Research And Implementation Of Automatic Recognition And Classification System Of Ascaris Eggs

The rapid development of digital image processing technology has brought a huge boost to medical image processing,and provides efficient and reliable technical support for medical diagnosis.But now,medical personnel usually judge the kind of parasitic ovum under a microscope relying on artificial experience,which has low accuracy and efficiency.Therefore,this thesis identified and classified three different physiological stages of ascaris eggs(infectious,fertilized,unfertilized)using digital image processing technology.Ascaris egg images are smoothed and sharpened in order to suppress noise and enhance edge.This thesis segments ascaris egg images,and compares segmentation results of edge detection method,threshold segmentation and U-net based on the image preprocess of ascaris egg such as smoothing and sharpening;OTSU combined with mathematical morphology has a better segmentation effect according to the experimental results.Geometry,texture,color aspects of a total of 11 features are selected and extracted from ascaris egg segmentation,which are used as data set of classifiers.Shortest distance algorithm,K-nearest neighbor algorithm,probabilistic neural network,random forest and deep residual network Res Net-34 are adopted respectively for constructing classifier of ascaris eggs,and the five classification algorithms’ performance are compared on problem of ascaris eggs’ classification.In the process of the construction of classifier,this thesis searched out the optimal K value in K-nearest neighbor algorithm,and used particle swarm optimization to find the optimal smoothing factor σ in probabilistic neural network,so that made the two classification algorithms achieve better classification effect.Probabilistic neural network has better classification effect on the classification of ascaris egg from experiment results.Finally,this thesis designs the system of ascaris egg’s identification and classification based on the results of the above experiments,and uses MATLAB to develop and implement physiological stage classification system of ascaris egg.