POSTURE CORRECTION APPLICATION USING DEEP LEARNING

Abstract

Many people in modern society suffer from muscle discomfort in the neck and shoulders, which is mostly caused by poor posture. As the use of digital devices becomes more common, the number of patients suffering from neck pain is that. The number of patients could be reduced if patients were told in real time about how inappropriate their postures are.  sitting posture of seated workers which will lead to chronic pain , sclerosis , kyphosis , skin and respiratory problems , loss of brain cells and many more physical problems. Different Sitting posture datas such as erect-sitting , forward inclination, left or right leg crossed , leaning left or right etc where collected using camera by capturing it and the collected data are pre-processed before they are trained by COCO algorithm. The processed data are made to classify depending on different weights of the proper and improper sitting posture image. In this system , we use Convolutional Neural Network (CNN) algorithm to train the processed data to give accurate deviation of angle from correct posture. This real-time posture detection will respond to the poor posture of the human body which will reduce major physical health issues .Has the result of the system which warns the user with problem that may occur in future due to incorrect posture and also suggestion of correct posture.