vehicles (UAVs) face GPS positioning errors that can lead to collisions with obstacles during inspections, as they lack expensive distance measurement and obstacle avoidance modules, necessitating a visual positioning-based waypoint correction technology.
Innovation Solution
A waypoint correction device and method that utilizes a processor to analyze real-time images for feature points, adjust the viewing angle, and generate a correction route based on three-dimensional feature points and adjusted images to accurately guide the UAV to its intended waypoint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If GPS navigation is used for UAV waypoint guidance, then the device complexity is reduced, but the positioning precision deteriorates leading to potential collisions with obstacles
Solution Approach 1:
The patent introduces visual feature points as an intermediary between the GPS navigation system and the UAV control system. By matching detected feature points with pre-stored three-dimensional map data, the system obtains accurate position and orientation information without requiring complex RTK or visual-inertial navigation systems, thus resolving the contradiction between simple device complexity and positioning precision.
Solution Approach 2:
The patent replaces the purely GPS-based mechanical navigation system with a visual positioning system that uses camera-based feature point detection and matching. This substitution eliminates the need for expensive distance measurement and obstacle avoidance modules while achieving high positioning precision through optical field information processing.
2Measurement precision
If the UAV rotates to capture adjusted real-time images for better feature point detection, then the positioning precision improves, but the inspection efficiency deteriorates due to additional rotation and capture time
Solution Approach 1:
The patent applies partial rotation rather than full 360-degree scanning. The UAV rotates only to the extent necessary to capture adjusted real-time images that contain sufficient feature points for accurate positioning. This partial action approach achieves the required positioning precision while minimizing the time loss and maintaining inspection efficiency.
Solution Approach 2:
The system implements feedback control where the detected feature point distribution quality determines whether rotation is needed. If feature points are sufficient, no rotation occurs; if insufficient, the UAV rotates to capture adjusted images. This feedback mechanism ensures positioning precision is achieved only when necessary, thereby maintaining overall inspection efficiency.
Data Source
AI summary
A waypoint correction device and waypoint correction method are provided. In response to a positioning signal of an unmanned aerial vehicle at a predetermined waypoint, the waypoint correction device obtains a real-time image from the unmanned aerial vehicle. The waypoint correction device calculates a feature point distribution in the real-time image based on the real-time image. The device generates an adjusted viewing angle signal based on the feature point distribution to control the unmanned aerial vehicle to rotate in place based on the adjusted viewing angle signal and capture an adjusted real-time image. The device generates a correction route based on a plurality of three-dimensional feature points, the adjusted real-time image, and a sampling number threshold to control the unmanned aerial vehicle to move from an actual position to a predetermined waypoint based on the correction route.