Architecture for Blocking Detection in Wireless Video Source Authentication

Blocking is a common artifact in wireless video streaming services, mainly attributed to packet loss degradation in real-time transmission scenarios. In this paper, we present a simple algorithm and architecture for robust detection of blocking artifact. We first take Discrete Wavelet Transform of the original video frame followed by utilizing a unique property of staircase sized repeated pattern in the videos. The variance of this pattern is measured as the extent of blocking in a video frame. We propose two architectures for blocking detection: one using orthogonal wavelets which can be seamlessly integrated to video source authentication, and the other based on bi-orthogonal wavelets, and can be used in robust stand-alone blocking detection. A prototype implementation on a Xilinx Virtex-6 XC6VLX75 FPGA device was optimized to obtain a clock frequency of 167 (396) MHz or orthogonal (and bi-orthogonal wavelets) using 4 (0) multipliers in the design respectively.