Effectiveness of Human Detection from Aerial Images Taken from Different Heights

Recently, drones have been regularly used to aid in search and rescue in places where it is normally to carry out some of the early forensic victim localization. There are many suitable human detectors for drone use, such as Histogram Oriented Gradient (HOG), You Only Looks Once (YOLO), and Aggregat...

Full description

Bibliographic Details
Published in:TEM Journal
Main Author: Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
Format: Article
Language:English
Published: UIKTEN - Association for Information Communication Technology Education and Science 2021
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85107725476&doi=10.18421%2fTEM102-06&partnerID=40&md5=1cd84c3b5eb439d27cb33c1470e1d50b
id 2-s2.0-85107725476
spelling 2-s2.0-85107725476
Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
Effectiveness of Human Detection from Aerial Images Taken from Different Heights
2021
TEM Journal
10
2
10.18421/TEM102-06
https://www.scopus.com/inward/record.uri?eid=2-s2.0-85107725476&doi=10.18421%2fTEM102-06&partnerID=40&md5=1cd84c3b5eb439d27cb33c1470e1d50b
Recently, drones have been regularly used to aid in search and rescue in places where it is normally to carry out some of the early forensic victim localization. There are many suitable human detectors for drone use, such as Histogram Oriented Gradient (HOG), You Only Looks Once (YOLO), and Aggregate Channel Features (ACF). In this paper, the height of the aerial images is analyzed for its effect on the accuracy of the detection. This works compares ACF, YOLO MobileNet, and YOLO ResNet50 using a different set of aerial images varying at 10m, 20m, and 30m heights. The results show that in a single-model test, with our proposed bounding-box standardization, YOLO MobileNet achieves significant increase in test precision (AP), with 0.7 AP recorded. For single-model test, YOLO MobileNet yield best AP using 20m training data where it obtained AP of 0.88 (10m test height), 0.82 (20m test height), and 0.91 (30m test height). © 2021. Muhammad Shahir Hakimy Salem, Fadhlan Hafizhelmi Kamaru Zaman, Nooritawati Md Tahir; published by UIKTEN. This work is licensed under the Creative Commons Attribution‐NonCommercial‐NoDerivs 4.0 License.
UIKTEN - Association for Information Communication Technology Education and Science
22178309
English
Article
All Open Access; Gold Open Access
author Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
spellingShingle Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
Effectiveness of Human Detection from Aerial Images Taken from Different Heights
author_facet Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
author_sort Salem M.S.H.; Zaman F.H.K.; Tahir N.M.
title Effectiveness of Human Detection from Aerial Images Taken from Different Heights
title_short Effectiveness of Human Detection from Aerial Images Taken from Different Heights
title_full Effectiveness of Human Detection from Aerial Images Taken from Different Heights
title_fullStr Effectiveness of Human Detection from Aerial Images Taken from Different Heights
title_full_unstemmed Effectiveness of Human Detection from Aerial Images Taken from Different Heights
title_sort Effectiveness of Human Detection from Aerial Images Taken from Different Heights
publishDate 2021
container_title TEM Journal
container_volume 10
container_issue 2
doi_str_mv 10.18421/TEM102-06
url https://www.scopus.com/inward/record.uri?eid=2-s2.0-85107725476&doi=10.18421%2fTEM102-06&partnerID=40&md5=1cd84c3b5eb439d27cb33c1470e1d50b
description Recently, drones have been regularly used to aid in search and rescue in places where it is normally to carry out some of the early forensic victim localization. There are many suitable human detectors for drone use, such as Histogram Oriented Gradient (HOG), You Only Looks Once (YOLO), and Aggregate Channel Features (ACF). In this paper, the height of the aerial images is analyzed for its effect on the accuracy of the detection. This works compares ACF, YOLO MobileNet, and YOLO ResNet50 using a different set of aerial images varying at 10m, 20m, and 30m heights. The results show that in a single-model test, with our proposed bounding-box standardization, YOLO MobileNet achieves significant increase in test precision (AP), with 0.7 AP recorded. For single-model test, YOLO MobileNet yield best AP using 20m training data where it obtained AP of 0.88 (10m test height), 0.82 (20m test height), and 0.91 (30m test height). © 2021. Muhammad Shahir Hakimy Salem, Fadhlan Hafizhelmi Kamaru Zaman, Nooritawati Md Tahir; published by UIKTEN. This work is licensed under the Creative Commons Attribution‐NonCommercial‐NoDerivs 4.0 License.
publisher UIKTEN - Association for Information Communication Technology Education and Science
issn 22178309
language English
format Article
accesstype All Open Access; Gold Open Access
record_format scopus
collection Scopus
_version_ 1809677783478042624