Low-cost real-time VLSI system for high-accuracy optical flow estimation using biological motion features and random forests  

作　　者：Cong SHI Junxian HE Shrinivas PUNDLIK Xichuan ZHOU Nanjian WU Gang LUO 

机构地区：[1]School of Microelectronics and Communication Engineering,Chongqing University,Chongqing 400044,China [2]Schepens Eye Research Institute,Massachusetts Eye and Ear,Harvard Medical School,Boston MA 02115,USA [3]State Key Laboratory for Superlattices and Microstructures,Institute of Semiconductors,Chinese Academy of Sciences,Beijing 100083,China

出　　处：《Science China(Information Sciences)》2023年第5期297-298,共2页中国科学（信息科学）（英文版）

基　　金：supported in part by National Key Research and Development Program of China(Grant No.2019YFB2204303);Science and Technology on Analogue Integrated Circuit Laboratory(Grant No.JCKY2019210C060);National Natural Science Foundation of China(Grant No.61434004);Key Project of Chongqing Science and Technology Foundation(Grant Nos.cstc2019jcyj-zdxm X0017,cstc2021ycjh-bgzxm0031)。

摘　　要：This study proposes a low-cost,real-time,very large-scale integration(VLSI)architecture for optical flow estimation.The architecture adopts parallel spatiotemporal filters to extract bio-inspired motion features at each pixel location and uses hardware random forests to infer the motion speed.Our system achieves higher estimation accuracy at low computational hardware costs under real-time constraint than previous biological motion estimation systems.

关 键 词：HARDWARE estimation MOTION 

分 类 号：TN40[电子电信—微电子学与固体电子学] TP181[自动化与计算机技术—控制理论与控制工程]