Python OpenCV Hand Gesture Recognition Trick

Hand gesture recognition is an exciting computer vision application that allows interaction with devices using hand movements. With OpenCV, we can create a simple yet effective hand gesture recognition system.

Step 1: Install OpenCV and Mediapipe

Ensure the required libraries are installed:

pip install opencv-python mediapipe numpy

Step 2: Import Libraries and Initialize Mediapipe

Mediapipe is a powerful library for real-time hand tracking.

import cv2

import mediapipe as mp

 mp_hands = mp.solutions.hands

mp_draw = mp.solutions.drawing_utils

hands = mp_hands.Hands(min_detection_confidence=0.7, min_tracking_confidence=0.7)

Step 3: Capture Video Feed

Open a video stream to detect hands in real time:

cap = cv2.VideoCapture(0)

 while cap.isOpened():

    ret, frame = cap.read()

    if not ret:

        break

    frame = cv2.flip(frame, 1)  # Flip for mirror effect

    rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

    results = hands.process(rgb_frame)

       if results.multi_hand_landmarks:

 for hand_landmarks in results.multi_hand_landmarks:

       mp_draw.draw_landmarks(frame, hand_landmarks, mp_hands.HAND_CONNECTIONS)

     cv2.imshow(“Hand Gesture Recognition”, frame)

    if cv2.waitKey(1) & 0xFF == ord(‘q’):

        break

 cap.release()

cv2.destroyAllWindows()

Step 4: Recognizing Specific Gestures

By analyzing landmark positions, we can classify different gestures. Here’s an example of recognizing an open palm:

def is_open_palm(hand_landmarks):

    thumb_tip = hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_TIP].y

    index_tip = hand_landmarks.landmark[mp_hands.HandLandmark.INDEX_FINGER_TIP].y

    middle_tip = hand_landmarks.landmark[mp_hands.HandLandmark.MIDDLE_FINGER_TIP].y

    ring_tip = hand_landmarks.landmark[mp_hands.HandLandmark.RING_FINGER_TIP].y

    pinky_tip = hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_TIP].y

     return (index_tip < thumb_tip and middle_tip < thumb_tip and

            ring_tip < thumb_tip and pinky_tip < thumb_tip)

Modify the video loop to check for gestures:

if results.multi_hand_landmarks:

    for hand_landmarks in results.multi_hand_landmarks:

mp_draw.draw_landmarks(frame, hand_landmarks, mp_hands.HAND_CONNECTIONS)

        if is_open_palm(hand_landmarks):

            cv2.putText(frame, “Open Palm Detected”, (50, 50),

cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)

Conclusion

Using OpenCV and Mediapipe, we can recognize hand gestures in real-time and map them to actions. Expand this by adding gesture-based commands for controlling applications, games, or IoT devices!