Cervical cancer is the second most common cancer among the women worldwide. A computer aided diagnosis system can help colposcopist to analyze cervical images more accurately. This work aims to detect lesion in cervical images based on Laws textural feature and nearest neighbor classifier and it can be used as a diagnostic tool. The images used for the detection of cervical cancer are taken by using colposcope which magnifies the cells of cervix. The Laws textural features are extracted from the cervical images and input to nearest neighbor classifier. A totally 240 images are used for the evaluation and an overall accuracy of 96% is obtained.
展开▼
