hand_gesturer_ecognition/gesture_recognition.py

import warnings
import threading
import cv2
import mediapipe as mp
import numpy as np
from tensorflow.keras.models import load_model
from tkinter import Tk, Canvas, Button, Label, LEFT, RIGHT, NW
from PIL import Image, ImageTk
import os

# 设置环境变量以关闭oneDNN自定义操作
os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'

warnings.filterwarnings("ignore", category=UserWarning, message='SymbolDatabase.GetPrototype() is deprecated')

# 初始化全局变量
hands = None
mp_draw = mp.solutions.drawing_utils
cap = None
keep_running = False
paused = False
popup_open = False  # 用于标记当前是否有弹窗打开

# 模型路径和加载
model_path = 'D:/hand/hand_gesture_model.h5'
model = load_model(model_path)

gesture_classes = ['00', '01', '02', '03', '04', '05', '06', '07', '08', '09']

def start_recognition(root, callback=None):
    global keep_running, cap, hands
    if cap is None or not cap.isOpened():
        cap = cv2.VideoCapture(0)
    if hands is None:
        hands = mp.solutions.hands.Hands(static_image_mode=False, max_num_hands=2,
                                         model_complexity=1, min_detection_confidence=0.5,
                                         min_tracking_confidence=0.5)
    keep_running = True
    threading.Thread(target=run_recognition, args=(root, callback)).start()

def run_recognition(root, callback=None):
    global keep_running, paused

    while keep_running and cap.isOpened():
        ret, img = cap.read()
        if not ret:
            break

        img = cv2.flip(img, 1)
        img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

        if not paused:
            results = hands.process(img_rgb)
            total_raised_fingers = 0

            if results.multi_hand_landmarks:
                for handLms in results.multi_hand_landmarks:
                    mp_draw.draw_landmarks(img_rgb, handLms, mp.solutions.hands.HAND_CONNECTIONS)
                    gesture, raised_fingers = detect_gesture_and_fingers(handLms)
                    total_raised_fingers += raised_fingers

            if total_raised_fingers > 0:
                handle_finger_detection(total_raised_fingers)

            cv2.putText(img_rgb, f'Total Raised Fingers: {total_raised_fingers}', (10, 30),
                        cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 0), 2, cv2.LINE_AA)

        if callback:
            root.after(0, callback, img_rgb)

def stop_recognition():
    global keep_running, cap
    keep_running = False
    if cap is not None and cap.isOpened():
        cap.release()
        cap = None
    cv2.destroyAllWindows()

def release_camera():
    global cap
    if cap is not None and cap.isOpened():
        cap.release()
        cap = None

def detect_gesture_and_fingers(hand_landmarks):
    gesture_image = get_hand_image(hand_landmarks)
    gesture = predict_gesture(gesture_image)

    raised_fingers = count_raised_fingers(hand_landmarks)

    return gesture, raised_fingers

def get_hand_image(hand_landmarks):
    img = np.zeros((150, 150, 3), dtype=np.uint8)
    return img

def predict_gesture(img):
    img = cv2.resize(img, (150, 150))
    img_array = np.expand_dims(img, axis=0) / 255.0
    predictions = model.predict(img_array)
    predicted_class = gesture_classes[np.argmax(predictions)]
    return predicted_class

def count_raised_fingers(hand_landmarks):
    fingers_status = [0, 0, 0, 0, 0]

    thumb_tip = hand_landmarks.landmark[mp.solutions.hands.HandLandmark.THUMB_TIP]
    thumb_ip = hand_landmarks.landmark[mp.solutions.hands.HandLandmark.THUMB_IP]
    thumb_mcp = hand_landmarks.landmark[mp.solutions.hands.HandLandmark.THUMB_MCP]
    thumb_cmc = hand_landmarks.landmark[mp.solutions.hands.HandLandmark.THUMB_CMC]

    angle_thumb = calculate_angle(thumb_cmc, thumb_mcp, thumb_tip)
    if angle_thumb > 160:
        fingers_status[0] = 1

