Font Size: a A A

Flight Target Attitude And Rotational Speed Measurement Technology Based On Stereo Vision

Posted on:2024-04-11Degree:MasterType:Thesis
Country:ChinaCandidate:Q X ZhaoFull Text:PDF
GTID:2542307061466694Subject:Electronic information
Abstract/Summary:PDF Full Text Request
With the wide application of machine vision measurement technology in military,aerospace,transportation and other important fields,the estimation of the motion state of the flight target with the help of optical measurement equipment has also become a hot issue and technical problem at present.The binocular stereo vision measurement technology,with its advantages of non-contact,high efficiency,high accuracy and high flexibility,is showing excellent development prospects and is receiving key attention from talents in various fields.This paper is dedicated to solving the problem of measuring the attitude and rotational speed of the flight target based on the measurement method of stereo vision,focusing on the calibration of the stereo vision measurement system,the detection of the foreground information of the flight target,the attitude measurement by the central axis method and the rotational speed measurement by the auxiliary marker method.In the end,the calibration,target detection,attitude solution and rotational speed measurement functions of the measurement system are realized.The main contents of the research and the obtained results of this paper are as follows:1)The overall test scheme of the measurement system is designed,and the components and test process of the measurement system are analyzed.The spatial position measurement model of the flight target with stereo vision is introduced based on the imaging model of monocular camera.Then,the calibration method of the stereo vision measurement system is derived according to the traditional Zhang Zhengyou’s plane calibration method,and the aberration parameters of the two lenses and the solution method of the position relationship between binocular cameras are given.Finally,the system calibration is completed.2)In order to complete the detection of the foreground information of the flight target,this paper investigates various methods for detecting the foreground information of the flight target,and suppresses its ghosting phenomenon by the improved Vi Be algorithm.And a method to measure the attitude of the flight target using the median method is studied.The target’s edge information is first obtained by using mathematical morphology,image noise reduction and Canny edge detection,and then Hough straight line detection and least squares method are used to extract the central axis of the flight target on the image plane.At last,the spatial position of the target’s central axis is solved by the plane-to-plane intersection method,and the attitude measurement of the flight target is achieved.3)The dynamic region segmentation method of the flight target is studied and completed by using image mask algorithm.And a method to measure the rotational speed of the flight target using the auxiliary marker method is studied.The target’s edge and auxiliary marker information are obtained by using binarization,image noise reduction,edge detection and edge refinement.The auxiliary flags are then extracted using Hough line detection and least-squares curve fitting.And finally,the auxiliary markers are resolved to achieve the measurement of the flight target rotational speed.4)The paper designs the overall process of measurement experiments,builds the experimental platform,and implements the software design.And an accuracy analysis model of binocular stereo vision measurement system is established.The accuracy analysis is carried out mainly for the parameters of baseline distance,camera focal length and angle between optical axis and baseline.After that,the comprehensive error is analyzed with the measurement results of attitude and rotational speed,and the experimental results are verified.The relative measurement error of attitude is about 0.73°,and the relative measurement error of rotational speed is about 2.61%,which meets the expected index.
Keywords/Search Tags:stereo vision, system calibration, moving target detection, attitude measurement, rotational speed measurement
PDF Full Text Request
Related items