Pre-Processing Irrelevant Information from Melanoma Infected Human Skin Images using Image Processing Techniques

The skin is the largest organ and it serves as the body's barrier against environmental factors and hazards from the outside world. An interesting and important part of dermatology is the study and identification of skin lesions, as these surface anomalies can be a sign of a number of underlying illnesses. Studying these lesions' categories, causes, and diagnosis techniques is necessary to comprehend the necessary further treatment. The image processing and computer vision plays a major role in evacuating many practical difficulties in finding the lesions and solving many issues related to skin. The approach used in the present study is unique in providing a proper solution in pre-processing techniques before analysis. The first step of the research project involved transforming the gathered photos into formats that would function well for simulations. The test histograms for RGB, HSV, LCH, and LAB show the colour distributions over the affected areas of the skin lesion. The graph's highest point represents the color's high intensity range. These gathered histogram peaks are used to test the different kinds of skin lesions. Variations in mild, moderate, and severe skin-infected disorders are also considered in this method. The study describes the feature extraction process used to choose the attributes required to detect the presence of melanoma in skin lesions.