The Flame Recognition And Background Disturbance Filtering Algorithm Based On Video Images

This article focuses on the flame detection and background interference filtering algorithm based on the video images. First of all, detailed analyze the characteristics of flame image, using the characteristics of highlight, process the flame image by use the image pre-processing technology, such as gray-scale images, optimal threshold segmentation, binarization and so on, then extracted the highlight suspicious region from the image. And then briefly introduced the basic theory of the traditional frame-difference method, analyze the advantage and disadvantage of its practical application, based on the traditional frame-difference method, proposed an improved frame-difference method, using the improved frame difference method to filter the interference of rapidly moving objects with high brightness. Finally, simple analysis of the characteristics of circular degree, for the circular degrees in the standard circular have the minimum and the more irregular of a shape, the greater its circular degree value is, so can use the circular degree to filter out the static highlight human disturbance, then the flame region be detected accurately.The results of the experiment show that the algorithm can effectively filter out background interference and detection flames accurately, and the calculation is simple, easy to implement.