Objective ? This paper proposed feature extraction of Pap smear slide images and present an automated method for extracting the features from the input images (cancerous and non-cancerous cells). Fuzzy logic technique gives better results in improving the image parameters and provides a better diagnosis of cervical cancer. Methods- In this work, there are 3 stages. In first stage, preprocessed images were detecting the edges using fuzzy logic. The detected edges are converted into gray-level co-occurrence matrix for extracting texture features. In second stage, the filtered images were selecting the particular region of nuclei and segment with threshold technique for extracting region features. In third stage, the filtered images were segmented with colors using Fuzzy C-means clustering method for color-intensity features. Results ? There are 228 slides of different 7 classes are used to extract the features for classifying cancerous and non-cancerous cells are present in the Pap smear slides. There are 22 features are extracted from the input slides of different stages of classes. Conclusion ? Feature extraction techniques provide the statistical measures for further implementation of feature selection and classification for classifying cancerous and non-cancerous cells are present in Pap smear image.
展开▼
