Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery

Most Convolution Neural Network (CNN) based object detectors, to date, have been optimized for accuracy and/or detection performance on datasets typically comprised of well exposed 8-bits/pixel/channel Standard Dynamic Range (SDR) images. A major existing challenge in this area is to accurately dete...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ratnajit Mukherjee, Maximino Bessa, Pedro Melo-Pinto, Alan Chalmers
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	High dynamic range (HDR) low dynamic range (SDR) multiple object detection faster RCNN SSD R-FCN Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/6fd7a2dddec745eebbd50455c851607f
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:6fd7a2dddec745eebbd50455c851607f
record_format	dspace
spelling	oai:doaj.org-article:6fd7a2dddec745eebbd50455c851607f2021-11-23T00:00:53ZObject Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery2169-353610.1109/ACCESS.2021.3082293https://doaj.org/article/6fd7a2dddec745eebbd50455c851607f2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9436775/https://doaj.org/toc/2169-3536Most Convolution Neural Network (CNN) based object detectors, to date, have been optimized for accuracy and/or detection performance on datasets typically comprised of well exposed 8-bits/pixel/channel Standard Dynamic Range (SDR) images. A major existing challenge in this area is to accurately detect objects under extreme/difficult lighting conditions as SDR image trained detectors fail to accurately detect objects under such challenging lighting conditions. In this paper, we address this issue for the first time by introducing High Dynamic Range (HDR) imaging to object detection. HDR imagery can capture and process ≈13 orders of magnitude of scene dynamic range similar to the human eye. HDR trained models are therefore able to extract more salient features from extreme lighting conditions leading to more accurate detections. However, introducing HDR also presents multiple new challenges such as the complete absence of resources and previous literature on such an approach. Here, we introduce a methodology to generate a large scale annotated HDR dataset from any existing SDR dataset and validate the quality of the generated dataset via a robust evaluation technique. We also discuss the challenges of training and validating HDR trained models using existing detectors. Finally, we provide a methodology to create an out of distribution (OOD) HDR dataset to test and compare the performance of HDR and SDR trained detectors under difficult lighting condition. Results suggest that using the proposed methodology, HDR trained models are able to achieve 10 – 12% more accuracy compared to SDR trained models on real-world OOD dataset consisting of high-contrast images under extreme lighting conditions.Ratnajit MukherjeeMaximino BessaPedro Melo-PintoAlan ChalmersIEEEarticleHigh dynamic range (HDR)low dynamic range (SDR)multiple object detectionfaster RCNNSSDR-FCNElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 77771-77783 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	High dynamic range (HDR) low dynamic range (SDR) multiple object detection faster RCNN SSD R-FCN Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	High dynamic range (HDR) low dynamic range (SDR) multiple object detection faster RCNN SSD R-FCN Electrical engineering. Electronics. Nuclear engineering TK1-9971 Ratnajit Mukherjee Maximino Bessa Pedro Melo-Pinto Alan Chalmers Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
description	Most Convolution Neural Network (CNN) based object detectors, to date, have been optimized for accuracy and/or detection performance on datasets typically comprised of well exposed 8-bits/pixel/channel Standard Dynamic Range (SDR) images. A major existing challenge in this area is to accurately detect objects under extreme/difficult lighting conditions as SDR image trained detectors fail to accurately detect objects under such challenging lighting conditions. In this paper, we address this issue for the first time by introducing High Dynamic Range (HDR) imaging to object detection. HDR imagery can capture and process ≈13 orders of magnitude of scene dynamic range similar to the human eye. HDR trained models are therefore able to extract more salient features from extreme lighting conditions leading to more accurate detections. However, introducing HDR also presents multiple new challenges such as the complete absence of resources and previous literature on such an approach. Here, we introduce a methodology to generate a large scale annotated HDR dataset from any existing SDR dataset and validate the quality of the generated dataset via a robust evaluation technique. We also discuss the challenges of training and validating HDR trained models using existing detectors. Finally, we provide a methodology to create an out of distribution (OOD) HDR dataset to test and compare the performance of HDR and SDR trained detectors under difficult lighting condition. Results suggest that using the proposed methodology, HDR trained models are able to achieve 10 – 12% more accuracy compared to SDR trained models on real-world OOD dataset consisting of high-contrast images under extreme lighting conditions.
format	article
author	Ratnajit Mukherjee Maximino Bessa Pedro Melo-Pinto Alan Chalmers
author_facet	Ratnajit Mukherjee Maximino Bessa Pedro Melo-Pinto Alan Chalmers
author_sort	Ratnajit Mukherjee
title	Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
title_short	Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
title_full	Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
title_fullStr	Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
title_full_unstemmed	Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery
title_sort	object detection under challenging lighting conditions using high dynamic range imagery
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/6fd7a2dddec745eebbd50455c851607f
work_keys_str_mv	AT ratnajitmukherjee objectdetectionunderchallenginglightingconditionsusinghighdynamicrangeimagery AT maximinobessa objectdetectionunderchallenginglightingconditionsusinghighdynamicrangeimagery AT pedromelopinto objectdetectionunderchallenginglightingconditionsusinghighdynamicrangeimagery AT alanchalmers objectdetectionunderchallenginglightingconditionsusinghighdynamicrangeimagery
_version_	1718417380420354048

Object Detection Under Challenging Lighting Conditions Using High Dynamic Range Imagery

Ejemplares similares