Skin diseases are quite common and prevail in all age groups. Dermatological diseases such as acne and lesions not only affect the appearance of the skin but also have a greater impact on their social and mental well-being. In the contemporary age, skin disease diagnosis relies on manual inspection and observation by physicians. Moreover, not all people have the courage to visit the doctor until it gets worse and unbearable. In addition, identification of distinction between different skin illnesses requires a significant deal of experience. So, the classification of the disease by a doctor is not always completely right and may result in misdiagnosis. The proposed system deals with the classification of the skin disease once the image of the infected skin is given as input. This task is carried out using image processing and deep learning models namely ResNet50 and Inception V3. While image processing is an essential phase to improve the accuracy of diagnostic procedures, deep learning has made a significant contribution in the classification of disease using images. The Inception V3 model outperformed ResNet50 with a classification accuracy of 0.97. The results would aid in the diagnosis of dermatological problems and be helpful for both professionals and the public.