/Human-Body-Measurements-using-Computer-Vision

Anthropometric measurement extraction using single image

Primary LanguagePython

Human body measurement using computer vision/ 3D modeling

Anthropometric measurement extraction

alt text

Getting human body measurements from image is a very hard problem. This repository provides a starting solution for any one who is working in this domain. It maps single input image onto 3-dimientional model using 3D human reconstruction and then extracts body measurements such as waist, chest etc. All meausrements are in centimeters. 3D reconstruction is done using HMR. Tested on tensorflow==1.13.1.

alt text

Inference

  1. Download the pre-trained models. Type the following command on terminal

wget https://people.eecs.berkeley.edu/~kanazawa/cachedir/hmr/models.tar.gz && tar -xf models.tar.gz

and save it in 'models' folder.

  1. Download CustomBodyPoints text file and place it in data(make new one) folder.

  2. Install packages pip install -r requirements.txt or pip3 install -r requirements.txt

  3. Run inference python3 inference.py -i <path to Image1> -ht <height in cm>

Acknowledgment

HMR

Deep lab v3 +

Humanbody shape