The Flocs Target Detection Algorithm Based on the Three Frame Difference and Enhanced Method of the Otsu
Xin Xie, Huiping Li, Fengping Hu · International Journal of Computer Theory and Engineering · 2014
At present, the commonly used floc tracking algorithms are susceptible to noise, light, and floc movement speed, etc., it is difficult to extract complete moving flocs, therefore put forward a flocs target detection algorithm based on the three frame difference and particle swarm optimization & enhanced method of the Otsu.The algorithm firstly conducts difference of three consecutive frames, then use particle swarm optimization & enhanced method of the Otsu to obtain the best threshold, the optimal threshold is adopted to the image binarization and then do image post-processing, eventually obtaining a clear floc target, to lay a foundation for subsequent automation analysis of floc.Experiments show that the algorithm has the characteristics of fast, accurate, and can extract flocs target effectively.