The Water Microbial Image Detection Technology Research

The rapid development of modern industry and continuous improvement of living standard exacerbate the disposal of sewage, so that sewage-disposal has become more and more important. Nowadays, one of the most effective approaches to treat waste water is to use microbes for cleaning away the waste in water. As a result of wide application of computer technology, the technology of detection of microorganisms in water, which is based on the technology of digit image processing, is paid attention in the area of monitoring the quality of water, prevention and treatment of water pollution. Besides, scientists make groundbreaking achievements for this detection technology.On the basis of the technology of image process, the classification and enumeration of microbes in water are realized by the pretreatment of microscopic image of microbes, target extraction, feature extraction, feature optimization and classification. First, this study introduces the overall procedure of image detection of microbes in water. Moreover, according to the diversity of microbes, the complexity of microscopic image background and other particularities, I propose further improvement and a new proposal for image pretreatment and target extraction. Furthermore, via the study about the technique of simulated annealing and genetic algorithm, I apply genetic simulated annealing algorithm to feature optimization to optimize the accuracy and efficiency of detection of microbes in water.The main work and achievements of this study includes:(1) Image pretreatment---this part detailedly shows the image-processing technique, which includes image grayness, grayness-correction, image filtering, mathematical morphology and so on, in the field of image pretreatment of microbes in water(2) Image target extraction---this part focuses on the algorithm of target extraction on the basis of target-extraction of Otsu method, multi structure element anti-noise morphological edge detection, and RFCM clustering. As to the target-extraction of images with complicated background, I still apply interactive target-extraction proposed by Chen Yao and her colleague. (3) Image feature optimization---this section systematically analyzes genetic algorithm and simulated annealing algorithm. Based on the combination of genetic algorithm and simulated annealing algorithm, I optimize the16extracted features and select10preferable features according to their area, perimeter and complexity.