Event-frame object detection under dynamic background condition
Neuromorphic vision sensors (NVS) with the features of small data redundancy and transmission latency are widely implemented in Internet of Things applications. Previous studies have developed various object detection algorithms based on NVS’s unique event data format. However, most of these methods...
Saved in:
Main Authors: | , , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/180590 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-180590 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1805902024-10-18T15:41:54Z Event-frame object detection under dynamic background condition Lu, Wenhao Li, Zehao Li, Junying Lu, Yuncheng Kim, Tony Tae-Hyoung School of Electrical and Electronic Engineering Engineering Neuromorphic vision sensor Event data Neuromorphic vision sensors (NVS) with the features of small data redundancy and transmission latency are widely implemented in Internet of Things applications. Previous studies have developed various object detection algorithms based on NVS’s unique event data format. However, most of these methods are only adaptive for scenarios with stationary backgrounds. Under dynamic background conditions, NVS can also acquire the events of non-target objects due to its mechanism of detecting pixel intensity changes. As a result, the performance of existing detection methods is greatly degraded. To address this shortcoming, we introduce an extra refinement process to the conventional histogram-based (HIST) detection method. For the proposed regions from HIST, we apply a practical decision condition to categorize them as either object-dominant or background-dominant cases. Then, the object-dominant regions undergo a second-time HIST-based region proposal for precise localization, while background-dominant regions employ an upper outline determination strategy for target object identification. Finally, the refined results are tracked using a simplified Kalman filter approach. Evaluated in an outdoor drone surveillance with an event camera, the proposed scheme demonstrates superior performance in both intersection over union and F 1 score metrics compared to other methods. Published version This work was supported by ST Engineering Advanced Networks and Sensors Pte. Ltd. 2024-10-14T05:30:38Z 2024-10-14T05:30:38Z 2024 Journal Article Lu, W., Li, Z., Li, J., Lu, Y. & Kim, T. T. (2024). Event-frame object detection under dynamic background condition. Journal of Electronic Imaging, 33(4), 043028-. https://dx.doi.org/10.1117/1.JEI.33.4.043028 1017-9909 https://hdl.handle.net/10356/180590 10.1117/1.JEI.33.4.043028 2-s2.0-85203182725 4 33 043028 en Journal of Electronic Imaging © 2024 SPIE and IS&T. All rights reserved. This article may be downloaded for personal use only. Any other use requires prior permission of the copyright holder. The Version of Record is available online at http://doi.org/10.1117/1.JEI.33.4.043028 application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Engineering Neuromorphic vision sensor Event data |
spellingShingle |
Engineering Neuromorphic vision sensor Event data Lu, Wenhao Li, Zehao Li, Junying Lu, Yuncheng Kim, Tony Tae-Hyoung Event-frame object detection under dynamic background condition |
description |
Neuromorphic vision sensors (NVS) with the features of small data redundancy and transmission latency are widely implemented in Internet of Things applications. Previous studies have developed various object detection algorithms based on NVS’s unique event data format. However, most of these methods are only adaptive for scenarios with stationary backgrounds. Under dynamic background conditions, NVS can also acquire the events of non-target objects due to its mechanism of detecting pixel intensity changes. As a result, the performance of existing detection methods is greatly degraded. To address this shortcoming, we introduce an extra refinement process to the conventional histogram-based (HIST) detection method. For the proposed regions from HIST, we apply a practical decision condition to categorize them as either object-dominant or background-dominant cases. Then, the object-dominant regions undergo a second-time HIST-based region proposal for precise localization, while background-dominant regions employ an upper outline determination strategy for target object identification. Finally, the refined results are tracked using a simplified Kalman filter approach. Evaluated in an outdoor drone surveillance with an event camera, the proposed scheme demonstrates superior performance in both intersection over union and F 1 score metrics compared to other methods. |
author2 |
School of Electrical and Electronic Engineering |
author_facet |
School of Electrical and Electronic Engineering Lu, Wenhao Li, Zehao Li, Junying Lu, Yuncheng Kim, Tony Tae-Hyoung |
format |
Article |
author |
Lu, Wenhao Li, Zehao Li, Junying Lu, Yuncheng Kim, Tony Tae-Hyoung |
author_sort |
Lu, Wenhao |
title |
Event-frame object detection under dynamic background condition |
title_short |
Event-frame object detection under dynamic background condition |
title_full |
Event-frame object detection under dynamic background condition |
title_fullStr |
Event-frame object detection under dynamic background condition |
title_full_unstemmed |
Event-frame object detection under dynamic background condition |
title_sort |
event-frame object detection under dynamic background condition |
publishDate |
2024 |
url |
https://hdl.handle.net/10356/180590 |
_version_ |
1814777708059557888 |