International Journal of Scientific Research and Engineering Development

Skin diseases are the most common medical conditions across the world wide affecting millions of people.According to the surveys every year approximately 1.8-3 billion people are affected by the skin diseases.Skin diseases majorly like melanoma pose significant challenges in the field of dermatology.. Skin Cancer is a major global health concern , with melanoma being in dangerous form due to its rapid progression. In recent years Convolution Neural Networks(CNN’s) emerged as a powerful tool for Image recognition and lesion detection. Skin Disease Requires Early Diagnosis For Effective Treatment, the Traditional Approach is Time - Consuming ,To overcome this we’re using the Deep Learning techniques which helps in identifying various skin disease with high accuracy and high precision .This study presents Deep learning application for Multi-Class Dermatological Lesion identification and melanoma Images from HAM100000 dataset.For this Study we use a Deep Learning Convolution Neural Network Model The CNN Model is MobileNetV2 which is a Deep Learning Convolution Neural Network Which works on with few Parameters This CNN Model is well known for its For its performance and it is also a Lightweight CNN model, MobileNet is widely used for tasks like image classification , object detection, face recognition ,Augmented Reality , semantic segmentation mainly on mobile , edge devices. MobileNet uses Inverted Residual Blocks unlike traditional residual blocks it connects layers of the different depth and reduces the computational complexity and uses ReLU6 Activation Function and introduces Linear bottleneck between Layers which reduces the information loss and improves overall accuracy.In conclusion this research paper presents a comprehensive study on skin diseases like lesion identification, melanoma screening using the MobileNet CNN.The findings of this research contribute to improving the diagnosis, classification, and severity assessment of skin diseases, ultimately enhancing treatment outcomes.

Yeadluri Prasanthi, P.L.Adithya, S.Sathvik, M.Shanmukha, MD.Ameenulla,"Multi-Class Dermatological Lesion Identification and Melanoma Screening Using Lightweight MobileNet Convolutional Models" International Journal of Scientific Research and Engineering Development, V9(2): Page(501-505) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.