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

Human Body Part Detection and External Injury Prediction Using Convolutional Neural Network

Download323 downloads
  • @INPROCEEDINGS{10.4108/eai.16-5-2020.2303952,
        author={Aliasgar  Haji and Riya  Saraf and Dipti  Pawade and Ashwini  Dalvi and Irfan  Siddavatam},
        title={Human Body Part Detection and External Injury Prediction Using Convolutional 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={body part detection injury detection convolutional neural network (cnn)},
        doi={10.4108/eai.16-5-2020.2303952}
    }
    
  • Aliasgar Haji
    Riya Saraf
    Dipti Pawade
    Ashwini Dalvi
    Irfan Siddavatam
    Year: 2021
    Human Body Part Detection and External Injury Prediction Using Convolutional Neural Network
    ICASISET
    EAI
    DOI: 10.4108/eai.16-5-2020.2303952
Aliasgar Haji1,*, Riya Saraf1, Dipti Pawade1, Ashwini Dalvi1, Irfan Siddavatam1
  • 1: Department of Information Technology, K. J. Somaiya College of Engineering, Vidyavihar, Mumbai, India
*Contact email: mustafa.aliasgar@somaiya.edu

Abstract

Fake callers continue to disrupt emergency ambulance services in the state with its call center registering 7 percent fake calls every day on an average. This is a growing problem and it needs to be curbed at the earliest as it not only wastes the time of the operators but also keeps the line busy hence causing a delay in emergency services which may even result in the death of the victim. Hence, we propose a solution to validate whether the request for ambulance services is genuine or not. The main aim is to detect and identity from an image whether a human body part is present or not. Even if the image does not contain the entire human in any particular pose and only a part of his/her body is present. We also detect any visible external injury on the body part. This would be a proof of concept in the form of a machine learning model that can successfully detect feet, face, hands and any external injury present in the image provided to it.