ViBee – segmentation algorithm applied to reduce video bitrates in bee weblabs

Abstract

Services provided by Internet allow the creation of new distance applications, as remote laboratories (Weblabs). In this context, the broadcasting of video and images are essential for the success and effectiveness of the experiments. However, the bandwidth of the communication channel does not allow transmission of high-definition video in real-time, which are basic requirements for remote monitoring and actuation activities. Considering a Weblab, also called Webbee, through which it is intended to analyze the behavior of bees, these requirements become more important. Commercial videos, CODEC (MPEG4 and WMV), applied to videos of bees reduce the bitrates, as well as the quality, what makes impracticable the applications for the Webbee. Therefore, this paper presents the proposal of a new algorithm for codification of video of bees and the reduction of bitrates, maintaining the quality of the information. The algorithm includes the video segmentation process by identification, recognition and tracking of objects. The test results show an adaptation of the proposal with reduction of data equivalent to the commercial CODECs, with 93% of similarity.