Optimized fixed point implementation of a local stereo matching algorithm onto C66x DSP

Judicaël Menant (1) , Muriel Pressigout (1) , Luce Morin (2, 1) , Jean-Francois Nezan (1)

1 IETR - Institut d'Électronique et des Technologies du numéRique
2 IRT b-com - Institut de Recherche Technologique b-com

Stereo matching techniques aim at reconstructing disparity maps from a pair of images. The use of stereo matching techniques in embedded systems is very challenging due to the complexity of the state-of-the-art algorithms. An efficient local stereo matching algorithm has been chosen from the literature and implemented on a c6678 DSP. Arithmetic simplifications such as approximation by piecewise linear functions and fixed point conversions are proposed. Thanks to factorisation and pre-computing, the memory footprint is reduced by a factor 13 to fit on the memory footprint available on embedded systems. A 14.5 fps speed (factor 60 speed-up) has been reached with a small quality loss on the final disparity map.

Mots clés: Algorithm design and analysis Approximation methods; Computer architecture; Digital signal processing; Embedded systems; Optimization; Real-time systems

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]; Traitement du signal et de l'image [eess.SP]; Systèmes embarqués

Judicaël Menant, Muriel Pressigout, Luce Morin, Jean-Francois Nezan. Optimized fixed point implementation of a local stereo matching algorithm onto C66x DSP. DASIP 2014, Oct 2014, Madrid, Spain. ⟨10.1109/DASIP.2014.7115636⟩. ⟨hal-01101822⟩