Keras opencv -- 识别是否佩戴口罩

流程分析

OpenCV为Keras提供数据

以摄像头截图人脸为Keras提供训练集测试集数据
带口罩的图片存放在 imgs/yes/
没带口罩的图片存放在 imgs/no/
每种准备400张

OpenCV摄像头截图

def CatchPICFromVideo(catch_num): # catch_num = 400
    face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')  # haarcascade_frontalface_alt2
    camera = cv2.VideoCapture(0, cv2.CAP_DSHOW)
    num = 0
    while True:
        ret, frame = camera.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.1, 5)
        for (x, y, w, h) in faces:
            img_name = "imgs/yes/" + str(num) + ".jpg"
            image = frame[y:y + h, x:x + w]
            print(img_name)

			# 将图片处理成固定大小
            img_resize2 = cv2.resize(image, dsize=(200, 200))

            cv2.imwrite(img_name, image)
            num += 1
            if num > catch_num:
                break

            # 画出矩形框圈出人脸
            cv2.rectangle(frame, (x, y), (x + w, y + h), (255, 0, 0), 2)
            # 显示捕捉了多少张人脸
            font = cv2.FONT_HERSHEY_SIMPLEX
            cv2.putText(frame, f'num:{str(num)}', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
        if num > catch_num:
            break
        # 显示图像
        cv2.imshow('camera', frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    camera.release()
    cv2.destroyAllWindows()

Keras 卷积神经

使用Keras快速搭建神经网络

加载数据到内存

def load_data():
    images = []
    labels = []
    for i in range(500):
        img = cv2.imread("imgs/yes/" + str(i) + ".jpg")
        images.append(img)
        labels.append(1)
    for i in range(500):
        img = cv2.imread("imgs/no/" + str(i) + ".jpg")
        images.append(img)
        labels.append(0)
    images = np.array(images)
    labels = np.array(labels)
    return images, labels

卷积神经网络构建以及训练保存

def cnn_train(images, labels):
    # sklearn 划分数据集
    x_train, x_test, y_train, y_test = train_test_split(images, labels, test_size=0.3, random_state=3)

    # （400, 200, 200, 3） -> (400, 3, 200, 200) -- (图片数量，宽，高，通道数) -> (图片数量，通道数，宽，高)
    x_train = x_train.reshape(x_train.shape[0], 3, 200, 200)
    x_test = x_test.reshape(x_test.shape[0], 3, 200, 200)

    # 目标值 转化为 One-Hot编码 ： 2 -> [0, 0, 1];  0 -> [1, 0, 0]
    y_train = np_utils.to_categorical(y_train, num_classes=2)
    y_test = np_utils.to_categorical(y_test, num_classes=2)

    # 标准化
    x_train = x_train.astype('float32')
    x_test = x_test.astype('float32')

    x_train /= 255.0
    x_test /= 255.0

    # Keras 模型构建
    model = Sequential()
    
    model.add(Convolution2D(filters=32, kernel_size=[3, 3], padding='same', input_shape=(3, 200, 200)))
    model.add(Activation('relu'))

    model.add(Convolution2D(filters=32, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
    model.add(Dropout(0.25))

    model.add(Convolution2D(filters=64, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(Convolution2D(filters=64, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
    model.add(Dropout(0.25))

    model.add(Flatten())
    model.add(Dense(512))
    model.add(Activation('relu'))
    model.add(Dropout(0.25))
    model.add(Dense(2))
    model.add(Activation('softmax'))

    model.summary()

    # 优化器
    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
    
    # 训练
    model.fit(x_train, y_train, epochs=4, batch_size=100)

    # 查看效果
    loss, acc = model.evaluate(x_test, y_test)
    print("最终loss\n", loss)
    print("最终acc\n", acc)

    # 模型保存
    model.save("wearmask.h5")

结合模型进行识别

def mask_check():

    # 加载模型
    model = load_model("wearmask.h5")
    
    # 人脸分类器
    face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')  # haarcascade_frontalface_alt2
    
    # 调度摄像头
    camera = cv2.VideoCapture(0, cv2.CAP_DSHOW)
    while True:
        
        #读取一帧
        ret, frame = camera.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.1, 5)
        for (x, y, w, h) in faces:
            image = frame[y:y + h, x:x + w]
            
            # 交给模型识别前的准备工作 
            img_resize = cv2.resize(image, dsize=(200, 200))
            
            test_x = []
            test_x.append(img_resize)
            images_test = np.array(test_x)
            images_test = images_test.reshape(images_test.shape[0], 3, 200, 200)
            
            # 获取到识别结果
            value = np.argmax(model.predict(images_test), axis=1)
            
            # 画出矩形框圈出人脸
            cv2.rectangle(frame, (x, y), (x + w, y + h), (255, 0, 0), 2)
            # 显示捕捉了多少张人脸
            font = cv2.FONT_HERSHEY_SIMPLEX
            print(value[0])
            if value[0] == 0:
                print('请带好口罩')
                cv2.putText(frame, 'Please wear a mask!', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
            else:
                print('已带口罩')
                cv2.putText(frame, 'Wearing a mask!', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
        # 显示图像
        cv2.imshow('camera', frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    camera.release()
    cv2.destroyAllWindows()

