Research On Image Stabilization Techniques Of Visual Measurement System

As an important branch of intelligent manufacturing and computer science,vision measurement technology has a wide range of applications in various detection tasks in industrial,medical,and military fields such as security monitoring,microscopic imaging,aerospace mapping,medical assistance,and scientific research.The visual measurement system image acquisition process is vulnerable to external environmental factors.Under complex conditions,vibration,noise,and other disturbances result in decreased performance and stability of the vision measurement system.The existing vision measurement technology has the effect of vibration caused by image jitter when the visual measurement system is applied under complex conditions and interference conditions,all the functions of the vision measurement system fail to meet the required stability and accuracy,therefore,it is of theoretical and practical significance to study the technology of image dejitter and solve the problem of image jitter.In this paper,the parallel method is verified through theoretical research and experimental testing,taking the visual measurement system as the research object,from the theoretical analysis of the image measurement system of the vision measurement system,the key technology analysis of the image de-shaking based on the vision measurement system and the image de-shaking algorithm design of the vision measurement system,a more systematic theoretical study and in-depth experimental analysis of the image de-shaking technology applied to visual measurement systems was carried out,the relevant research results have been verified experimentally in the visual measurement system independently developed by the research group.The main work of the thesis can be summarized as follows:1.In view of the jitter effect of visual measurement systems,the sources and causes of jitter are analyzed,the specific analysis of the effects of various jitter sources and jitters was established,and the model of the vision measurement system movement wasestablished,in-depth exploration of the key technologies of all aspects of the vision measurement system,including image motion model,motion estimation,motion compensation and algorithm effect evaluation methods.Based on the corresponding relationship between camera shake and image jitter in the analyzed vision measurement system,establish a mathematical model for image measurement of a vision measurement system.2.For the limitation that the traditional projection algorithm is not suitable for grayscale single image in image debounce in vision measurement system,the application range and advantages and disadvantages of image debounce algorithm are compared systematically.Gradient transformation of the collected image,the idea of combining gradient projection and image segmentation into the gray projection algorithm,using gradient image gradient projection and vertical gradient projection to determine the jitter vector,using this method to do the debounce Simulation,at the same time,the effectiveness of the algorithm is evaluated from the four indicators of accuracy,noise fitness,peak signal-to-noise ratio,and algorithm speed.3.The proposed image debouncing algorithm is applied to the visual measurement system independently developed by the research group.For the on-line and off-line measurement state,the tool parameters of the NC milling cutter are collected separately,the experimental images of the collected images were simulated using the traditional algorithm and the algorithm,the visual measurement system independently developed in the laboratory validates the effectiveness of the proposed algorithm.