OpenCV-Python

Language: Python

ML/AI

OpenCV (Open Source Computer Vision Library) was initially developed by Intel in 1999. OpenCV-Python provides bindings to access the full power of OpenCV in Python, making it popular for real-time image and video processing, machine learning, and robotics applications.

OpenCV-Python is a Python wrapper for the OpenCV C++ library, providing tools for computer vision, image processing, and video analysis. It allows rapid prototyping and deployment of vision applications in Python.

Installation

pip: pip install opencv-python
conda: conda install -c conda-forge opencv

Usage

OpenCV-Python provides functions for image I/O, color space conversions, filtering, transformations, object detection, and machine learning integration. It supports reading and writing images and videos, as well as performing operations on arrays representing images.

Read and display an image

import cv2
img = cv2.imread('image.jpg')
cv2.imshow('Image', img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Reads an image from file and displays it in a window until any key is pressed.

Convert image to grayscale

import cv2
img = cv2.imread('image.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.imshow('Gray', gray)
cv2.waitKey(0)
cv2.destroyAllWindows()

Converts a color image to grayscale using OpenCV’s color conversion function.

Edge detection using Canny

import cv2
img = cv2.imread('image.jpg', 0)
edges = cv2.Canny(img, 100, 200)
cv2.imshow('Edges', edges)
cv2.waitKey(0)
cv2.destroyAllWindows()

Detects edges in a grayscale image using the Canny algorithm.

Video capture from webcam

import cv2
cap = cv2.VideoCapture(0)
while True:
    ret, frame = cap.read()
    cv2.imshow('Webcam', frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
cv2.destroyAllWindows()

Captures video from the default webcam and displays it in real-time.

Drawing shapes

import cv2
import numpy as np
img = np.zeros((512,512,3), np.uint8)
cv2.line(img, (0,0), (511,511), (255,0,0), 5)
cv2.rectangle(img, (100,100), (300,300), (0,255,0), 3)
cv2.circle(img, (256,256), 50, (0,0,255), -1)
cv2.imshow('Shapes', img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Draws lines, rectangles, and circles on an empty image using OpenCV drawing functions.

Face detection using Haar cascades

import cv2
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
img = cv2.imread('group.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
for (x,y,w,h) in faces:
    cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
cv2.imshow('Faces', img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Detects faces in an image using pre-trained Haar cascade classifiers.

Error Handling

cv2.error: OpenCV(…): error: (-215:Assertion failed): Check that image paths are correct and files exist before reading.
AttributeError: module 'cv2' has no attribute '…': Ensure that the correct OpenCV version is installed; some functions may require opencv-contrib-python.

Best Practices

Use NumPy arrays for image processing to leverage vectorized operations.

Release video capture objects and destroy windows to avoid resource leaks.

Use proper color conversion codes when changing image color spaces.

Profile and optimize processing pipelines for real-time applications.

Keep pre-trained classifiers and models organized for reuse.

Official DocsGithub