import cv2
import numpy as np
import osclass CaffeModelLoader:    @staticmethoddef load(proto, model):net = cv2.dnn.readNetFromCaffe(proto, model)return netclass FrameProcessor:   def __init__(self, size, scale, mean):self.size = sizeself.scale = scaleself.mean = meandef get_blob(self, frame):img = frame(h, w, c) = frame.shapeif w>h :dx = int((w-h)/2)img = frame[0:h, dx:dx+h]resized = cv2.resize(img, (self.size, self.size), cv2.INTER_AREA)blob = cv2.dnn.blobFromImage(resized, self.scale, (self.size, self.size), self.mean)return blob

CaffeModelLoader类具有一个静态方法来加载来自磁盘的Caffe模型。之后的FrameProcessor类旨在将数据从图像转换为DNN的特定格式。该类的构造函数接收三个参数。size参数定义用于DNN处理的输入数据的大小。用于图像处理的卷积网络几乎总是使用正方形图像作为输入，因此我们仅为宽度和高度指定一个值。scale和mean参数用于将数据缩放到训练SSD所用的值范围。该类的唯一方法是get_blob，它接收图像并返回blob——神经网络处理的一种特殊结构。为了接收blob，首先将图像调整为指定的正方形。然后，使用OpenCV的DNN模块中的blobFromImage方法使用指定的scale、size和mean值，从调整大小后的图像创建Blob 。

class SSD: def __init__(self, frame_proc, ssd_net):self.proc = frame_procself.net = ssd_netdef detect(self, frame):blob = self.proc.get_blob(frame)self.net.setInput(blob)detections = self.net.forward()# detected object countk = detections.shape[2]obj_data = []for i in np.arange(0, k):obj = detections[0, 0, i, :]obj_data.append(obj)return obj_datadef get_object(self, frame, data):confidence = int(data[2]*100.0)(h, w, c) = frame.shaper_x = int(data[3]*h)r_y = int(data[4]*h)r_w = int((data[5]-data[3])*h)r_h = int((data[6]-data[4])*h)if w>h :dx = int((w-h)/2)r_x = r_x+dxobj_rect = (r_x, r_y, r_w, r_h)return (confidence, obj_rect)def get_objects(self, frame, obj_data, class_num, min_confidence):objects = []for (i, data) in enumerate(obj_data):obj_class = int(data[1])obj_confidence = data[2]if obj_class==class_num and obj_confidence>=min_confidence :obj = self.get_object(frame, data)objects.append(obj)return objects

上面的类的构造函数有两个参数：frame_proc用于将图像转换为blob而ssd_net用于检测对象。主要方法detect接收帧（图像）作为输入，并使用指定的帧处理器从该帧获取blob。该blob用作网络的输入，我们使用该forward方法获得检测结果。这些检测结果显示为4级数组（张量）。我们不会分析整个张量。我们只需要数组的第二维。我们将从检测结果中提取出来并返回结果——对象数据列表。

[array([ 0.        , 15.       ,  0.90723044,  0.56916684,  0.6017439 ,0.68543154,  0.93739873], dtype=float32)]

该数组包含七个数字：

class Utils:   @staticmethoddef draw_object(obj, label, color, frame):(confidence, (x1, y1, w, h)) =  objx2 = x1+wy2 = y1+hcv2.rectangle(frame, (x1, y1), (x2, y2), color, 2)y3 = y1-12text = label + " " + str(confidence)+"%"cv2.putText(frame, text, (x1, y3), cv2.FONT_HERSHEY_SIMPLEX, 0.6, color, 1, cv2.LINE_AA)@staticmethoddef draw_objects(objects, label, color, frame):for (i, obj) in enumerate(objects):Utils.draw_object(obj, label, color, frame)

现在我们可以编写代码来启动人检测算法：

proto_file = r"C:\PI_RPD\mobilenet.prototxt"
model_file = r"C:\PI_RPD\mobilenet.caffemodel"
ssd_net = CaffeModelLoader.load(proto_file, model_file)
print("Caffe model loaded from: "+model_file)proc_frame_size = 300
# frame processor for MobileNet
ssd_proc = FrameProcessor(proc_frame_size, 1.0/127.5, 127.5)
person_class = 15ssd = SSD(ssd_proc, ssd_net)im_dir = r"C:\PI_RPD\test_images"
im_name = "woman_640x480_01.png"
im_path = os.path.join(im_dir, im_name)
image = cv2.imread(im_path)
print("Image read from: "+im_path)obj_data = ssd.detect(image)
persons = ssd.get_objects(image, obj_data, person_class, 0.5)
person_count = len(persons)
print("Person count on the image: "+str(person_count))
Utils.draw_objects(persons, "PERSON", (0, 0, 255), image)res_dir = r"C:\PI_RPD\results"
res_path = os.path.join(res_dir, im_name)
cv2.imwrite(res_path, image)
print("Result written to: "+res_path)

该代码实现了一个size= 300的帧处理器，因为我们使用的模型适用于300 x 300像素的图像。scale和mean参数都被用于MobileNet模型训练，同样的值。这些值必须始终分配给模型的训练值，否则模型的精度会降低。person_class的值是15，因为人体是模型上下文中的第15个类。