Detection of Cervical Cancer Disease Using Adaptive Thresholding Method bu Image Processing
Detection of Cervical Cancer Disease Using Adaptive Thresholding Method bu Image Processing
Author : EGGI INTAN PUTRI; LEDYA NOVAMIZANTI Published on : 2nd International Conference on Global Trends in Academic Research
Abstract
Cervical cancer is one of the cancer disease caused by human papilloma virus type 16 and 18, attacking woman cervix. The cervical cancer detection method which frequently used is Pap-Smear. But, error often occuring this method in diagnosing the level of cervical cancer. Therefore, a system is needed that is able to help identifying the result of Pap-Smear. In this study has design a system to detect the symptoms of cervical cancer using Matlab 2009a software to solve these problems. The image processing starts from converting the type of image, thresholding, noise removal using filter, until the image is ready to be detected. For thresholding process, used Adaptive Thresholding method, thresholding used local threshold value. This system is able to classify the image into two types, normal and abnormal (precancerous). For abnormal type divided into three types, that are mild, moderate, and severe. This system used 500 test images, 250 for training and 250 for testing. Based on testing process, obtained an accuracy rate of 100% and the average time to process this sytem is 25.4 seconds with WS value is 10 and C value is -2
