SPEEDUPIMAGESPATIO TEMPORALFEATURE EXTRACTION USING GPGPU

Abstract

The robust representation of image features becomes fundamental to most machine vision and image registration applications. Spatio-temporal feature extraction algorithms are favored because of their robust generated features. However, they have high computational complexity. In this thesis, we propose new parallel implementations, using GPU computing, for the two most widely used Spatio-temporal feature extraction algorithms: Scale-Invariant Feature Transform (SIFT) and Speeded-Up Robust Feature (SURF). In our implementations, we solve problems with previous parallel implementations, such as load imbalance, thread synchronization, and the use of atomic operations. We compare our presented implementations to previous CPU and GPU parallel implementations of the two algorithms. Results used in Human action recognition and achieve accuracy 96% for SIFT and 94.5% for SURF.