Drone Wind Turbine Blade Detection via Binocular Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current drone-based methods for detecting wind turbine blade damages suffer from low detection accuracy due to discrepancies between actual and expected postures during image capture, affecting image quality and damage identification.
Innovation Solution
A method and apparatus that determine initial three-dimensional position coordinates and posture of wind turbines, adjust the drone's camera orientation using binocular vision techniques to ensure uniform orientation, and utilize laser ranging to establish models for damage detection, allowing for precise image capture and damage assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a drone-mounted camera is used to detect wind turbine blade surfaces, then the detection can be performed remotely, but the detection accuracy is reduced due to discrepancies between actual and expected postures during image capture
Solution Approach 1:
The patent employs feedback mechanisms through real-time posture monitoring and image quality assessment. The system continuously monitors the drone's actual posture during flight and compares it with the expected posture for optimal imaging. Based on this feedback, the system adjusts the camera orientation and drone positioning to minimize posture discrepancies, thereby maintaining high detection accuracy while preserving remote operation capabilities.
Solution Approach 2:
The patent replaces manual mechanical positioning with automated control systems. Instead of relying on operators to manually position the drone and camera, the system uses automated posture adjustment mechanisms, GPS-based positioning, and algorithmic control to maintain optimal imaging conditions. This substitution of mechanical manual positioning with automated systems resolves the contradiction by enabling both remote operation and high precision simultaneously.
2Measurement precision
If the drone adjusts camera orientation to ensure uniform orientation for improved image quality, then detection accuracy improves, but the complexity of the detection system increases
Solution Approach 1:
The patent integrates multiple functions into unified components. The camera system serves both as the primary detection instrument and as a posture reference source. The same imaging sensors used for damage detection also provide feedback on camera orientation and positioning. This multi-functionality reduces system complexity while maintaining improved detection accuracy through uniform camera orientation.
Solution Approach 2:
The system implements self-service through automated posture correction. The camera system automatically monitors its own orientation and makes necessary adjustments without external intervention. The drone's control system automatically compensates for posture deviations based on real-time data from onboard sensors, eliminating the need for complex manual adjustment mechanisms while maintaining high detection accuracy.
3Measurement precision
If laser ranging is used to establish detailed blade models, then damage identification capability is enhanced, but the time required for detection increases
Solution Approach 1:
The patent applies preliminary action by pre-establishing reference models of healthy blade surfaces and pre-programming detection algorithms. Before actual detection, the system prepares standardized blade models and detection protocols. During field detection, the laser ranging data is immediately compared against pre-prepared references, significantly reducing processing time while maintaining enhanced damage identification capability through detailed 3D modeling.
Solution Approach 2:
The system implements partial action by selectively applying laser ranging only to critical blade areas or regions where anomalies are suspected. Instead of performing exhaustive scanning of entire blade surfaces, the system focuses laser measurement resources on high-priority zones, achieving effective damage identification with reduced detection time through targeted rather than comprehensive scanning.
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
Improves image quality and detection accuracy by ensuring consistent camera orientation and utilizing laser ranging for detailed blade modeling, enhancing the identification of damages on wind turbine blades.
Implementation Method 1
controlling the drone to track at least one first key point of blades of the wind turbine in a first posture with a binocular camera at the actual three-dimensional coordinates, and determining a three-dimensional movement trajectory of the first key point in a camera coordinate system
Implementation Method 2
controlling the laser device to perform laser ranging on the blades multiple times by scanning all the time in a process of flying around the blades, so as to obtain a plurality of laser ranging results
Data Source
AI summary
A method, an apparatus, an electronic device, and a computer-readable storage medium for detecting a wind turbine blade based on drone aerial photography are provided. The method includes: determining first initial three-dimensional position coordinates and a posture of a wind turbine; determining second initial three-dimensional coordinates of a first expected position with respect to a front shooting position of the wind turbine, based on the first initial three-dimensional position coordinates, the posture, and a preset distance; inputting the second initial three-dimensional coordinates of the first expected position into a flight control system of the drone, and controlling the drone to fly to the second initial three-dimensional coordinates based on a GPS system; and controlling the drone to track at least one first key point of blades of the wind turbine, and determining a three-dimensional movement trajectory of the first key point in a camera coordinate system.