Use of a Raspberry PI single-board computer for image acquisition and transmission

The article presents an analysis of the use of a single-board computer Raspberry PI for video acquisition and transmission. The article focuses on requirements necessary for the recorded image to be used for face analysis to identify facial expressions and microexpressions. The quality of the recorded video frames was verified for different resolutions and fps using PSNR (Peak Signal-to-Noise Ratio). Tests of CPU cores usage were also carried out for simultaneous recording and transmission of different types of video streams. The results show that the size of the effective image area depends on the resolution of recorded video stream. Increasing the frame rate for the given video resolution has a significant impact on the value of PSNR. And the resultant CPU usage, for the available resolutions and frame rates of the video recorded, in most cases does not exceed 15% of the total computing power of the CPU.