Real-Time Quantized Optical Flow

Algorithms based on the correlation of image patches can be robust in practice but are computationally intensive due to the computational complexity of their search-based nature. Performing the search over time instead of over space is linear in nature, rather than quadratic, and results in a very eecient algorithm. This, combined with implementations which are highly eecient on standard computing hardware, yields performance of 9 frames per second on a scientiic workstation. Although the resulting velocities are quan-tized with resulting quantization error, they have been shown to be suuciently accurate for many robotic vision tasks such as time-to-collision and robotic navigation. Thus, this algorithm is highly suitable for real-time robotic vision research.