Sitting Posture Recognition Based on OpenPose

Abstract

Abstract Sedentary and poor sitting posture can damage the health of adolescents. Therefore, it is very practical to effectively detect the sitting posture of students in the classroom and to warn the bad sitting posture. This paper proposed an in-class student sitting posture recognition system based on OpenPose, which uses the monitor in the classroom to detect the sitting posture of the students, and uses OpenPose to extract the posture feature. Keras deep learning framework is used to construct the convolutional neural network, which is used to train the datasets and recognize sitting posture of students. Experiments show that the accuracy is more than 90% after 100 epoch training.