利用Dragonbaord 410c开发板搭建具有运动追踪功能摄像头

很多时候我们不在家的时候，都希望能够了解到家里的情况，这时候你可能想到需要一个web摄像头来进行监控，但是作为一个410c开放爱好者，我们去买一个这样的摄像头就太浪费了，今天就带大家一起来用手中的dragonbaord 410c开发板来实现一些个简单的带有运动追踪功能的摄像头，废话不多说，接下来上干货。

一、准备工作

准备一块dragonboard 410c开发板、一个USB摄像头、鼠标、键盘当然还有适配器和能够接入网络的wifi热点。

这里鼠标和键盘主要是用来进行开发，如果你的dragonboard 410c开发板已经安装好了debain操作系统，这里我们接下来的工作将全部在dragonboard 410c开发板上完成，不需要PC机，具体环境搭建参考如下：

按照上述环境搭建后，就可以启动你的dragonboard410c开发板了，这里首先我们不连接摄像头，直接连接鼠标，因为还需要配置你的dragonboard 410c开发板的网络和在上面进行编程，编号程序后运行程序的时候再讲鼠标拔掉，连接USB摄像头。

二、代码实现

整个开发过程中采用Python脚本语言来进行设计，并且借助于opencvpython库来进行相关的图形处理操作，废话不多说，具体代码如下：

#!/usr/bin/env python

import argparse

import datetime

import imutils

import time

import cv2

import numpy as np

ap = argparse.ArgumentParser()

ap.add_argument("-v","--video", help="path to the video file")

ap.add_argument("-a","--min-area", type=int, default=5000, help="minimum areasize")

args = vars(ap.parse_args())

 ifargs.get("video", None) is None:

   camera = cv2.VideoCapture(0)

   time.sleep(0.1)

 else:

   camera = cv2.VideoCapture(args["video"])

 framenext = None

firstFrame = None

i=0

while True:

   (grabbed, frame) = camera.read()

   text = "Unoccupied"

   if not grabbed:

       break

   frame = imutils.resize(frame, width=500)

   framenext = frame

   gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

   #gray_pre = np.float32(gray)

   gray = cv2.GaussianBlur(gray, (21, 21), 0)

   if firstFrame is None:

       firstFrame = gray

       continue           

   frameDelta = cv2.absdiff(firstFrame, gray)

   thresh = cv2.threshold(frameDelta, 115, 255, cv2.THRESH_BINARY)[1]

   thresh = cv2.dilate(thresh, None, iterations=2)

   (cnts, _) = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL,

       cv2.CHAIN_APPROX_SIMPLE)

    for c in cnts:

       # if the contour is too small, ignore it

       if cv2.contourArea(c) < args["min_area"]:

           continue

        # compute the bounding box for the contour  draw it on the frame

       # and update the text

       (x, y, w, h) = cv2.boundingRect(c)

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

       text = "Occupied"

# caculation feature points

   graynext  =cv2.cvtColor(framenext, cv2.COLOR_BGR2GRAY)

   #graynext = np.float32(graynext);

   detector = cv2.xfeatures2d.SIFT_create()

   keypoints = detector.detect(gray,None)

   framenext=cv2.drawKeypoints(graynext,keypoints)

# draw the text and timestamp on the frame

   cv2.putText(frame, "Room Status: {}".format(text), (10, 20),

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

   cv2.putText(frame, datetime.datetime.now().strftime("%A %d %B %Y%I:%M:%S%p"),

       (10, frame.shape[0] - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.35, (0, 0, 255),1)

   cv2.imshow("Security Feed", frame)

   cv2.imshow("Thresh", thresh)

   cv2.imshow("Frame Delta", frameDelta)

   cv2.imshow("keypoints",framenext)

   #key = cv2.waitKey(1) & 0xFF

   if cv2.waitKey(10) == 27:

       break 

                   camera.release()

    cv2.destroyAllWindows()

         #Send the message via our own SMTP server (sendmail)

    sender = '[e-mail]'

    receivers = ['[e-mail]']

    text = MIMEText('Hello Qualcomm my name is NO1','plain','utf-8')

    image_url = "./detected_face.jpg"

     image = MIMEImage(open(image_url,'rb').read())

    image.add_header('Content-ID','')

    message = MIMEMultipart('related')

    message['From'] = Header("workshop",'utf-8')

    message['To'] = Header("Test",'utf-8')

    subject = 'Python SMTP Test'

    message['Subject'] = Header(subject,'utf-8')

    message.attach(text)

    message.attach(image)

try:

         smtpObj=smtplib.SMTP('localhost')

         #smtpObj.set_debuglevel(1)

         smtpObj.sendmail(sender,receivers, message.as_string())

         smtpObj.quit()

         print"done!"

except smtplib.SMTPException:

 print "error:send failed"

三、测试

编写好代码后保存，然后使用Python monition_camera.py命令就可以运行上述代码了，这里我们收到的邮件及附件效果如下，由于涉及到隐私具体的跟踪人体目标就不展示了，大家下载代码实现即可以看到效果。