A Vlsi-Implementation-Friendly EGO-Motion Detection Algorithm Based on Edge-Histogram Matching

An ego-motion detection algorithm compatible to hardware implementation has been developed. The algorithm utilizes local motion detection scheme based on edge-histogram matching, which enables us to detect local motions robustly in segmented blocks in a visual field. An 18-dimension motion field vector is generated by summarizing local motions. Then the vector quantization is carried out to recognize the ego-motion. In order to achieve further robustness, two thresholding techniques, block thresholding and median processing, are employed in the procedure. In computer simulation, over 93% of detecting accuracy has been experimentally demonstrated by template matching using 30 template vectors generated from each of four ego-motion types

[1]  Jake K. Aggarwal,et al.  On the computation of motion from sequences of images-A review , 1988, Proc. IEEE.

[2]  Tadashi Shibata,et al.  An image representation algorithm compatible with neural-associative-processor-based hardware recognition systems , 2003, IEEE Trans. Neural Networks.

[3]  T. Shibata,et al.  A high-speed median filter VLSI using floating-gate-MOS-based low-power majority voting circuits , 2005, Proceedings of the 31st European Solid-State Circuits Conference, 2005. ESSCIRC 2005..

[4]  Truong Q. Nguyen,et al.  Global motion estimation in frequency and spatial domain , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[5]  B. Bhanu,et al.  Estimating 3D Egomotion from Perspective Image Sequence , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  T. Shibata,et al.  A real-time image-feature-extraction and vector-generation VLSI employing arrayed-shift-register architecture , 2005, Proceedings of the 31st European Solid-State Circuits Conference, 2005. ESSCIRC 2005..