International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 132 - Number 8 |
Year of Publication: 2015 |
Authors: Md. Al-Amin, Mohammad Badrul Alam Miah, Md. Ronju Mia |
10.5120/ijca2015907513 |
Md. Al-Amin, Mohammad Badrul Alam Miah, Md. Ronju Mia . Detection of Cancerous and Non-cancerous Skin by using GLCM Matrix and Neural Network Classifier. International Journal of Computer Applications. 132, 8 ( December 2015), 44-49. DOI=10.5120/ijca2015907513
Day by day the use of image processing is increasing. Now a days image processing is the part and parcel of medical science. By image processing many types of cancer are easily detected. Skin cancer is one of them. In this paper the proposed method detects two types of skin one is cancerous skin and another is affected but not cancerous skin. Skin cancers are most common cancer in human. Skin cancers are curable cancer after early detection. The system can distinguish cancerous skin and non-cancerous skin based on some values of features. Some value extracted from Grey Level Co-occurrence Matrix (GLCM). GLCM features include Contrast, Correlation, Energy, Entropy and Homogeneity. Besides those MajorAxisLength, MinorAxisLength, Solidity, Equivdiameter, Perimeter, Mean, Standard Deviation, ConvexArea, Area, Euclidean Distance, Manhattan Distance, Minkowski Distance and Hamming Distance. There are several steps for evaluating the process. The first step is preprocessing, in this step the noise is removed by using filter. The filtered image is segmented into gray level and black and white (BW) image. All the features are calculated on black and white (BW) image. The neural network is used to classify the images. It is an easy system rather than the doctor biopsy procedure. The system consumes less time and gets better result than ordinary systems.