Performance improvement of noise content video using Raspberry Pi

This paper proposes the implementation of a service running on Raspberry Pi which captures video from the camera and streams it to processor, which processes it and displays on the screen. Implementation of this scenario needs following topics to be addressed. For controlling and capturing the video, Raspberry provides the console application raspivid which can be executed with parameters specifying how the video shall be captured. You can specify parameters like width, height or frames per seconds as if the video be produced to a file to out. The service application implemented in this example internally uses the raspivid. For video capturing the service starts raspivid and then reads incoming video data from raspivid's out. The service executes raspivid with above specified parameters. Live video data continuously coming from raspivid's out needs to be transferred across the network to the connected client. To transfer the data for inter-process communication the implementation uses the lightweight cross-platform library. The communication is based directly on duplex channels to avoid serialization or deserialization. It means the service application running on Raspberry Pi uses input duplex channel and the .NET client running on PC as output channel. When a raspivid's out provides chunk of video data, the service uses the duplex channels to send data to connected clients. Then it uses output channel to receive the video data and to notify it for further processing.