Android Application for Posture Analysis using Tensorflow and Computer Vision

Abstract

Human posture is an important body part that indicates the fundamental structure of the human body. Hence, the aim of this project is to develop an android application that can analyze on the human body posture by capturing an image of the body and provide guide to improve it. It is decided that using Microsoft Azure AI: Custom Vision and Tensorflow as the machine learning approaches for this project is reasonable. The development of this project used the Rapid Application Development (RAD) model which consists of the requirement planning that required the collection requirements for the project from the users. The user design that needed to build prototype for deciding system design, the construction phases that focus on coding and testing, while lastly the cutover phase that last test will be conducted, and the project will be published. The expected outcome of this project is that the application can conveniently help the people that needed to improve their body posture in a more simple and effective way. In the future, this project is hoped to be enhanced with more features and reliable functionalities that improve the effectiveness of improving human body posture.