Multi Object Detection and Tracking

This project implements a multi-object detection and tracking system using YOLOv5 for object detection and DeepSORT for tracking. The system is capable of detecting multiple objects in video frames and tracking their movements across frames.

Features

• Object detection using YOLOv8
• Multi-object tracking using DeepSORT
• Real-time FPS measurement and overlay on video frames

Setup and Installation

1. Clone the repository:

   git clone https://github.com/karpenet/DeepSORT-Multi-Object-Tracking
   cd multi-object-detection-tracking

2. Install dependencies:

   pip install -r requirements.txt

3. Download the YOLOv8 model:

   • Download the YOLOv8 model and place it in the models directory. You can use the provided script to convert the model to TensorRT format:

   python:models/utils/compress_model.py

4. Prepare the dataset:

   • Place your dataset in the dataset directory. Ensure the images are in the correct format and directory structure.

Running the Inference

1. Run the inference script:

   python inference.py

   This script will process the images in the dataset and perform object detection and tracking. The results will be saved in the output directory.

Results

The results of the object detection and tracking will be saved as images or videos in the output directory. You can visualize the results using any image or video viewer.

Logging

Logs are saved in the log.txt file. You can check this file for detailed information about the processing steps and any errors encountered.