监控器物检测object detection实战
实战目的：根据家里的监控器，实时检测出监控拍到的物体，包括人、车等。
基本情况：
1、家里安装有JOVISION监控器（中维），摄像头与存储设备通过路由器连接。
2、JOVISION对应有一个客户端，名称为：云视通网络监控系统，可以在JOVISION官网下载。
——————————————————————————————————————————
思路：
用一个训练好的小的Faster RCNN模型，结合opencv，将摄像头采集到图片输送模型，
标注出图片中检测出的物体。
——————————————————————————————————————————
困难：
JOVISION客户端处理监控视频有两种方法：（1）点击，录制好一段存起来；（2）点一个按钮保存一个快照
这两种方法都不好实施，方案（1）也需要点击，并且视频存储起来后才可使用，无法做到实时，方案（2）
如果能手动一直操作鼠标点击保存快照，而另一边程序实时检测快照图片，那就能做到实时。
但人工一直点击不现实。
——————————————————————————————————————————
解决方案：
用按键精灵编写了一个小脚本，用于定时点击监控，定时点击快照生成按钮。
按键精灵代码：
Plugin.Window.MousePoint
点击 = WaitClick()
Delay 10
Hwnd = Plugin.Window.MousePoint()
GetCursorPos x, y
sRect = Plugin.Window.GetClientRect(Hwnd)
MyArray = Split(sRect, "|")
L = CLng(MyArray(0))
T = CLng(MyArray(1))
R = CLng(MyArray(2))
B = CLng(MyArray(3))
点击1 = WaitClick()
Delay 10
Hwnd1 = Plugin.Window.MousePoint()
GetCursorPos x1, y1
sRect1 = Plugin.Window.GetClientRect(Hwnd1)
MyArray1 = Split(sRect1, "|")
L1 = CLng(MyArray1(0))
T1 = CLng(MyArray1(1))
R1 = CLng(MyArray1(2))
B1 = CLng(MyArray1(3))
Tick = 2000
Do
Delay Tick
Call Plugin.Bkgnd.LeftClick(Hwnd1, x1 - L1, y1 - T1)
Delay Tick
Call Plugin.Bkgnd.LeftClick(Hwnd, x - L, y - T)
Loop
按键精灵按F5运行，切换到JOVISION客户端，按F10启动，然后操作。
按键精灵界面：
JOVISION客户端界面：
监控快照需要首先点击需要截图的监控页面，然后点击按钮“现场拍照”
——————————————————————————————————————————
生成的快照图片被保存在一个文件夹中。
物体检测代码：
# USAGE
# python deep_learning_object_detection_files.py --prototxt MobileNetSSD_deploy.prototxt.txt --model MobileNetSSD_deploy.caffemodel
# import the necessary packages
import numpy as np
import argparse
import time
import cv2
import os
FindPath = "E:\\JdvrFile\\CapFile\\"
def deleteimage(fullfilename):
os.remove(fullfilename)
# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-p", "--prototxt", required=True,
help="path to Caffe 'deploy' prototxt file")
ap.add_argument("-m", "--model", required=True,
help="path to Caffe pre-trained model")
ap.add_argument("-c", "--confidence", type=float, default=0.2,
help="minimum probability to filter weak detections")
args = vars(ap.parse_args())
# initialize the list of class labels MobileNet SSD was trained to
# detect, then generate a set of bounding box colors for each class
CLASSES = ["background", "aeroplane", "bicycle", "bird", "boat",
"bottle", "bus", "car", "cat", "chair", "cow", "diningtable",
"dog", "horse", "motorbike", "person", "pottedplant", "sheep",
"sofa", "train", "tvmonitor"]
COLORS = np.random.uniform(0, 255, size=(len(CLASSES), 3))
# load our serialized model from disk
print("[INFO] loading model...")
net = cv2.dnn.readNetFromCaffe(args["prototxt"], args["model"])
while True:
FileNames = os.listdir(FindPath)
for file_name in FileNames:
fullfilename = os.path.join(FindPath,file_name)
# print "fullfilename:", fullfilename
image = cv2.imread(fullfilename)
if image is None:
continue
(h, w) = image.shape[:2]
blob = cv2.dnn.blobFromImage(cv2.resize(image, (300, 300)), 0.007843, (300, 300), 127.5)
net.setInput(blob)
detections = net.forward()
for i in np.arange(0,detections.shape[2]):
confidence = detections[0,0,i,2]
if confidence > args["confidence"]:
idx = int(detections[0,0,i,1])
box = detections[0,0,i,3:7] * np.array([w,h,w,h])
(startX,startY,endX,endY) = box.astype("int")
label = "{}: {:.2f}%".format(CLASSES[idx],confidence*100)
cv2.rectangle(image,(startX,startY),(endX,endY),COLORS[idx],2)
y = startY - 15 if startY - 15 > 15 else startY + 15
cv2.putText(image,label,(startX,y),
cv2.FONT_HERSHEY_SIMPLEX,0.5,COLORS[idx],2)
cv2.imshow("Frame",image)
key = cv2.waitKey(1)
deleteimage(fullfilename)
——————————————————————————————————————————
物体检测代码可以参考github：https://github.com/miracleyhj/object-detection-with-jovision-monitor
关于模型，可以参考博客：https://www.pyimagesearch.com/2017/09/11/object-detection-with-deep-learning-and-opencv/
——————————————————————————————————————————
欢迎微信骚扰：yuan136315miracle