Avaliação da qualidade de vídeos transmitidos via vídeo streaming em ambientes residenciais.

Abstract

The use of streaming services has grown significantly in recent years, through systems such as Youtube, Hulu, Netflix, Vimeo, etc. Using the stream, videos are transmitted and displayed in real time and from anywhere in the world via the Internet. Many users of these systems use the same in their homes where it is quite common to find a wireless network (due to the mobility we can achieve through this type of computer network). In spite of having more mobility, a wireless network is more susceptible to interference than the wired network so a video can have its picture degraded more easily just because it is transmitted over a wireless network. We evaluate the transmission of videos via stream to find out whether the video is actually degraded by transmission; if different types of content affect the quality of the received video; and if there are compression standarts (H.264 and MPEG-4 HEVC, the most currently used) which ensure a better received video quality. Through an experimental approach with a design of full factorial experiments, several transmissions of videos were made in the three chosen standarts. In addition to defining the standarts, other variables were defined as: transmission channel occupancy percentage (in order to evaluate the degradation of videos according to occupation); power transmission signal (in order to assess the impact of network signal quality in the video received); amount of motion in the video (to evaluate whether the amount of motion the video displays impacts on its quality). For the experiments we used a residential environment that has all the interference from other networks, just as it can happen in a real environment. To set the occupancy percentages, experiments were performed to measure the maximum transmission capacity of the test network. Concerning the quantity of movement, it was necessary to make a preliminary classification of videos according to spatial and temporal characteristics of each one. The videos were separated into three categories and within these categories, three videos were chosen at random to participate in the experiments. The results show that the HEVC achieved the highest average for quality metricsofthechosenvideos,andthecodecistheonethatloseslessqualityinawireless transmission. The amount of movement affects the quality of the received video, and the greater the amount of motion, the bigger the loss of image quality.