GPU and multi-threaded CPU enabled normalized cross correlation

A. Nafis, E. Teters, R. Aktar, G. Seetharaman, and K. Palaniappan

Proc. SPIE Conf. Geospatial InfoFusion X (Defense + Commercial Sensing), 2020

PlainText, Bibtex, URL, DOI, Google Scholar


Image matching has been a critical research topic in many computer vision applications, such as stereo vision, feature tracking, motion tracking, image registration and mosaicing, object recognition, and 3D reconstruction. Normalized Cross Correlation (NCC) is a template-based image matching approach which is invariant to linear brightness and contrast variations. As a first step in mosaicing, we use NCC to a great extent for matching images which is an expensive and time consuming operation. Thus, an attempt is made to implement NCC in GPU and multi-CPU to improve the execution time for real-time applications. We also show performance differences for our different parallelization techniques: dense and sparse NCC. Finally, we compare the enhancement in performance and efficiency in timing by switching NCC implementation from CPU to GPU.