Classification of Urinary Sediments Image Based on Bayesian Classifier

In this paper a new method of classification of image in medical domain was introduced. Since the traditional way of diagnosis is slow, time-consuming and heavy-workload, the result is largely influenced by the experience of doctor. The new way of classification can largely improve the efficiency and accuracy in diagnosis. The method is based on Naive Bayesian classification. After distinguishing the visible compositions, feature extraction and feature selection, we can get a naive Bayesian classifier which can be used in classifying of object entities. Experiments show that the new method is suitable for image classification.