Skin Cancer ? Melanoma Detection in Skin Images Using Local Binary Pattern (LBP) and GLCM
Abstract: To improve the accuracy level, a k-means clustering is proposed followed by local binary pattern. This not only clearly detects the melanoma but also segment the cancerous part from the back ground. Further, the image is confirmed by using the local binary pattern in order to do the dimensional analysis of the skin cancer. The algorithm is tested on different skin image data base covering different stages of skin cancer and b=normal images. The results very accurate and later stage could be predicted in consultation with medical practioner. The prime concern in the presented work is on extracting the skin image features in textural domain as well as radial domain i. e. area, perimeter and standard deviation of radii. This enables in analyzing the cancer spot analysis and guides for the direction of spread of the cancer.
Keywords: Histogram Equalization, K-means clustering, LBP, GLCM, Dermoscopic Images
How to Cite?: Ramandeep Kaur, Gurmeen Kaur, "Skin Cancer ? Melanoma Detection in Skin Images Using Local Binary Pattern (LBP) and GLCM", Volume 4 Issue 7, July 2015, International Journal of Science and Research (IJSR), Pages: 134-139, https://www.ijsr.net/getabstract.php?paperid=SUB156304, DOI: https://dx.doi.org/10.21275/SUB156304