Employee Monitoring System

Overview

This Employee Monitoring System is a video processing application that performs real-time analysis of employee behavior in a workplace setting. It uses computer vision and machine learning techniques to identify employees, verify dress code compliance, and detect greeting behaviors.

Demo

Employee.monitor.demo.mp4

System Architecture

┌─────────────────────────────────────────────────────────────────────────┐
│                             Main Thread                                 │
│                                                                         │
│  ┌───────────────────┐        ┌───────────────────┐                     │
│  │  Initialize       │        │ Create Queues &   │                     │
│  │  Embedding Index  │───────▶│  Start Threads    │                     │
│  └───────────────────┘        └───────────────────┘                     │
│                                   │                                     │
└───────────────────────────────────┼─────────────────────────────────────┘
                                    │
                 ┌──────────────────┴───────────────────┐
                 │                                      │
                 ▼                                      ▼
┌────────────────────────────────┐      ┌────────────────────────────────┐
│    Human Detection Thread      │      │  Face Verification, Dress &    │
│                                │      │      Greeting Thread           │
│  ┌───────────────────────────┐ │      │  ┌───────────────────────────┐ │
│  │ Initialize Model & Video  │ │      │  │   Get Frame from Queue    │ │
│  └───────────────────────────┘ │      │  └───────────────────────────┘ │
│              │                 │      │              │                 │
│              ▼                 │      │              ▼                 │
│  ┌───────────────────────────┐ │      │  ┌───────────────────────────┐ │
│  │    Process Video Frames   │ │      │  │      Process Boxes        │ │
│  └───────────────────────────┘ │      │  └───────────────────────────┘ │
│              │                 │      │              │                 │
│              ▼                 │      │              ▼                 │
│  ┌───────────────────────────┐ │      │  ┌───────────────────────────┐ │
│  │  Put Results in Queue     │ │      │  │    Face Verification      │ │
│  └───────────────────────────┘ │      │  └───────────────────────────┘ │
│              │                 │      │              │                 │
│              ▼                 │      │              ▼                 │
│  ┌───────────────────────────┐ │      │  ┌───────────────────────────┐ │
│  │   Plot Results on Frame   │ │      │  │    Dress Verification     │ │
│  └───────────────────────────┘ │      │  └───────────────────────────┘ │
│              │                 │      │              │                 │
│              ▼                 │      │              ▼                 │
│  ┌───────────────────────────┐ │      │  ┌───────────────────────────┐ │
│  │      Display Frame        │ │      │  │    Greeting Detection     │ │
│  └───────────────────────────┘ │      │  └───────────────────────────┘ │
│                                │      │              │                 │
└────────────────────────────────┘      │              ▼                 │
               ▲                        │  ┌───────────────────────────┐ │
               │                        │  │  Update Identify Queue    │ │
               │                        │  └───────────────────────────┘ │
               │                        └────────────────────────────────┘
               │                                       │
               │                                       │
               └───────────────────────────────────────┘

Features

• Human detection in video streams
• Face verification against a database of known employees
• Dress code compliance verification
• Greeting behavior detection
• Real-time tracking of individuals across video frames
• Event logging with timestamps
• Live visualization of processed video frames

Prerequisites

• Python 3.7+
• OpenCV
• PyTorch
• Ultralytics YOLOv8

Installation

1. Clone the repository:

   git clone https://github.com/vTuanpham/Employee_monitor.git
   cd Employee_monitor
   

2. Install the required packages:

   bash setup.sh
   

Usage

Run the main script:

python monitor_pipe/pipe.py

By default, the script will use the video file specified in VIDEO_FILE_1. To use a different video file, modify the VIDEO_FILE_1 constant in the script. To use Webcam, change VIDEO_FILE_1 to 0

Configuration

The following constants can be adjusted in the script:

• VIDEO_FILE_1: Path to the input video file, can be set to 0 for Webcam
• MODEL_FILE: Path to the YOLOv8 model file
• DEVICE_ID: GPU device ID
• confidence_threshold: Threshold for face verification confidence (default: 0.65)
• max_age_seconds: Maximum age of a track ID in seconds (default: 10)
• cooldown_seconds: Cooldown period for greeting detection in seconds (default: 5)

Logging

The system logs important events to identification_log.txt. This includes:

• Employee identifications
• Dress code compliance
• Greeting detections

Disclaimer

This system is designed for educational and research purposes. Ensure compliance with all applicable laws and regulations regarding workplace monitoring and data privacy before deploying in a real-world setting.