Velocity Measurement Of Mountain Stream Based On Image Processing

According to classic floating particles monitoring system, velocity measurement of mountainous stream based on image processing is promoted. It utilizes video monitor technology to overcome inherent flaw. It monitor natural, or manual floating particle, apply solid geometry and photography, establish transforming model, precisely track floating particles and position, and get velocity of flow. This method need only simple equipment, less labor; it can simultaneously acquisition multiple data; when there is a searchlight, or other lighting device, it can not miss peak; Velocity data can be automatically displayed, and saved, convenient reviewed, etc. With the rapid development of science and technology, it is significant subject to either the improvement of automatic measurement technology, or flood and disaster prevention.The paper addresses the principle of velocity measurement system of mountainous stream based on image processing in detail, design and achieve the whole system, get higher precision than classic time difference way. This article mainly as follows:1. The paper elaborates main calibration contents and conventional methods, adopts pigeon-hole parameter model and Zhengyou Zhang's way to do off-line calibration.2. This paper analyzes the basic principles of the median filtering and average filtering and their characteristics, uses median filter to weaken noise. And for the possible effect of blurring the edges, the Laplacian Operator is adopted to enhance image contrast.3. Detailed analysis of the common method of moving target detection, and for the stream of the river's character, the continuous frame difference method is adopted to detect moving object.4. Detailed analysis of the Match tracking algorithm and the basic principles of Kalman tracking, according to the stream of the river for laboratory simulation platform, this paper Propose a fast target tracking.5. Through the trajectories of floating objects, combined with spatial resolution, the river's surface flow was calculated .Compared to the velocity from classic time difference, the results show that river flow measurement based on image processing can meet practical engineering requirements.