Impact of Image Sensor Output Data on Power Consumption of the Image Processing System

Gernot Fiala*, Johannes Loining, Christian Steger

*Corresponding author for this work

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

Abstract

Nowadays, image sensors are widely used for industrial or augmented and virtual reality (AR)/(VR) applications. Often, multiple sensors are used and the host platform needs to process each high resolution image frame with low latency. Therefore, the processing units and the communication have a big impact on the power consumption of the image processing system. This paper describes a method to simulate a smart image sensor and the impact of different output resolutions on the image processing system. In-sensor processing is simulated, by sending different image resolutions to the host platform, where a shape matching algorithm was processed. It shows an analysis of the power consumption of the communication from the sensor with region-of-interest (ROI) data to the host and the energy consumption of the whole image processing system with a Mira220 and an Nvidia Jetson Nano. The power consumption of the MIPI interface can be reduced by over 83% for ROIs smaller than 640×480. Furthermore, the system energy consumption can be reduces by 18% at 90 fps and 21% at 60 fps.
Translated title of the contributionEinfluss von Bildsensordaten auf die Leistungsaufnahme des Bildverarbeitungssystems
Original languageEnglish
Title of host publicationIntelligent Systems and Applications - IntelliSys 2022
Subtitle of host publicationProceedings of the 2022 Intelligent Systems Conference (IntelliSys)
EditorsKohei Arai
PublisherSpringer Nature Switzerland AG
Pages618 - 636
Volume 1
ISBN (Electronic)978-3-031-16072-1
ISBN (Print)978-3-031-16071-4
DOIs
Publication statusPublished - 2022
EventIntelligent Systems Conference 2022: IntelliSys 2022 - Park Plaza Amsterdam Airport, Amsterdam, Netherlands
Duration: 1 Sept 20222 Sept 2022
Conference number: IntelliSys2022
https://saiconference.com/Conferences/IntelliSys2022

Publication series

NameLecture Notes in Networks and Systems
PublisherSpringer Cham
Number1
Volume542
ISSN (Print)2367-3370
ISSN (Electronic)2367-3389

Conference

ConferenceIntelligent Systems Conference 2022
Abbreviated titleIntelliSys 2022
Country/TerritoryNetherlands
CityAmsterdam
Period1/09/222/09/22
Internet address

Keywords

  • Communication bandwidth
  • Energy
  • Mira220
  • Nvidia Jetson Nano
  • Power
  • Region-of-interest
  • Smart image sensor

ASJC Scopus subject areas

  • Computer Science(all)
  • Signal Processing
  • Control and Systems Engineering
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Impact of Image Sensor Output Data on Power Consumption of the Image Processing System'. Together they form a unique fingerprint.

Cite this