Evaluation of an OpenMP Parallelization of Lucas-Kanade on a NUMA-Manycore

Olfa Haggui (1, 2, 3) , Claude Tadonki (1, 2) , Fatma Sayadi (4, 5) , Ouni Bouraoui (5, 3)

1 Mines Paris - PSL (École nationale supérieure des mines de Paris)
2 CRI - Centre de Recherche en Informatique
3 ENISo - National Engineering School of Sousse / Ecole Nationale d'Ingénieurs de Sousse
4 UM - Université de Monastir - University of Monastir
5 FSM - Faculté des Sciences de Monastir

Lucas-Kanade algorithm is a well-known optical flow estimator widely used in image processing for motion detection and object tracking. As a typical image processing algorithm, the procedure is a series of convolution masks followed by 22 linear systems for the optical flow vectors. Since we are dealing with a stencil computation for each stage of the algorithm, the overhead from memory accesses is expected to stand as a serious scalability bottleneck, especially on a NUMA manycore configuration. The objective of this study is therefore to investigate an OpenMP parallelization of Lucas-kanade algorithm on a NUMA manycore, including the performance impact of NUMA-aware settings at runtime. Experimental results on a dual-socket INTEL Broadwell-E/EP is provided together with the corresponding technical discussions

Mots clés: Optical flow; Lucas-Kanade; multicore; manycore; OpenMP; NUMA; scalability.

Domaines: Algorithme et structure de données [cs.DS]; Compilation

Olfa Haggui, Claude Tadonki, Fatma Sayadi, Ouni Bouraoui. Evaluation of an OpenMP Parallelization of Lucas-Kanade on a NUMA-Manycore. 9th Workshop on Applications for Multi-Core Architectures (WAMCA 2018), Sep 2018, Lyon, France. ⟨hal-01859701⟩