The Analysis Of Collimator Load Dynamic Parameters Based On Image Recognition

In the recent year, the application of the high precision guidance weapons in military affairs has been paid attention to widely, the precision of the guidance weapons is important for qualities and operations of arsenals' products. It becomes the focus of the research to process huge image information in good time and improve speed in recognition.The paper has researched the key technology of precision test of collimator. In the research, The paper mainly uses computer vision for the important measure, and adopts other measures including digital image processing, Hough transform and so on. In this paper, it can be described as follow:(1) Presents some basic strategies of computer vision, the development and application. This paper emphasizes on introducing algorithms of Hough transform theory on lines, circles detection.(2) Chooses operator as the best edge detection algorithm for basic geometric shapes detection by comparing the performance with other operators.(3) Hough Transform was used to detect the line and the crossing point of lines.(4) Proposes the improved Randomized Hough Transform of the traditional methods to detect circles. Experiments show that the improvements reduce the dependence on parameters of detection algorithm, enhances the speed, stability and accuracy of detection.(5) Obtains the distances based on crossing points of lines and centre points. (6) The design and exploitation of software for experiments is also completed based on MATLAB7.0.