Computer Vision Based Skin Cancer Classification by Using Texture Features

Cancer of the skin is now one of the most prevalent forms of the disease among people. As a result, accurate diagnosis of malignant lesions is of utmost significance in treating skin cancer. Dermatoscopic images can be used with computer-aided diagnostic tools, which may include machine learning models, to assist medical professionals in diagnosing skin cancer. In this particular research project, skin lesions were classified using image processing and machine learning strategies. Several distinct mathematical techniques have been implemented in the field of image processing in order to improve image quality. Image segmentation utilizing the watershed approach was conducted after an image preparation step, which included filtering the undesired pixels in the pictures. Following that, the lesioned regions were separated, and texture feature extraction was carried out. In the end, the classification was completed using the SVM algorithm, which stands for support vector machines. When the results acquired from the classifiers were compared, it was seen that the SVM classifier had an accuracy of 94.33%.