Font Size: a A A

Research On Simple Monocular Spatial Relative Position And Attitude Measurement Of Cooperative Target

Posted on:2018-07-01Degree:MasterType:Thesis
Country:ChinaCandidate:Y CaoFull Text:PDF
GTID:2348330542984923Subject:Instrumentation engineering
Abstract/Summary:PDF Full Text Request
Based on the problem of relative pose measurement between two free coordinate systems,a new attitude angle definition method is introduced and a plane object position and attitude solution algorithm based on monocular vision is proposed.And this paper use FPGA platform to realize the image preprocessing algorithm which is used to obtain the binarized image to extract the image coordinates of the target point.The results of simulation analysis and experimental verification show that the measurement accuracy and speed of the system meet the system design requirements.This measurement system consists of four parts: CMOS module(image acquisition module),FPGA module(image preprocessing module),DSP module(pose solving module)and the planar double-loop LED target.In the measurement model which uses a plane object position and attitude solution algorithm based on monocular vision,a new attitude angle definition method is introduced,and the three attitude angles are not restricted by the order of rotation.The three attitude angles and position vectors of the plane model with respect to the camera coordinate system can be directly derived from the plane homography matrix which can be solved by using 4 coplanar LED light sources' coordinate information.The algorithm in this paper satisfies the system precision requirement,reduces the computational complexity and improves the effectiveness of the system to a certain extent.This system takes use of advantages of FPGA,including high-speed processing speed and parallel operation to implement the configuration of CMOS camera parameters,median filtering algorithms,histogram equalization and binarization algorithms of the real-time video images.Through this way,this system meets the realtime requirements of the system,and we can conduct field performance maintenance and function expansion according to the needs of system optimization.Using FPGA increases flexibility and versatility of the system.With the error analysis of two kinds of error sources which are the pixel image coordinates and target location error in object coordinate in this pose measurement algorithm,the error range of these two parameters is determined.And to verify the accuracy of the position and attitude of the proposed algorithm,we build the simulation platform to carry out simulation experiment outdoors and compare with Fischler's P3 P algorithm;we verify the effectiveness of FPGA module by the comparison of the processed image and original image,and use the real-time measurement data to illustrate the real-time performance of the system.Through the experimental verification and comparative analysis,it fully proves that this system meets the requirements of accuracy and speed,and has a certain practicality and effectiveness in the relative position and attitude measurement between two free coordinate systems.
Keywords/Search Tags:Monocular vision, FPGA, Image processing, Pose measurement, Two-dimensional cooperation objectives
PDF Full Text Request
Related items