엉금엉금 - Turtle Project 🐢

카메라 모듈을 이용해 사용자가 거북목인지 아닌지 체크하고 웹, 앱 어플리케이션을 통해 사용자에게 알림을 보내 주는 서비스입니다.

Installation

1. Download zip file
2. Download OpenPose zip file
3. Unpack the zip file and install the model from that file

cd openpose-master

for Linux 
./models/getModels.sh

for Windows
./models/getModels.bat

4. install opencv-python

pip install opencv-python

OpenPose

OpenPose has represented the first real-time multi-person system to jointly detect human body, hand, facial, and foot keypoints (total 135 keypoints) on single images.

Raspberry pi

take a picture of the user with the Raspberry Pi camera module.

Server

• Encode the photo and deliver it to the server in JSON format.
• The server decodes the received image and analyzes the photo.
• If a turtle neck is detected in the photo analysis results, the result is sent to the client.

Reference

OpenPose

• https://github.com/CMU-Perceptual-Computing-Lab/openpose

OpenPose를 이용해 이미지를 띄우고 분석하는 소스 참고

• https://hanryang1125.tistory.com/6