最终完整代码

import numpy as np
from keras.utils import np_utils
from sklearn.model_selection import train_test_split
import cv2
from keras.models import Sequential, load_model
from keras.layers import Convolution2D, Dense, Activation, Flatten, Dropout, MaxPooling2D


def CatchPICFromVideo(catch_num):
    face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')  # haarcascade_frontalface_alt2
    camera = cv2.VideoCapture(0, cv2.CAP_DSHOW)
    num = 0
    while True:
        ret, frame = camera.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.1, 5)
        for (x, y, w, h) in faces:
            img_name = "imgs/other/" + str(num) + ".jpg"
            image = frame[y:y + h, x:x + w]
            print(img_name)

            # 将图片处理成固定大小
            img_resize2 = cv2.resize(image, dsize=(200, 200))

            cv2.imwrite(img_name, img_resize2)
            num += 1
            if num > catch_num:
                break

            # 画出矩形框圈出人脸
            cv2.rectangle(frame, (x, y), (x + w, y + h), (255, 0, 0), 2)
            # 显示捕捉了多少张人脸
            font = cv2.FONT_HERSHEY_SIMPLEX
            cv2.putText(frame, f'num:{str(num)}', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
        if num > catch_num:
            break
        # 显示图像
        cv2.imshow('camera', frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    camera.release()
    cv2.destroyAllWindows()


def load_data():
    images = []
    labels = []
    for i in range(500):
        img = cv2.imread("imgs/other/" + str(i) + ".jpg")
        images.append(img)
        labels.append(1)
    for i in range(500):
        img = cv2.imread("imgs/me/" + str(i) + ".jpg")
        images.append(img)
        labels.append(0)
    images = np.array(images)
    labels = np.array(labels)
    return images, labels


def cnn_train(images, labels):
    # sklearn 划分数据集
    x_train, x_test, y_train, y_test = train_test_split(images, labels, test_size=0.3, random_state=3)

    # （400, 200, 200, 3） -> (400, 3, 200, 200) -- (图片数量，宽，高，通道数) -> (图片数量，通道数，宽，高)
    x_train = x_train.reshape(x_train.shape[0], 3, 200, 200)
    x_test = x_test.reshape(x_test.shape[0], 3, 200, 200)

    # 目标值 转化为 One-Hot编码 ： 2 -> [0, 0, 1];  0 -> [1, 0, 0]
    y_train = np_utils.to_categorical(y_train, num_classes=2)
    y_test = np_utils.to_categorical(y_test, num_classes=2)

    # 标准化
    x_train = x_train.astype('float32')
    x_test = x_test.astype('float32')

    x_train /= 255.0
    x_test /= 255.0

    # Keras 模型构建
    model = Sequential()

    model.add(Convolution2D(filters=32, kernel_size=[3, 3], padding='same', input_shape=(3, 200, 200)))
    model.add(Activation('relu'))

    model.add(Convolution2D(filters=32, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
    model.add(Dropout(0.25))

    model.add(Convolution2D(filters=64, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(Convolution2D(filters=64, kernel_size=[3, 3], padding='same'))
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
    model.add(Dropout(0.25))

    model.add(Flatten())
    model.add(Dense(512))
    model.add(Activation('relu'))
    model.add(Dropout(0.25))
    model.add(Dense(2))
    model.add(Activation('softmax'))

    model.summary()

    # 优化器
    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

    # 训练
    model.fit(x_train, y_train, epochs=4, batch_size=100)

    # 查看效果
    loss, acc = model.evaluate(x_test, y_test)
    print("最终loss\n", loss)
    print("最终acc\n", acc)

    # 模型保存
    model.save("wearmask.h5")


def mask_check():

    # 加载模型
    model = load_model("wearmask.h5")

    # 人脸分类器
    face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')  # haarcascade_frontalface_alt2

    # 调度摄像头
    camera = cv2.VideoCapture(0, cv2.CAP_DSHOW)
    while True:

        #读取一帧
        ret, frame = camera.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.1, 5)
        for (x, y, w, h) in faces:
            image = frame[y:y + h, x:x + w]

            # 交给模型识别前的准备工作
            img_resize = cv2.resize(image, dsize=(200, 200))

            test_x = []
            test_x.append(img_resize)
            images_test = np.array(test_x)
            images_test = images_test.reshape(images_test.shape[0], 3, 200, 200)

            # 获取到识别结果
            value = np.argmax(model.predict(images_test), axis=1)

            # 画出矩形框圈出人脸
            cv2.rectangle(frame, (x, y), (x + w, y + h), (255, 0, 0), 2)
            # 显示捕捉了多少张人脸
            font = cv2.FONT_HERSHEY_SIMPLEX
            print(value[0])
            if value[0] == 0:
                print('请带好口罩')
                cv2.putText(frame, 'Please wear a mask!', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
            else:
                print('已带口罩')
                cv2.putText(frame, 'Wearing a mask!', (x + 30, y + 30), font, 1, (255, 0, 255), 4)
        # 显示图像
        cv2.imshow('camera', frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    camera.release()
    cv2.destroyAllWindows()


if __name__ == '__main__':
    # images, labels = load_data()
    # cnn_train(images, labels)
    mask_check()

效果演示

为佩戴口罩

> 已佩戴口罩