SVM and CNN based skin tumour classification using WLS smoothing filter

Skin cancer is considered as one of the most hazardous types of cancer, with a sharp rise in mortality due to a lack of knowledge of symptoms and prevention. In order to stop cancer from spreading, early identification at an early stage is required. Automatic classification of skin lesions is always a challenging task due to the different shape and size of tumour, low contrast, light reflections from the skin surface etc. so, we propose an image classification, by decomposing the input image into base and detailed layer, then applying the bilinear interpolation to both the layers and then applying the WLS filter to the detailed layer and then merging the base layer and modified detailed layer. After obtaining the enhanced image, the enhanced images are used for classification by training the tumour images from the dataset and the enhanced images are used for testing. With the SVM and CNN classifiers, we are achieving the classification accuracy around 98% and 98.5%.