Machine Vision-based Automatic Identification Device For Yaw Angle

Steering gear is an important part of missile control system.It executes aircraft commands and drives the blades of air rudder to deflect.In the comprehensive test,the response performance of the steering gear to the command is usually tested by reading the deflection angle of the blade.There are some potential safety hazards in the way of manual reading,which can not ensure the reading accuracy.With the development of computer and image technology,machine vision technology has become more and more mature.At present,it is a general trend to use machine vision for non-contact testing.Therefore,I designed a set of automatic yaw angle recognition device based on machine vision.This paper describes my research process.Firstly,this paper introduces the tooling used for auxiliary testing and camera installation.The algorithm processing is divided into three parts: preprocessing,region of interest processing and angle acquisition.After graying the image,Gaussian filtering,image enhancement and corrosion expansion are used to preprocess the image;The preprocessed image pointer and scale line are filled.In the process of processing the region of interest,calibration,edge detection and template matching are used;Calibration can correct the distortion of the image.Marr operator is used for edge detection,and then the pointer template is extracted to quickly locate the pointer on the image.The angle is divided into three parts: integer part,whether to add 0.5 °,and the angle between the pointer and the scale line.OCR function,Hough line detection and angle measurement are used in the calculation process.In addition,the special positions of the pointer and the scale line are analyzed.While the design is completed,several groups of data are compared with this design by using manual reading.The maximum absolute error of this design is 0.08 °,and the average absolute error is 0.039 °.The test stability and accuracy meet the design requirements.