A 52mW full HD 160-degree object viewpoint recognition SoC with visual vocabulary processor for wearable vision applications

Author

Su, Yu-Chi ; Huang, Keng-Yen ; Chen, Tse-Wei ; Tsai, Yi-Min ; Chien, Shao-Yi ; Chen, Liang-Gee

Author_Institution

Dept. of Electr. Eng., Nat. Taiwan Univ., Taipei, Taiwan

fYear

2011

Firstpage

258

Lastpage

259

Abstract

A wearable 1920×1080 160-degree object viewpoint recognition SoC is realized on a 6.38mm² die with 65nm CMOS technology. This system focuses on enhancing the capability for wide viewpoint and long-distance recognition while reducing the computation of feature matching process. The recognition accuracy is improved from 29% to 94% under full HD resolution for a 50m-far traffic light compared with the performance under VGA (640×480). Object viewpoint prediction (OVP) supports 160-degree object viewpoint differences. 85% of power consumption and 75% of memory bandwidth are reduced via proposed visual vocabulary processor (VVP). 52mW power consumption with 25.9GOPS/mm² area efficiency is achieved.

Keywords

CMOS integrated circuits; image recognition; image resolution; system-on-chip; CMOS technology; distance 50 m; full HD resolution; long-distance recognition; object viewpoint recognition SoC; power 52 mW; size 65 nm; visual vocabulary processor; wearable vision applications; Cameras; Computer architecture; Feature extraction; Histograms; Image recognition; System-on-a-chip; Vocabulary;

fLanguage

English

Publisher

ieee

Conference_Titel

VLSI Circuits (VLSIC), 2011 Symposium on

ISSN

2158-5601

Print_ISBN

978-1-61284-175-5

Type

conf

Filename

5986442