Wrong posture detection with OpenCV and Support vector machine

EasyChair Preprint 2369

Abstract

Since the technological revolution, the health of normal human being has tainted a lot. Over the past ten years, there is a heavy demise in the health (both physical and mental health) of the people. The life expectancy rate has seen a downfall and has gone below 65 years of age. To sustain this life expectancy rate and to make people aware about their physical fitness, we have tried to design a process with the help of OpenCV, machine learning and deep learning algorithm to detect wrong postures and subsequently asking the user not to be in that particular position which can harm the body. For the entire process, we have used OpenCV for image and video processing, support vector machine (SVM), recurrent convolutional neural network(R-CNN) (proposed) to determine the wrong posture by ensuing training the dataset and testing the unknown input images. We have also provided analysis for the physical health of people of India (region wise) over the past few years.

Keyphrases: Life-expectancy-rate, Normal-Weight, OpenCV, Predictive Analysis, RCNN, Recurrent Convolutional Neural Network, SVM, Support Vector Machine, bmi analysis, clinical anthropometric biochemical cab, deep learning, postural deformity, revised consensus body mass, survey data district, underweight male, wrong posture

@booklet{EasyChair:2369,
  author    = {Piyush Keshari},
  title     = {Wrong posture detection with OpenCV and Support vector machine},
  howpublished = {EasyChair Preprint 2369},
  year      = {EasyChair, 2020}}