基于MATLAB图像处理技术预测絮凝效果

Research on Forecasting Flocculation Effect Based on MATLAB Image Processing Technology

针对现阶段大多数水厂絮凝效果检测时间较长的问题，提出一个新方案：用电子摄像头采集水厂絮凝过程中不同时段水样絮凝状况图片，利用MATLAB软件进行相关地图像处理与分析，寻找出最终絮凝效果与不同时段絮凝状况之间存在地统计相关规律，并以此来提前预知絮凝效果。以华北某水厂实际运行中栅条絮凝池为研究对象，水样中矾花个数在50～500个／mL，总面积在600～4000，平均粒径在160～190um。本方法检测絮凝效果比传统的检测方法在时间上提前了58％。

In the present stage, for most of the water treatment factories＇ flocculation effect testing lag, we put forward a new plan： Using electronic camera capture flocculation water samples images in the different periods in the process of flocculation, which treated with MATLAB software are related to image processing and analysis, we find out the statistical correlation rule between final flocculation effect and the flocculation conditions in the different periods. It will predict flocculation effect in advance, we choose the operating grid flocculating tank in north China as research object. The number of alum floc in water at 50 - 500/mL, the total area in 600- 4000, the average particle size in 160 - 190 microns. This method to detect flocculation effect than the traditional in time by 58% in advance.