OpenCV Python Capture Video from Camera Tutorial: Webcam Streaming Guide

OpenCV Python – Capture Video from Camera

Video capture is one of the most important features in OpenCV. It allows you to access a webcam or camera and process live video streams in real time.

In this tutorial, you will learn how to capture video from a camera using OpenCV Python.

---

1. What is Video Capture?

Video capture means reading frames continuously from a camera or video device.

It is used for:

• Face detection
• Object tracking
• Motion detection
• Real-time AI applications

---

2. Import OpenCV

import cv2

---

3. Access Webcam

Syntax:

cv2.VideoCapture(index)

• 0 → Default webcam
• 1 → External camera

---

Example:

cap = cv2.VideoCapture(0)

---

4. Read Video Frames

while True:
    ret, frame = cap.read()

    cv2.imshow("Camera", frame)

    if cv2.waitKey(1) & 0xFF == 27:  # ESC key
        break

cap.release()
cv2.destroyAllWindows()

---

5. Check Camera Status

if not cap.isOpened():
    print("Cannot access camera")

---

6. Change Frame Properties

You can modify camera settings:

cap.set(cv2.CAP_PROP_FRAME_WIDTH, 640)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 480)

---

7. Convert Video to Grayscale

while True:
    ret, frame = cap.read()

    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    cv2.imshow("Grayscale Video", gray)

    if cv2.waitKey(1) & 0xFF == 27:
        break

---

8. Why Video Capture is Important

Video capture is used in:

• Face recognition systems
• Security surveillance
• Motion tracking
• AI-based camera apps
• Real-time object detection

---

9. Common Mistakes

❌ Camera not opening

✔ Solution:

• Try different index: 0, 1, 2

---

❌ Blank video window

✔ Solution:

• Check cap.read() return value

---

10. Conclusion

Capturing video from a camera in OpenCV Python is the foundation of real-time computer vision applications. It allows you to process live frames for AI, detection, and tracking systems.

Once you master video capture, you can move to advanced topics like video recording and object tracking.