myjournal manuscript No. (will be inserted by the editor) Feature Tracking and Matching in.pdf

myjournal manuscript No. (will be inserted by the editor) Feature Tracking and Matching in Video Using Programmable Graphics Hardware Sudipta N. Sinha 1, Jan-Michael Frahm 1, Marc Pollefeys 1, Yakup Genc 2? 1Department puter Science, CB# 3175 Sitterson Hall, University of North Carolina at Chapel Hill, NC 27599 2Real-time Vision and Modeling Department, Siemens Corporate Research, 755 College Road East, Princeton, NJ 08540 Received: date / Revised version: date Abstract This paper describes novel implementations of the KLT feature track- ing and SIFT feature extraction algorithms that run on the graphics processing unit (GPU) and is suitable for video analysis in real-time vision systems. While signi?cant acceleration over standard CPU implementations is obtained by ex- ploiting parallelism provided by modern programmable graphics hardware, the CPU is freed up to run putations in parallel. Our GPU-based KLT im- plementation tracks about a thousand features in real-time at 30 Hz on 1024×768 resolution video which is a 20 times improvement over the CPU. The GPU-based Send offprint requests to: ?Present address:Insert the address here if needed 2 Sudipta N. Sinha et al. SIFT implementation extracts about 800 features from 640×480 video at 10Hz which is approximately 10 times faster than an optimized CPU implementation. 1 Introduction Extraction and matching of salient 2D feature points in video is important in puter vision tasks like object detection, recognition, structure from motion and marker-less augmented reality. While certain sequential tasks like structure from motion for video [18] require online feature point tracking, others need features to be extracted and matched across frames separated in time (eg. wide-baseline stereo). The increasing programmability putational power of the graph- ics processing unit (GPU) present in modern graphics hardware provides great scope for acceleration puter vision algorith