OpenCV-Python

This repository is a place where all the codes related to OpenCV is posted.

Topics

• Introduction to Images
• Installations
• Chapter 01: Read Images, Videos and Webcams
• Chapter 02: Basic Functions
• Chapter 03: Resize and Crop
• Chapter 04: Shapes and Text
• Chapter 05: Warp Perspective
• Chapter 06: Joining Images
• Chapter 07: Colour Detection
• Chapter 08: Contour / Shape Detection
• Chapter 09: Face Detection

Installations

• Install Python 3.7.6 -> It works well with OPENCV
• Install PyCharm -> A free comfortable IDE

Packages to install

• opencv-python --> This is the OPENCV Package
• numpy --> Used in arrays to analyse the images

Imported as

• import cv2
• import numpy as np

Introduction to Images

Consider a 2d-box of dimension (10 x 10). Draw equal lines such that each interior square is (1 x 1). Now there are 100 such small squares in the 2d-box.

Every shaded region (black) is denoted by a value 0. And every non-shaded region (white) is denoted by a value 1. In technical terms, each sqaure is a pixel.

Types of Images

• VGA -> 640 x 480
• HD -> 1280 x 720
• FHD -> 1920 x 1080
• 4K -> 3840 x 2160

Black and White Images

Generally 2^8 = 256 bit resolution is used.

• 0 is black
• 255 is white
• The rest are shades of gray

Those images are called 8 bits or 256 levels.

Coloured Images

Coloured images have 3 channels. They are RGB.

• RGB stands for Red Green Blue In OpenCV Convention, BGR term is used instead of RGB.

Thanks to Murtaza's Workshop - Robotics and AI