Object Detection using OpenCV

About the Project

Principle

• OpenCV (Open Source Computer Vision Library) is an open source computer vision and machine learning software library.
• The Program uses the WebCam to capture video until interrupted
• First Frame of the video is taken as reference and any movement after that is treated as an object
• Times of object entering and leaving the frame is stored and can be viewd later from a csv file
• Changes can be made in terms of video used for observation, extent of contouring for objects etc.

Dependencies

• Python 3.8.6
• OpenCV 4.4.0
• Pandas 1.1.5

Saving files

Using Git

• Open Terminal
• Traverse to the directory where you want to save project files
• Type git clone https://github.com/manankohlii/object-detection-opencv.git

Download from GitHub

• Go to Code in the Repository
• Download ZIP
• Unzip and save files

Setting up Virtual Environment

With Conda

• Open Terminal
• Traverse to the directory where files were saved
• Use conda create --name <env> --file requirements.txt

Executing the Project

Running the Program

• Open terminal and move to the directory where program files are stored
• Enter python capture.py
• Four windows appear with different tasks
• Press q after observation is complete

Viewing Results

• A Times.csv file has been created in the directory where program files are saved
• Open a the file in a Text Editor or MS Excel to view the results