CLOTH SIZE PREDICTION SYSTEM USING COMPUTER VISION

Abstract

In this project we decided to make a size prediction system that would measure and predict the size of people based on camera footage, The reason for this is the growing amount of uneasiness between people due to the covid outbreak, policies and behavioural changes that came along with that, going to a store and asking someone to measure your size is very anxiety inducing for most people, not to mention the workers at the shop which would like to be covid free as well, secondly changing or try out rooms which were a common place to check clothing before buying have been deemed unusable due to being occupied by people, and anything that has to be shared has the highest chance of passing on covid. So, we decided to make a system that would measure size without needing any human intervention. We had multiple plans of action to achieve the means but we settled on using purely camera and computer vision due to the vast of amount of research that has already been done in detecting human bodies through computer vision, also the python libraries that help and facilitate machine learning and model creation are a blessing to have, working any kind of third party non-conventional system would make things harder and not on scalable. So, We decided to use computer vision to identify human bodies and measure them using distance equations, Once done we outputted the data to a excel sheet, from which we can read the data into a website and recommend clothes from there. In this endeavour we chose to make a size prediction framework that would gauge and foresee the size of individuals in light of camera perspective using Computer Vision