Hexagonal pixel-array for efficient spatial computation for motion-detection pre-processing of visual scenes
DOI:
https://doi.org/10.14738/aivp.22.153Keywords:
Hexagonal pixel-array, rectilinear pixels, geometric optimization, computational efficiency, spatial computation, parallel pre-processingAbstract
Motion-detection, edge-detection, and orientation-detection often require spatial computation of the light intensity difference between neighboring pixel cells. Pre-processing the image at the retinal level can improve the computational efficiency when the parallel processing can be achieved naturally by the spatial arrangement of the pixel-array. Pixel-arrays that require spatial pre-processing computation have different geometric constraints than plain pixel-arrays that do not require such local computation. Our analysis shows that geometric optimization of pixel-arrays can be achieved by using hexagonal arrays over rectilinear arrays. Hexagonal arrays improve the packing density, and reduce the complexity of the spatial computation compared to rectilinear square and octagonal arrays. They also provide geometric symmetry for efficient computation not only at the contiguous neighboring cell level, but also at the higher-order neighboring cell level. The light intensity difference at the higher-order cell level is used to compute the first-order and second-order time-derivatives for velocity and acceleration detections of the visual scene, respectively. Thus, hexagonal arrays increase the computational efficiency by using a symmetric configuration that allows pre-processing of spatial information of the visual scene using hardware implementations that are repeatable in all higher-order neighboring pixels.References
Blum, M. and T. Labhart, Photoreceptor visual fields, ommatidial array, and receptor axon projections in the polarisation-sensitive dorsal rim area of the cricket compound eye. J Comp Physiol A, 2000. 186(2): p. 119-28.
Staunton, R.C., The design of hexagonal sampling structures for image digitization and their use with local operators. Image and Vision Computing, 1989. 7(3): p. 162-166.
Coleman, B.D. and G.H. Renninger, Theory of delayed lateral inhibition in the compound eye of limulus. Proc Natl Acad Sci U S A, 1974. 71(7): p. 2887-91.
Schultz, A.M. and M.J. Wilcox, Parallel image segmentation using the L4 network. Biomed Sci Instrum, 1999. 35: p. 117-21.
Srinivasan, M.V. and G.D. Bernard, The effect of motion on visual acuity of the compound eye: a theoretical analysis. Vision Res, 1975. 15(4): p. 515-25.
Laughlin, S., A simple coding procedure enhances a neuron's information capacity. Z Naturforsch C, 1981. 36(9-10): p. 910-2.
Tam, D.C., Computational efficiency circuitry for detecting shadows in hexagonal array of compound eye. Neurocomputing, 2004. 58–60(0): p. 1073-1078.
Davies, E.R., Optimising computation of hexagonal differential gradient edge detector. Electronics Letters, 1991. 27(17): p. 1526 – 1527.
He, X., et al., Bilateral Edge-detection on a Virtual Hexagonal Structure, in Advances in Visual Computing, G. Bebis, et al., Editors. 2006, Springer Berlin Heidelberg. p. 176-185.
Abu-Baker, S. and R.J. Green. Detection of edges based on hexagonal pixel formats. in Signal Processing, 1996., 3rd International Conference on. 1996.
Staunton, R.C. Hexagonal image sampling: A practical proposition. in Proc. SPIE 1008, Expert Robots for Industrial Use. 1989.
Gardiner, B., S. Coleman, and B. Scotney. Multi-scale Feature Extraction in a Sub-pixel Virtual Hexagonal Environment. in Machine Vision and Image Processing Conference, 2008. IMVIP '08. International. 2008.
Davies, E.R., Low-level vision requirements. Electronics & Communication Engineering Journal, 2000. 12(5): p. 197 – 210.
