Description of the project

This project aims to detect and track shoulder movements, particularly focusing on the action of shoulder shrugging, in videos depicting natural upper body movements. 

1.Video Capture: Users are instructed to provide a video demonstrating natural upper body movements with 
  clear visibility of the head, shoulders, and upper body against a smooth, uncluttered background.

2.Head Position Detection: The head position is detected in each frame of the video using established 
  methods such as Viola-Jones or Convolutional Neural Networks (CNNs) for face detection. This serves 
  as a reference point for subsequent steps.

3.Shoulder Line Detection and Tracking: Shoulder lines are detected and tracked in the video, with options
  for tracking each shoulder independently or both shoulders together. The goal is to ensure robustness in 
  detecting shoulder movements.

4.Shoulder Shrugging Detection: The system identifies shoulder shrugging events throughout the video, ranging
  from subtle to exaggerated movements. These events are clearly indicated in the corresponding frames through 
  text superimposed on the frames or changes in the color of the calculated shoulder lines.

By combining these steps, the project provides a comprehensive solution for analyzing and understanding shoulder 
movements in videos, enabling applications in areas such as gesture recognition, motion analysis, and human-computer interaction.