CoChat

Object Detection in Very High-Resolution Aerial Images Using One-Stage Densely Connected Feature Pyramid Network

Open access
Tayara, Hilal, Chong, Kil To · 2018 · Sensors

The DOI matches its Crossref record; 49 of the 49 references it cites were found in a registry or an index.

Abstract

Object detection in very high-resolution (VHR) aerial images is an essential step for a wide range of applications such as military applications, urban planning, and environmental management. Still, it is a challenging task due to the different scales and appearances of the objects. On the other hand, object detection task in VHR aerial images has improved remarkably in recent years due to the achieved advances in convolution neural networks (CNN). Most of the proposed methods depend on a two-stage approach, namely: a region proposal stage and a classification stage such as Faster R-CNN. Even though two-stage approaches outperform the traditional methods, their optimization is not easy and they are not suitable for real-time applications. In this paper, a uniform one-stage model for object detection in VHR aerial images has been proposed. In order to tackle the challenge of different scales, a densely connected feature pyramid network has been proposed by which high-level multi-scale semantic feature maps with high-quality information are prepared for object detection. This work has been evaluated on two publicly available datasets and outperformed the current state-of-the-art results on both in terms of mean average precision (mAP) and computation time.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/s18103341

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Comparison between the scales of the objects in natural images given by COCO dataset (a) and the scale of the objects in VHR aerial images given by NWPU VHR-10 dataset (b). It can be seen that the vehicles in natural images occupy a larger area compared with the vehicles in VHR aerial images.
Figure 1 Comparison between the scales of the objects in natural images given by COCO dataset (a) and the scale of the objects in VHR aerial images given by NWPU VHR-10 dataset (b). It can be seen that the vehicles in natural images occupy a larger area compared with the vehicles in VHR aerial images.CC BY · PubMed Central
Figure 2 The overall architecture of the proposed model.
Figure 2 The overall architecture of the proposed model.CC BY · PubMed Central
Figure 3 The architecture of the densely connected feature pyramid network.
Figure 3 The architecture of the densely connected feature pyramid network.CC BY · PubMed Central
Figure 4 The architecture of classification and regression heads.
Figure 4 The architecture of classification and regression heads.CC BY · PubMed Central
Figure 5 Examples of data augmentation technique. First row represents the input images while the second and third rows represent the augmented output.
Figure 5 Examples of data augmentation technique. First row represents the input images while the second and third rows represent the augmented output.CC BY · PubMed Central
Figure 6 Detection results of the proposed model in terms of AP using different backbones: VGG-16, Resnet 50, and Resnet 101.
Figure 6 Detection results of the proposed model in terms of AP using different backbones: VGG-16, Resnet 50, and Resnet 101.CC BY · PubMed Central
Figure 7 Comparison of area under precision-recall curve with different state-of-the-art models.
Figure 7 Comparison of area under precision-recall curve with different state-of-the-art models.CC BY · PubMed Central
Figure 8 Some object detection results from NWPU VHR-10 dataset. Yellow, red, and blue colors represent true positive, false negative, and false positive cases, respectively. (a) airplane, (b) ship, (c) storage tank, (d) baseball diamond, (e) tennis court, (f) basketball court, (g) ground track field, (h) harbor, (i) bridge, (j) vehicle, (k–o) show some false positive and false negative cases.
Figure 8 Some object detection results from NWPU VHR-10 dataset. Yellow, red, and blue colors represent true positive, false negative, and false positive cases, respectively. (a) airplane, (b) ship, (c) storage tank, (d) baseball diamond, (e) tennis court, (f) basketball court, (g) ground track field, (h) harbor, (i) bridge, (j) vehicle, (k–o) show some false positive and false negative cases.CC BY · PubMed Central
Figure 9 Some object detection results from RSOD dataset. Yellow, red, and blue colors represent true positive, false negative, and false positive cases, respectively. (a–c) show examples of true positive detection of oil tank, (d–f) show examples of true positive detection of overpass, (g–i) show examples of true positive detection of playground, (j–l) show examples of true positive detection of aircraft, and (m–o) show examples of false positive and false negative cases.
Figure 9 Some object detection results from RSOD dataset. Yellow, red, and blue colors represent true positive, false negative, and false positive cases, respectively. (a–c) show examples of true positive detection of oil tank, (d–f) show examples of true positive detection of overpass, (g–i) show examples of true positive detection of playground, (j–l) show examples of true positive detection of aircraft, and (m–o) show examples of false positive and false negative cases.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 49 references

Research works printed but not linked: 2

  1. Barrett, E. (2003). Introduction to Environmental Remote Sensing, Routledge.
  2. Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.-Y., and Berg, A.C. (2018, October 04). Ssd: Single Shot Multibox Detector. Available online: https://link.springer.com/chapter/10.1007%2F978-3-319-46448-0_2. · DOI · Paper page

Cited by

See all 117

Metadata

Authors
Tayara, H., & Chong, K. T.
Affiliations
Jeonbuk National University (KR)
Corresponding author
Chong, Kil To
Venue
Sensors
Publisher
MDPI AG
Published
2018-10-06
Volume
18
Issue
10
Pages
3341
ISSN
1424-8220 (ISSN-L)
OpenAlex
W2897898606
Cited by
117
References
49
Citations by year
2026: 4 · 2025: 6 · 2024: 11 · 2023: 10 · 2022: 5 · 2021: 34 · 2020: 27 · 2019: 18 · 2018: 2
FWCI
10.13
Type
journal article
Language
en
Open access
Locations
  • Sensors (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
National Research Foundation (NRF) funded by the Korean government (MSIT) (NRF-2017M3C7A1044815)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Remote-Sensing Image Classification
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
Keywords
object detection, one-stage object detection, multi-scale feature maps, real-time detection, scale variation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Object Detection in Very High-Resolution Aerial Images Using One-Stage Densely Connected Feature Pyramid Network. https://app.cochat.ai/papers/10.3390/s18103341
The paper
Tayara, H., & Chong, K. T. (2018). Object Detection in Very High-Resolution Aerial Images Using One-Stage Densely Connected Feature Pyramid Network. Sensors. https://doi.org/10.3390/s18103341

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.