A novel local image descriptor is proposed in this paper, which combines intensity orders and gradient distributions in multiple support regions. The novelty lies in three aspects...
Abstract. We investigate a number of approaches to pose invariant face recognition. Basically, the methods involve three sequential functions for capturing nonlinear manifolds of f...
In this work, we demonstrate examples about uses of practical image processing techniques over a new interpretation of power quality event data. Power quality event data are 1D da...
In this paper, we present a method to speed up video encoding of GPU rendered scenes. Modern video codecs, like H.264/AVC, are based on motion compensation and support partitionin...
H.264 is the current state-of-the-art monoscopic video codec providing almost twice the coding efficiency with the same quality comparing the previous codecs. With the increasing ...