Riotboard: IoT based intrusion detection & alert system: Part2 : Capturing image

In this mini blog series we will build an IoT based intruder detection and alert system.

Blogs:

Part1 : Connecting Pir Sensor

Part2 : Capturing intruder image  <--- this blog

Part3 : Email Alerts

In previous blog we integrated a PIR sensor to Riotboard for human movement detection.

In this very short guide we will try to capture image once an IR detection is made.

I am using a  Logitech C310Logitech C310 USB camera for this purpose.

Connect pir sensor HC-Sr501 and Camera to Riotboard.

to capture image we will use fswebcam tool.

in terminal type:

    apt-get install fswebcam

Now we determine the usb device

type "ls /dev/video*" without pluggin camera.

Now pluggin camera to Riotboard, and type "ls /dev/video*"

check for video device in /dev.

the camera hardware is accessible at the device /dev/video2

now add below line to previous code  below ln28.

system("fswebcam -d /dev/video2 capture.jpg");

Next we will compile the below code and test.

#include <stdio.h>
#include <fcntl.h>
#include <unistd.h>
#include <string.h>

int main()
{
int ret =0;
int fd=0;
char buf[1];

system("devmem 0x20E00A0 w 0x5  ");
system("echo 113 > /sys/class/gpio/unexport");
system("echo 113 > /sys/class/gpio/export");
usleep(100000);
system("echo in > /sys/class/gpio/gpio113/direction");
usleep(100000);

fd = open("/sys/class/gpio/gpio113/value",O_RDONLY);
while(1)
{
ret = read(fd,buf,sizeof(1));
if(lseek(fd,0,SEEK_SET) < 0) return 1;

if(ret>0 && (atoi(buf)==1))
    {
        printf("\n\033[031;43m HUMAN DETECTED \033[0m ");
        fflush(stdout);
        system("fswebcam -d /dev/video2 capture.jpg");
        usleep(500000);
    }
else
    printf("\nNo Detection");

fflush(stdin);
fflush(stdout);
usleep(500000);

}
close(fd);
}

Now perform some hand movement infront of Pir sensor, it will show human detected and captures image as "capture.jpg".

before capture:

after capture:

Now with every pir interrupt we can capture the image of the intruder.

Next time we will add an image based email alert for the above.