Multipurpose spybot

SpyBot, designed to assist combat personnels in hostile areas, is a Raspberry Pi & Arduino based surveillance robot capable of live video streaming by means of an IP webcam and intelligent surveillance. The vehicle when in surveillance mode watches a frame of interest and detects intruders. The weapon in the vehicle can be used in case of emergency and distance details of the intruder will also be available.

In the context of increasing terrorist attacks,a surveillance vehicle which is manually controlled from a laptop via GUI by means of Zigbee transmitter receiver module can be used to assist combat personnels in hostile environments. There will be live video streaming to the laptop by means of an IP webcam whenever the vehicle is moving. Once the vehicle reaches the frame to be surveyed live video stream can be stopped and surveillance mode can be turned on. The room will be continuously surveyed and once an intruder is detected, it sends a notification along with the  distance  of the intruder from the vehicle . At this point, the controlling person will be able to turn on the live stream from any device connected to the private network by providing the IP address. The camera will be kept on a servo arm which is capable of 3 axis motion. Camera position can be adjusted for a better view and the vehicle may be retracted,if needed. The vehicle will also be equipped with a weapon which can be triggered under emergency situations.Raspberry Pi is used for motion detection and Arduino is used for driving the vehicle and triggering the weapon.

Video streaming using IP based webcam. The vehicle has two cameras, one is the IP webcam used for live video streaming and the second camera is used for motion detection. Peripherals such as monitor, keyboard and mouse can be connected to Pi for programming, via USB ports. When con figured with wifi the Raspberry Pi will be assigned an IP address. This IP is used to stream the video to any other device connected to the same network. Live video stream can be used to watch the frame from a distance.

Processed image using Raspberry Pi, intruder detection. Background subtraction is employed and number of nonzero pixels are estimated. Once the vehicle is stationed to watch the frame, the motion detection will be employed and if any intruder is detected, the system sends an alert to the controlling personnel via the Zigbee connected to the Raspberry Pi. The camera connected to the USB port of Rasp-berry Pi will be capturing video continuously. The screenshots are taken at a default rate and are processed in opencv. Adjacent frames are compared and di erential images are generated. The nonzero pixels are estimated from the corresponding matrix and the number is compared with a pre de termined threshold value. If the nonzero pixels exceed threshold a count is incremented and if the count reaches a particular value the user will be alerted. After processing three successive frames, the background will be updated to avoid false alarms.

The army personnel will be able to control the vehicle from his laptop by means of visual basics GUI. The laptop with GUI forms the transmitter section. The GUI has controls for vehicle motion and direction control. The vehicle can be moved forward, backward, left or right and stopped when needed. Each time a control but-ton is pressed, characters are transmitted serially via Zigbee to the receiver Zigbee connected to the At-mega 328 in the vehicle. The micro controller board has been programmed to undertake necessary actions corresponding to the serial characters received.Once motion is  detected the controlling person will be notified and provided the distance details of the intruder