Automated Skin Lesion Analysis System for Melanoma using Feature extraction

Melanoma is the deadliest form of skin cancer can be treated successfully if it is detected at an early stage. Incidence rates of melanoma have increased, but the survival rates are high if detected early. In order to reduce the cost of dermatologists to screen each patient, there is a need for an automated melanoma screening system. Melanoma is a malignant pigmented skin lesion, and is currently the most dangerous existing cancers. Malignant melanoma and benign cases are different, is a difficult task, even for experienced specialists and a computer-aided diagnosis system can be a useful tool. Normally, the system starts by the image preprocessing, which means removing unwanted artifacts such as hair, freckles or shadow effects. Next, the system performs a segmentation step of identifying the lesion borders. Several features are identified finally, based on the range image as a lesion calculated and a classification is provided. Skin cancer is the most common form of cancer and represents 50 percent of all new cancer cases per year determined. The most deadly form of skin cancer is melanoma, and the frequency at a rate of 3 percent was rising annually. Due to monitor the cost of dermatologist every patient, there is a need for a computerized system a patient's risk of melanoma with pictures of their skin lesions taken using a standard digital camera to evaluate. In the proposed system, Gabor filters used for extracting features from the input medical image. It characteristics of an image with skin lesions, and an image, no skin injury, which are extracted.