HumanPoseEstimation

This project is about a HPC(HumanPoseClassifier) implemented by Machine Learning and Deep Learning.

EXTEND

This program is extended to WEB UI

PREVIEW

FUNCTIONALITY

1. count squat,lunge,pushup
2. angle Feedback

Squat: Leg angle

Pushup: Elbow anple

Lunge : Leg angle and waist angle

3. MSE Feedback : compare with the expert pose (User must stand with 45 degree)

TREE

src : Main project source code

src/Posture_DeepLearning : the HPE implemented by DeepLearning(main_tf.py)

src/Posture_MachineLearning : the HPE implemented by MachineLearning(main.py)

custom : You can collect data, extract features, and also train by the custom data set

custom/Posture_CollectData : Collect custom data by cv2, extract features, and save the extracted features to .CSV

custom/Posture_Data : The extracted feature I used to train

custom/Posture_Train : You can train the data by ML and DL