An 81.6 GOPS Object Recognition Processor Based on NoC and Visual Image Processing Memory

Author

Kim, Donghyun ; Kim, Kwanho ; Kim, Joo-Young ; Lee, Seungjin ; Yoo, Hoi-Jun

Author_Institution

Korea Adv. Inst. of Sci. & Technol., Daejeon

fYear

2007

fDate

16-19 Sept. 2007

Firstpage

443

Lastpage

446

Abstract

An 81.6 GOPS object recognition processor is developed by using NoC and visual image processing (VIP) memory. SIFT (scale invariant feature transform) object recognition requires huge computing power and data transactions among tasks. The chip integrates 10 SIMD PEs for data/task level parallelism while the NoC facilitates inter-PE communications. The VIP memory searches local maximum pixel inside a 3times3 window in a single cycle providing 65.6 GOPS. The proposed processor achieves 15.9 fps SIFT feature extraction at 200 MHz.

Keywords

digital signal processing chips; feature extraction; image recognition; network-on-chip; object recognition; transforms; GOPS object recognition processor; SIMD processing elements; data transactions; feature extraction; frequency 200 MHz; network-on-chip; scale invariant feature transform; visual image processing memory; Computer architecture; Concurrent computing; Energy consumption; Filtering; Image processing; Intelligent robots; Network-on-a-chip; Object recognition; Parallel processing; Vehicle detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Custom Integrated Circuits Conference, 2007. CICC '07. IEEE

Conference_Location

San Jose, CA

Print_ISBN

978-1-4244-1623-3

Electronic_ISBN

978-1-4244-1623-3

Type

conf

DOI

10.1109/CICC.2007.4405769

Filename

4405769