pip install mediapipe opencv-python numpy scikit-learn

import cv2
import mediapipe as mp

mp_hands = mp.solutions.hands
hands = mp_hands.Hands()
mp_draw = mp.solutions.drawing_utils

cap = cv2.VideoCapture(0)

while True:
    success, frame = cap.read()
    if not success:
        break
    rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    result = hands.process(rgb_frame)
    if result.multi_hand_landmarks:
        for hand_landmarks in result.multi_hand_landmarks:
            mp_draw.draw_landmarks(frame, hand_landmarks, mp_hands.HAND_CONNECTIONS)
    cv2.imshow("Hand Tracking", frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

import numpy as np

data = []
labels = []

gesture_name = input("Nhập tên cử chỉ (ví dụ: thumbs_up): ")

while True:
    success, frame = cap.read()
    if not success:
        break
    rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    result = hands.process(rgb_frame)
    if result.multi_hand_landmarks:
        for hand_landmarks in result.multi_hand_landmarks:
            landmarks = []
            for lm in hand_landmarks.landmark:
                landmarks.extend([lm.x, lm.y, lm.z])
            data.append(landmarks)
            labels.append(gesture_name)
    cv2.imshow("Thu Thập Dữ Liệu", frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

np.save('gesture_data.npy', np.array(data))
np.save('gesture_labels.npy', np.array(labels))

from sklearn.neighbors import KNeighborsClassifier

X = np.load('gesture_data.npy')
y = np.load('gesture_labels.npy')

knn = KNeighborsClassifier(n_neighbors=3)
knn.fit(X, y)

while True:
    success, frame = cap.read()
    if not success:
        break
    rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    result = hands.process(rgb_frame)
    if result.multi_hand_landmarks:
        for hand_landmarks in result.multi_hand_landmarks:
            landmarks = []
            for lm in hand_landmarks.landmark:
                landmarks.extend([lm.x, lm.y, lm.z])
            prediction = knn.predict([landmarks])
            cv2.putText(frame, f'Cử chỉ: {prediction[0]}', (10, 50),
                        cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
    cv2.imshow("Nhận Diện Cử Chỉ", frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()