Automated Melanoma Detection Using Haar Wavelet-Based Neural Networks

Melanoma is one of the most dangerous forms of skin cancer, with high potential for metastasis if not detected and treated early. Traditional diagnostic methods rely heavily on visual examination and dermoscopic analysis, which can be subjective and prone to errors. In this paper, we propose an automated melanoma detection system using Haar wavelet-based feature extraction combined with neural networks for classification. The Haar wavelet transform is employed to extract critical features from dermoscopic images, such as texture and edge details, which are crucial in distinguishing between benign and malignant lesions. These features are then used to train a MultiLayer Perceptron (MLP) neural network for accurate classification. Experimental results demonstrate that Levenberg-Marquardt (LM) algorithm achieves high accuracy, sensitivity, and specificity, effectively reducing false positives and negatives. This system presents a reliable and efficient solution for early melanoma detection, which can assist dermatologists in clinical decision-making and potentially improve patient outcomes.