Skin Disease Recognition using AutoML and CNN
Skin diseases seriously affect individual’s life and health. Current research proposes an effective way to dealing with and recognizing only singular type of skin diseases. It is important to create programmed techniques to expand the precision of analysis for multi-type skin diseases. In this model skin disease could be recognized by new recognition technique. Skin images were pre-processed to remove noise and irrelevant background by filtering and transformation. The texture and color features of various skin infection pictures could be acquired precisely. At the end, using the AutoML model and CNN model skin infections were distinguished and results are compared. The test results exhibit the adequacy and attainability of the proposed technique.