Automatic Calibration And Accuracy Evaluation Of Camera Parameters

With the continuous improvement of hardware and computer performance,machine vision technology is widely used in artificial intelligence,vision measurement,robotics and image processing.Camera is the most important sensor in vision measurement technology,camera calibration technology is the key step of vision measurement technology.Through the camera calibration process,we can quickly establish the accurate position relationship between different coordinate systems,and further realize the position tracking and size of the measurement object.In the process of camera calibration,there is a large amount of work and the calculation of calibration parameters is complex.Different camera or different purposes may adopt different calibration methods.In the production of machine vision system,there are a lot of cameras need to be calibrated,which will be a very time-consuming,laborious and technical work.The traditional calibration method is to collect the image of the calibration target plate by hand-held camera,and import the collected image into the professional camera calibration software,and finally calculate the internal and external parameters of the camera.This kind of method depends on the technical level of the staff,which is not only time-consuming,laborious and error prone,but also can not evaluate the calibration accuracy effectively.Aiming at this objective demand,this paper plans to build a set of automatic camera calibration device,which can realize the automatic calibration of camera parameters,and develop the methods and algorithms that can effectively evaluate camera parameters.It can make the whole process of camera calibration automatic and improve the production efficiency.The camera parameter automatic calibration device developed in this paper realizes the automatic acquisition of the target plate image by using the mechanical arm instead of the manual and software control,and automatically transfers the collected image data into the camera calibration module to automatically calculate the camera parameters.The calibrated camera parameters are imported into the accuracy verification module,and four evaluation methods are used to comprehensively evaluate the calibrated parameters.The device realizes the closed-loop of the whole system and effectively improves the efficiency of camera calibration.The main contents of this paper are as follows:1? According to the technical specifications of the camera parameter automatic calibration device,the hardware involved in the hardware platform is selected and designed,and the software platform is built to realize the synchronous control among the hardware;2? In this paper,the algorithm and technical method used in the development of automatic camera parameter calibration device are studied.The algorithms of feature point center location,Tsai calibration and brown model based beam self calibration are studied.These algorithms are encapsulated in the solution module of camera parameter automatic calibration device;3? Four methods of camera calibration accuracy evaluation are developed.Based on the common point conversion algorithm and the single image space resection algorithm,four kinds of camera parameter evaluation methods are proposed,which are absolute space evaluation,image residual evaluation,relative rotation angle error evaluation and target measurement system error evaluation.With these four evaluation methods,the camera calibration parameters are evaluated as a whole to improve the reliability of accuracy verification.Finally,the calibration results and accuracy are evaluated The estimation results are output in the form of word.