A DEEP LEARNING APPROACH FOR THE AUTOMATED PROGNOSIS OF SKIN DISEASES

People may now more easily obtain current information thanks to the growth of mobile applications. Users are looking for solutions to problems in the virtual world, especially medical issues. More people have skin conditions than other illnesses. Skin conditions can be brought on by viruses, germs, allergies, fungal infections, and more. A skin condition can alter the skin’s tone or texture. Skin conditions are typically persistent, contagious, and occasionally carcinogenic. This research discusses the image-based skin disease diagnosis application for mobile devices. The collection of photos of unhealthy skin is used by the algorithm to examine the ailment. This technique is intended to identify skin conditions from unwholesome photos. By comparing pre processed photos, the threshold value difference is found. The differential in the given threshold will be used in decision making when suspicious unpleasant skin is detected. The app was created using the Android Platform and the OpenCV library to implement the Machine learning algorithm. Android-based mobile applications that can diagnose skin infections have indeed been extensively developed. The aim is to introduce a deep learning approach for the automated prognosis of skin diseases, leveraging image analysis techniques. The methodology integrates diverse deep learning strategies to enhance the accuracy and efficiency of skin disease prognosis. By combining various neural network architectures, including Densenet and VGG achieves a comprehensive analysis of medical images. The incorporation of advanced image analysis techniques aims to provide a robust and reliable prognosis system. The effectiveness of the proposed deep learning approach is evaluated through rigorous experimentation, demonstrating its potential as an innovative solution for automated skin disease prognosis based on image data.