Object Detection or Segmentation With Tracking and Counting

Table of Contents

• Features
• Installation

Features

Outline the key features of your project:

• Object detection using YOLOv8
• Object segmentation using YOLOv8
• Object tracking using DeepSORT
• Tracking with IDs and Trail
• Object Counting

Installation and Usage

To use the Coccidiosis Detection Web App, follow these steps:

1. Clone the Repository:

   git clone https://github.com/junaid-o/CV-DL-12_Object_Segmentation_and_Tracking.git
   cd CV-DL-12_Object_Segmentation_and_Tracking

2. Create a conda environment after opening the repository

   conda create --prefix ./envYOLO8 python=3.10.12 -y

   conda activate ./envYOLO8

3. Install Dependencies:

   pip install -e '.[dev]'

   pip install easydict

4. Test File Location:

   Keep all your test video files inside input_data_USER/test

5. Run From Terminal:

• When running without any argument it will take dafault values for model, test files etc

  python run_Yolo_DeepSORT_tracking.py 

• For help

  python run_Yolo_DeepSORT_tracking.py -h

  OPTIONS AVAILABLE FOR ARGUMENTS

  $ python run_Yolo_DeepSORT_tracking.py -h
  usage: run_Yolo_DeepSORT_tracking.py [-h] [-task TASK] [-ds DETECTION_SCRIPT] [-ss SEGMENT_SCRIPT]
                                      [-dm DETECTION_MODEL] [-sm SEGMENTATION_MODEL] [-t TEST_FILE] [-w WORKING_DIR]
  
  Run segment prediction script.
  
  options:
  -h, --help            show this help message and exit
  -task TASK, --task TASK
                          Task to be performed (Detection+DeepSORT Tracking ID + TRAIL) or (Segmentation + DeepSORT
                          Tracking ID + TRAIL)
  -ds DETECTION_SCRIPT, --detection-script DETECTION_SCRIPT
                          Path to predict.py script for detection with DeepSORT
  -ss SEGMENT_SCRIPT, --segment-script SEGMENT_SCRIPT
                          Path to predict.py script
  -dm DETECTION_MODEL, --detection-model DETECTION_MODEL
                          Path to detection model
  -sm SEGMENTATION_MODEL, --segmentation-model SEGMENTATION_MODEL
                          Path to segmentation model
  -t TEST_FILE, --test-file TEST_FILE
                          Path to test video file
  -w WORKING_DIR, --working-dir WORKING_DIR
                          Working directory having all the files for of YOLOv8 for detection or segmentation along with
                          DeepSORT folder. In this repository this working dir points to ultralytics/yolo/v8/detect or
                          segment
  

6. View Output Video Files:

   Find output files in run folder