Smart Surveillance System Using Tensor Flow

Abstract: In this modern era, surveillance security camera system plays a vital role in day-to-day life, as it full-fills the safety aspects against burglary, theft related problems and reduces the need of the watchman. This work proposes a smart surveillance system that would act as a solution for security issues. The proposed system for smart surveillance involves a security camera with the night vision capabilities interfaced with raspberry pi. OpenCV is used to perform real-time image processing. This system proves to be a cost-effective way of surveillance, as it uses a credit card-sized chip Raspberry Pi (RPI). The methodology of the proposed system involves breaking down a video, captured by the camera in real-time, into separate frames. Each frame is processed using image processing tools in python. The pre- processed images (frames) are compared for change in pixel values, to detect the movement of an object. Then, an object detection algorithm like Tensor Flow is used to detect the object and classify it. Later, a buzzer is actuated to cognize the security system and personnel. Also, an alert notification is sent to the owner.

Keywords: Tensor Flow, Raspberry Pi, OpenCV, Python