A CMOS vision sensor with on-the-fly histogram equalization using TFS encoding and AER read-out

Shoushun Chen*, Amine Bermak, Dominique Martinez

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

This paper presents a CMOS vision sensor performing image capture and on-chip histogram equalization. Both image capture and the processing are based on time-to-first spike encoding combined with Address Event Representation (AER) read-out scheme. The proposed approach not only provides a compact and efficient VLSI implementation but also eliminates voltage/current domain analog processing and hence provides immunity to dynamic range reduction caused by low voltage operation in advanced sub-micron technologies. The pixel values are read-out using an asynchronous handshaking type of read-out while the processing is carried out using simple and yet robust digital timer occupying a very small silicon area (0.1 × 0.6mm2 in 0.35μm CMOS). Low power operation (10nA per pixel) is achieved since the pixels are only allowed to switch once per frame and only active pixels are allowed to asynchronously access the bus. The prototype was implemented in 0.35μm CMOS technology with a silicon area of 3.1 × 3.2mm2.

Original languageEnglish
Title of host publication2005 IEEE Asian Solid-State Circuits Conference, ASSCC 2005
PublisherIEEE Computer Society
Pages433-436
Number of pages4
ISBN (Print)0780391624, 9780780391628
DOIs
Publication statusPublished - 2005
Externally publishedYes
Event1st IEEE Asian Solid-State Circuits Conference, ASSCC 2005 - Hsinchu, Taiwan, Province of China
Duration: 1 Nov 20053 Nov 2005

Publication series

Name2005 IEEE Asian Solid-State Circuits Conference, ASSCC 2005

Conference

Conference1st IEEE Asian Solid-State Circuits Conference, ASSCC 2005
Country/TerritoryTaiwan, Province of China
CityHsinchu
Period1/11/053/11/05

Fingerprint

Dive into the research topics of 'A CMOS vision sensor with on-the-fly histogram equalization using TFS encoding and AER read-out'. Together they form a unique fingerprint.

Cite this