    for i, finger_tip_id in enumerate([mp.solutions.hands.HandLandmark.INDEX_FINGER_TIP,
                                       mp.solutions.hands.HandLandmark.MIDDLE_FINGER_TIP,
                                       mp.solutions.hands.HandLandmark.RING_FINGER_TIP,
                                       mp.solutions.hands.HandLandmark.PINKY_TIP]):
        finger_tip = hand_landmarks.landmark[finger_tip_id]
        finger_pip = hand_landmarks.landmark[finger_tip_id - 2]
        finger_mcp = hand_landmarks.landmark[finger_tip_id - 3]

        angle_finger = calculate_angle(finger_mcp, finger_pip, finger_tip)
        if angle_finger > 160:
            fingers_status[i + 1] = 1

    return sum(fingers_status)

def calculate_angle(point1, point2, point3):
    angle = np.arctan2(point3.y - point2.y, point3.x - point2.x) - np.arctan2(point1.y - point2.y, point1.x - point2.x)
    angle = np.abs(angle)
    if angle > np.pi:
        angle = 2 * np.pi - angle
    return angle * 180 / np.pi

def handle_finger_detection(finger_count):
    global paused, popup_open
    if not popup_open:
        if finger_count == 5:
            paused = True
            popup_open = True
            show_stop_recognition_window()
        # if finger_count == 1:
        #     paused = True
        #     popup_open = True
        #     show_stop_recognition_window()

def show_finger_window(message):
    def on_continue():
        global paused, popup_open
        paused = False
        popup_open = False  # 关闭弹窗后将标志设置为False
        finger_window.destroy()
        start_recognition(show_frame)

    finger_window = Tk()
    finger_window.title("手指检测")

    label = Label(finger_window, text=message, font=('Helvetica', 24, 'bold'))
    label.pack(pady=20)

    continue_button = Button(finger_window, text="继续识别", command=on_continue)
    continue_button.pack(pady=10)

    finger_window.protocol("WM_DELETE_WINDOW", on_continue)
    finger_window.mainloop()

def show_stop_recognition_window():
    def on_continue():
        global paused, popup_open
        paused = False
        popup_open = False  # 关闭弹窗后将标志设置为False
        stop_window.destroy()
        start_recognition(show_frame)

    def on_stop():
        global popup_open
        stop_recognition()
        popup_open = False  # 关闭弹窗后将标志设置为False
        stop_window.destroy()

    stop_window = Tk()
    stop_window.title("停止识别")

    label = Label(stop_window, text="您竖起了五根手指，是否停止识别？", font=('Helvetica', 24, 'bold'))
    label.pack(pady=20)

    continue_button = Button(stop_window, text="继续识别", command=on_continue)
    continue_button.pack(side=LEFT, padx=10, pady=10)

    stop_button = Button(stop_window, text="停止识别", command=on_stop)
    stop_button.pack(side=RIGHT, padx=10, pady=10)

    stop_window.protocol("WM_DELETE_WINDOW", on_continue)
    stop_window.mainloop()

def show_frame(img=None):
    global paused, canvas
    if keep_running and cap.isOpened():
        if img is not None:
            frame_rgb = img
        else:
            ret, frame = cap.read()
            if not ret:
                return

            frame = cv2.flip(frame, 1)
            frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

        img = Image.fromarray(frame_rgb)
        imgtk = ImageTk.PhotoImage(image=img)
        canvas.create_image(0, 0, anchor=NW, image=imgtk)
        canvas.image = imgtk

        if not paused:
            root.after(10, show_frame)
        else:
            root.update_idletasks()
            root.update()

if __name__ == "__main__":
    root = Tk()
    root.title("手势识别")

    canvas = Canvas(root, width=640, height=480)
    canvas.pack()

    start_button = Button(root, text="开始识别", command=lambda: start_recognition(show_frame))
    start_button.pack(side=LEFT, padx=10, pady=10)

    stop_button = Button(root, text="停止识别", command=stop_recognition)
    stop_button.pack(side=RIGHT, padx=10, pady=10)

    root.mainloop()