An improved BP neural network for wastewater bacteria recognition based on microscopic image analysis
Li Xiaojuan, Cunshe Chen · 2009
Abstract: The microscopic images of wastewater bacteria are analysed, and a scheme for classification and recognition for wastewater bacteria based on microscopic images analysis are put forward in the paper. An adaptive and enhanced edge detection solution for the images of wastewater bacteria is proposed, which can effectively remove noises in the images and get clear edges of microscopic image by optimizing segmentation threshold and the varied order of edge detection. Seven contour invariant moment features and four morphological features are extracted by analysis of microscopic images of wastewater bacteria in which six features are chosen by PCA in order to reduce the dimensionality of the features extracted from the images. A self-adaptive accelerated BP algorithm is developed for training the classification of bacteria microscopic images. The proposed method is tested with CECC database and the results show that the presented image recognition solution is effective and can greatly improve the speed and consistency in performing large-scale surveys or rapid determination of bacterial abundance, morphology.