Vision-Based Skin Disease Identification Using Deep Learning
R. Bhavani1, V. Prakash2, R.V Kumaresh3, R .Sundra Srinivasan4
1R.Bhavani, Assistant Professor, Department of Computer Science, Sastra Deemed To Be University, Kumakonam, (Tamil Nadu), India.
2V. Prakash,  Assistant Professor, Department of  MCA,  Sastra Deemed To, Be University, Thanjavur, (Tamil Nadu), India.
3R.V Kumaresh,  B. Tech CSE, Department of CSE, Sastra Deemed To Be University, Kumakonam, (Tamil Nadu), India.
4R.Sundra Srinivasan,  B. Tech CSE, Department of CSE, Sastra Deemed To Be University, Kumakonam, (Tamil Nadu), India.

Manuscript received on February 06, 2019. | Revised Manuscript received on February 14, 2019. | Manuscript published on August 30, 2019. | PP: 3784-3788 | Volume-8 Issue-6, August 2019. | Retrieval Number: F9391088619/19©BEIESP | DOI: 10.35940/ijeat.F9391.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: Skin disease is the most common health problems worldwide. Human skin is one of the difficult areas topr edict. The difficulty is due to rough areas, irregular skin tones, various factors like burns, moles. We have to identify the diseases excluding these factors. In a developing country like India, it is expensive for a large number of people to go to the dermatologist for their skin disease problem. Every year a large number of population in developing countries like India suffer due to different types of skin diseases. So the need for automatic skin disease prediction is increasing for the patients and as well as the dermatologist. In this paper, a method is proposed that uses computer vision-based techniques to detectvarious kinds of dermatological skin diseases. Inception_v3, Mobilenet, Resnetare three deep learning algorithms used for feature extraction in a medical image and machine learning algorithm namely Logistic Regression is used for training and testing the medical images. Using the combined architecture of the three convolutional neural networks considerable efficiency can be achieved.
Keywords: Convolutional neural networks, Inception_v3, Mobilenet, Resnet, Logistic Regression.