Research On Electronic Image Stabilization System Of Video Monitoring For CBM Drainage Equipment

Due to severe operation environment of the CBM drainage equipment,the monitoring device in such environment is easily influenced by external environment,and thus random jitter would occur to cause false alarm often.The technology of electronic image stabilization could well solve this issue.It is a kind of technology to stabilize the video sequences,with the advantages of high accuracy of image stabilization,small size,light weight and low power consumption and so on.It draws much attention both in the fields of military and civilian technology.First,this paper elaborates basic principles and the processing of electronic image stabilization.It introduces several common algorithms,including block match algorithm,bit-plane matching algorithm,gray projection,and feature tracking.It also analyzes the strengths and weaknesses of these algorithms from the precision of image,the operation speed of algorithm and algorithm application environment.According to this research’s requirement,this paper chooses the feature point tracking algorithm based on Harris corner detection to achieve the object of electronic image stabilization.During the image preprocessing,aiming at the deficiency of the original switching median filter in preserving fine lines and details,an improved switching median filter algorithm is put forward.Through the simulation experiment,to filter the noise signal is increased from 5%to 30%of the images,the improved method than the original peak signal-to-noise ratio increased an average of 3.641 dB,image damage degree is reduced by 1.2%.In contrast to the original algorithm,the improved algorithm is better in terms of saving the thin lines and the detailed features in the images.Then based on the feature points of images extracted by Harris corner detection,it matches the features,estimates the motion parameters and then compensates the motion vectors to acquire the final stabilized image consequences.The second,transplants the electronic image stabilization algorithm to the platform of TMS320DM6467 from the company TI,and then optimizes the algorithm.Configure the DAVINCI development environment based on the structure and framework of DM6467.And the process of transplanting the algorithm described in detail.The algorithm is optimized by two methods of program optimization and memory optimization,and the running speed of the algorithm is improved.Among them,the program optimization includes C language optimization and assembly language optimization.Finally,it utilizes the subjective evaluation and the objective evaluation to evaluate quality of the image after stabilization.Subjective evaluation provides the method of the image quality evaluation,it compares the image of the same image before and after the experiment with a number of students as judges,and then take the average value,after the image stabilization,the average value of the image increased about 1.3 point.The objective evaluation includes peak signal to noise ratio,weighted peak signal to noise ratio and difference comparison method,based on the analysis of the results of image processing on the scene it can be seen that the value of the peak signal to noise ratio of the stable image is improved by 7.542dB,the weighted peak signal to noise ratio is improved by 6.614dB,it can also be seen that the gray value of the image is small after comparing with the image of the two adjacent frames before and after the image stabilization.Combine the two methods,it could be concluded that this method could effectively achieve electronic image stabilization in the platform of DM6467,satisfied with the final result.