Research On Pyramid Angle Detection System Of Helicopter Rotor Blades Based On Stereo Vision

Pyramid angle of helicopter rotor blades is an important parameter that reflects the dynamic balance characteristics of the helicopter rotor,which directly affects the safety,stability and maneuverability of the helicopter.Therefore,the precision measurement of pyramid angle of helicopter rotor blades provides an important foundation for the production and maintenance of the helicopter rotor.In the portable measurement system of pyramid angle of helicopter rotor blades based on stereo vision,the collected rotor images often have the large scene,complex background and contain some disturbing noise,which leads to the slow and low accuracy of the traditional mark detection algorithm.Aiming at the above problems,a deep research in the fast and precise positioning method of circular marker of the rotor image has been conducted in this paper.The primary research work and achievements of this paper are as follows:(1)The traditional methods and base theory were overviewed about pyramid angle detection system of helicopter rotor blades based on stereo vision.Firstly,the measurement principle of helicopter rotor taper and the principle of binocular stereo vision are introduced.Secondly,the portable measurement system of pyramid angle of helicopter rotor blades based on stereo vision is described in detail.The system consists of hardware system and software system.The hardware system includes photoelectric sensor subsystem,binocular image acquisition subsystem,PC/104 embedded subsystem,power supply subsystem and configuration tool subsystem.The software system includes configuration software and rotor taper measurement software.Finally,the reasons of slow positioning and low precision of the circular marker detection algorithm in the system are analyzed,and a deep research in the location method of circular marker has been conducted.(2)A fast and accurate locating method for circular marker based on CIC was proposed(CIC).Firstly,the Canny edge detector is adopted to inspect the image’s edge and to extract the connected domain of the edge image.Secondly,the circularity,inertia,and convexity of the connected domain are employed to filter the non-circular contour from the connected domain in the image.Finally,least-square circle fitting is used to complete the accurate locating of the circular marker.The simulation and real experiments show that the circular markers can be inspected and located with the proposed algorithm in complex scenes with high speed and good accuracy.However,when the rotor is rotating at high speed and the focal length of the camera is not allowed,result in the edge of the circular marker is blurred,which leads to the failure in detection of circular marker.(3)Regarding the issue above,a fast and accurate locating method for circular marker based on EDCIC was improved(EDCIC).Firstly,the image is passed the Gaussian filter to suppress noise and smooth the image,then obtain the gradient magnitude and direction at each pixel by Sobel,and the anchors are searched by the gradient magnitude and direction.Secondly,the anchors are joined via the Smart Routing algorithm,and then obtain the edge segments.Thirdly,the Helmholtz principle is used to judge the edge segment and remove the clutter in the edge image.Finally,the CIC method is used to detect the circular marker of the edge image.The simulation and real experiments show that the EDCIC compared with the traditional circular marker detection algorithm works real-time,has high detection rates,produces accurate results,solves the missed problem of circular marks due to blurry edge,and could be used to improve the speed and accuracy of measurement for the helicopter blades motion parameters against a complex background.