import cv2 import numpy as np import mindspore from mindspore.train.serialization import load_checkpoint from resnet import ResNet, BasicBlock

加载ResNet模型

network = ResNet(BasicBlock, [2, 2, 2, 2], num_classes=100) model = load_checkpoint('checkpoint_resnet-10_156.ckpt') mindspore.load_param_into_net(network, model) network.set_train(False)

加载人脸识别模型

face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')

打开摄像头

cap = cv2.VideoCapture(0)

while True: # 读取摄像头中的图像 ret, frame = cap.read()

# 对图像进行人脸识别
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)

# 对每一个人脸进行识别
for (x, y, w, h) in faces:
    # 提取人脸图像
    face_img = frame[y:y+h, x:x+w]
    # 缩放人脸图像
    face_img = cv2.resize(face_img, (224, 224))
    # 转换为MindSpore可用的格式
    face_img = np.transpose(face_img, (2, 0, 1)).astype(np.float32) / 255.
    face_img = np.expand_dims(face_img, axis=0)
    # 使用ResNet模型进行识别
    output = network(Tensor(face_img))
    # 获取结果
    predict = np.argmax(output.asnumpy(), axis=1)[0]
    # 在图像上显示结果
    cv2.rectangle(frame, (x, y), (x+w, y+h), (0, 255, 0), 2)
    cv2.putText(frame, str(predict), (x, y-10), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)

# 显示图像
cv2.imshow('frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
    break

关闭摄像头

cap.release() cv2.destroyAllWindows()