Proceedings of the First International Conference on Advanced Scientific Innovation in Science, Engineering and Technology, ICASISET 2020, 16-17 May 2020, Chennai, India

Research Article

DETECTION OF SKIN DISEASE IN PERVASIVE HEALTH CARE USING CONVOLUTION NEURAL NETWORK

Download409 downloads
  • @INPROCEEDINGS{10.4108/eai.16-5-2020.2304021,
        author={Sharon R.K. Sarojini and Sruthy  Simon and Rajalakshmi Shenbaga Moorthy and P.  Pabitha},
        title={DETECTION OF SKIN DISEASE IN PERVASIVE HEALTH CARE USING CONVOLUTION NEURAL NETWORK},
        proceedings={Proceedings of the First  International Conference on Advanced Scientific Innovation in Science, Engineering and Technology, ICASISET 2020, 16-17 May 2020, Chennai, India},
        publisher={EAI},
        proceedings_a={ICASISET},
        year={2021},
        month={1},
        keywords={cnn keras pervasive healthcare},
        doi={10.4108/eai.16-5-2020.2304021}
    }
    
  • Sharon R.K. Sarojini
    Sruthy Simon
    Rajalakshmi Shenbaga Moorthy
    P. Pabitha
    Year: 2021
    DETECTION OF SKIN DISEASE IN PERVASIVE HEALTH CARE USING CONVOLUTION NEURAL NETWORK
    ICASISET
    EAI
    DOI: 10.4108/eai.16-5-2020.2304021
Sharon R.K. Sarojini1,*, Sruthy Simon1, Rajalakshmi Shenbaga Moorthy1, P. Pabitha2
  • 1: St.Joseph’s Institute of Technology
  • 2: Madras Institute of Technology
*Contact email: sharon110699@gmail.com

Abstract

This paper is about detecting skin disease using convolutional neural network in pervasive health care. Various optimizers in keras model are compared and the optimizer with the highest accuracy percentage is employed to predict skin disease through abnormalities in skin images. Detecting skin diseases by viewing images of skin can be an advancement to a great extent. There are many algorithms in machine language which would help to serve the above mentioned scenario. One of the most efficient algorithms that is being used here is convolutional neural networks. It simplifies the input image by reducing its dimensionality which makes prediction of patterns much easier. Convolutional neural network is used along with keras using tensor flow as backend which enables higher efficiency and accuracy.