Adaptívne rezervácie sieťových zdrojov podľa klasifikácie typu video scény

Abstract

Video quality assessment requires a combined approach that includes objective metrics, subjective testing, and monitoring of the network. Carefully conducted video quality subjective tests are extremely reliable and repeatable. A subjective video quality tests use a small set of short video sequences (10 sec. according to standard) to measure people’s opinions of the quality of different video processing options and with degradation of this video. These tests focus on people’s current opinion, as opposed, e.g., to opinions of an entire movie. The goal is to make an impartial judgment about opinion trends. The aim of this work is the subjective evaluation of the quality at the different bitrate. This evaluation is helpful to define the threshold for setting of the bitrate and his change based on the spatial and temporal information of the video. I compared the results of subjective tests with the results of the objective tests. Based on this comparison I defined correlation. After, I created optimum mapping function for to define the threshold for setting of the variable bitrate. I used standard video sequences for the evaluation of the video quality. These video sequences were carefully select based on the location of the individual quartile of the temporal and spatial information. I compared coding algorithms H.264/AVC (Advanced Video Coding) and H265/HEVC (High Efficiency Video Coding) and also different resolutions, FHD (Full High Definition) and UHD (Ultra High Definition).