Vitiligo (सफेद दाग) is a disease that causes loss of skin colour in patches. The colourless areas usually get bigger in size with time. The condition is not body part specific; it can affect the skin on any part of the body, even in hair and the inside of the mouth too. There are various applications that detects the vitiligo by detecting human skin, but no one of them detects the skin tissue. Histo graphical image of tissue of vitiligous skin detects the extend of disease, a person is suffering from. By knowing the current damages in skin tissue due to vitiligo, proper and precise treatment can be given to sufferer. The present work is done in four phases. In first phase, it takes the microscopic image of a tissue. In second phase, it applies the machine learning algorithm on that image, to check for a tissue whether the image is having the characteristics that mark the tissue infected or not. In third phase, it makes a graph of different colours present in that image. In fourth phase, based upon that graph, it gives the result of whether the skin is vitiligo infected or not. The present approach is based on machine learning and annotated data increases the efficiency and reduce the false positive rate.

Vitiligo, Machine learning, PyTorch, OpenCV, melanocyte