Diao, Y, Tang, X, Wang, H orcid.org/0000-0002-2281-5679 et al. (4 more authors) (2022) A large-scale container dataset and a baseline method for container hole localization. Journal of Real-Time Image Processing, 19 (3). pp. 577-589. ISSN 1861-8200
Abstract
Automatic container handling plays an important role in improving the efficiency of the container terminal, promoting the globalization of container trade, and ensuring worker safety. Utilizing vision-based methods to assist container handling has recently drawn attention. However, most existing keyhole detection/localization methods still suffer from coarse keyhole boundaries. To solve this problem, we propose a real-time container hole localization algorithm based on a modified salient object segmentation network. Note that there exists no public container dataset for researchers to fairly compare their approaches, which has hindered the advances of related algorithms in this domain. Therefore, we propose the first large-scale container dataset in this work, containing 1700 container images and 4810 container hole images, for benchmarking container hole location and detection. Through extensive quantitative evaluation and computational complexity analysis, we show our method can simultaneously achieve superior results on precision and real-time performance. Especially, the detection and location precision is 100% and 99.3%, surpassing the state-of-the-art-work by 2% and 62% respectively. Further, our proposed method only consumes 70 ms (on GPU) or 1.27s (on CPU) per image. We hope the baseline approach, the first released dataset will help benchmark future work and follow-up research on automatic container handling. The dataset is available at https://github.com/qkicen/A-large-scale-container-dataset-and-a-baseline-method-for-container-hole-localization.
Metadata
Item Type: | Article |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2022. This version of the article has been accepted for publication, after peer review (when applicable) and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/s11554-022-01199-y. Uploaded in accordance with the publisher's self-archiving policy. |
Keywords: | Container keyhole localization; Salient object segmentation; Deep learning; Container dataset |
Dates: |
|
Institution: | The University of Leeds |
Academic Units: | The University of Leeds > Faculty of Engineering & Physical Sciences (Leeds) > School of Computing (Leeds) |
Depositing User: | Symplectic Publications |
Date Deposited: | 01 Apr 2022 15:11 |
Last Modified: | 02 Mar 2023 01:13 |
Status: | Published |
Publisher: | Springer |
Identification Number: | 10.1007/s11554-022-01199-y |
Related URLs: | |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:185339 |