Real-Time and Approximate Iterative Optical Flow Implementation on Low-Power Embedded CPUs

Maxime Millet (1) , Adrien Cassagne (1) , Nicolas Rambaux (2) , Lionel Lacassagne (1)

1 ALSOC - Architecture et Logiciels pour Systèmes Embarqués sur Puce
2 IMCCE - Institut de Mécanique Céleste et de Calcul des Ephémérides

Optical flow estimation is used in many embedded computer vision applications, and it is known to be computationally intensive. In the literature, many methods exist to estimate optical flow. Thus, the challenge is to find a method that matches the applicative constraints. In an embedded system, a trade-off between power consumption and execution time has to be made to meet both energy and framerate constraints. This work proposes methods to implement an approximate HORN & S CHUNCK optical flow estimation that meets embedded CPUs constraints. This is achieved thanks to architectural optimizations, software optimizations and algorithm tuning. For instance, on the NVIDIA Jetson Nano, and for HD video sequences, the achieved frame latency is 12 ms for 5 Watts. To the best of our knowledge, this is the fastest optical flow implementation on embedded CPUs.

Mots clés: computer vision; optical flow; SIMD; approximate computing; low power; tradeoffs; embedded systems; computer vision optical flow SIMD approximate computing low power tradeofs embedded systems; tradeofs

Domaines: Informatique [cs]; Vision par ordinateur et reconnaissance de formes [cs.CV]; Systèmes embarqués

Maxime Millet, Adrien Cassagne, Nicolas Rambaux, Lionel Lacassagne. Real-Time and Approximate Iterative Optical Flow Implementation on Low-Power Embedded CPUs. 2023 IEEE 34th International Conference on Application-specific Systems, Architectures and Processors (ASAP), Jul 2023, Porto, Portugal. pp.135-138, ⟨10.1109/ASAP57973.2023.00032⟩. ⟨hal-04247806⟩