def __init__(self):

self.model = self.build_model()

def build_model(self) -> Any:

"""

opencv_version = cv2.__version__.split(".")

if not len(opencv_version) >= 2:

raise ValueError(

f"OpenCv的版本必须包含主版本和次版本，但当前版本为 {opencv_version}"

)

opencv_version_major = int(opencv_version[0])

opencv_version_minor = int(opencv_version[1])

if (opencv_version_major < 4 or (opencv_version_major == 4 and opencv_version_minor < 8)):

# 最低要求: https://github.com/opencv/opencv_zoo/issues/172

raise ValueError(f"YuNet 需要 opencv-python >= 4.8，但当前版本为 {cv2.__version__}")

# pylint: disable=C0301

weight_file = r'../DeepFace.Console/bin/Debug/net8.0/models/face_detection_yunet_2023mar.onnx'

try:

face_detector = cv2.FaceDetectorYN_create(weight_file, "", (0, 0))

except Exception as err:

raise ValueError(

"调用 opencv.FaceDetectorYN_create 模块时发生异常。"

+ "这是一个可选依赖。"

+ "你可以通过 pip install opencv-contrib-python 安装它。"

) from err

return face_detector

def detect_faces(self, img: np.ndarray) -> List[FacialAreaRegion]:

"""

# FaceDetector.detect_faces不支持score_threshold参数

# 我们可以通过环境变量设置它

score_threshold = float(os.environ.get("yunet_score_threshold", "0.9"))

resp = []

faces = []

height, width = img.shape[0], img.shape[1]

# 如果图像太大则调整大小(Yunet有时在大尺寸输入上无法检测人脸)

# 选择640作为阈值是因为这是Yunet中max_size的默认值

resized = False

r = 1 # 调整大小的系数

if (height > 640 or width > 640):

r = 640.0 / max(height, width)

img = cv2.resize(img, (int(width * r), int(height * r)))

height, width = img.shape[0], img.shape[1]

resized = True

self.model.setInputSize((width, height))

self.model.setScoreThreshold(score_threshold)

_, faces = self.model.detect(img)

print('r:',r, 'width:',width, 'height:',height)

if faces is None:

return resp

for face in faces:

"""

(x, y, w, h, x_le, y_le, x_re, y_re) = list(map(int, face[:8]))

# YuNet在认为检测到的人脸部分在框架外时会返回负坐标

x = max(x, 0)

y = max(y, 0)

if resized:

x, y, w, h = int(x / r), int(y / r), int(w / r), int(h / r)

x_re, y_re, x_le, y_le = (

int(x_re / r),

int(y_re / r),

int(x_le / r),

int(y_le / r),

)

confidence = float(face[-1])

facial_area = FacialAreaRegion(

x=x,

y=y,

w=w,

h=h,

confidence=confidence,

left_eye=(x_re, y_re),

right_eye=(x_le, y_le),

)

resp.append(facial_area)

# 初始化一个FaceDetector对象

yunet = YuNetClient()

# 读取图像

img = cv2.imread(r"../SampleImages/000000001000.jpg")

# 检测人脸

faces = yunet.detect_faces(img)

# 打印检测到的人脸数量和详细信息

print(f"检测到{len(faces)}张人脸。")

# 在图像上绘制检测结果

for face in faces:

# 绘制人脸边界框

cv2.rectangle(img, (face.x, face.y), (face.x + face.w, face.y + face.h), (0, 255, 0), 2)

# 绘制左眼

cv2.circle(img, face.left_eye, 2, (255, 0, 0), 2)

# 绘制右眼

cv2.circle(img, face.right_eye, 2, (0, 0, 255), 2)

# 添加置信度文本

cv2.putText(img, f"{face.confidence:.2%}", (face.x, face.y - 10),

cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

# 保存结果图像

cv2.imwrite("out.jpg", img)

# 打印检测信息

for i, face in enumerate(faces, 1):

print(f"\n人脸 #{i}:")

print(f"位置: x={face.x}, y={face.y}, 宽度={face.w}, 高度={face.h}")

print(f"置信度: {face.confidence:.2%}")

print(f"左眼坐标: {face.left_eye}")