An Efficient Hardware Architecture for Block Based Image Processing Algorithms

Tomasz Kryjak, Marek Gorgon, Mateusz Komorkiewicz

Abstract: In this paper a novel method for data reordering in a video stream is presented. It can be used in high performance vision algorithms that require block based image processing. For case study a block based optical flow histogram computation application was selected. The proposed solution allows for a 2.3x and 6.3x speed-up for fixed-point and floating-point calculations respectively. This enables real-time operations for \(1920 \times 1080\) pixels or higher resolutions.