{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T19:02:36Z","timestamp":1778698956186,"version":"3.51.4"},"reference-count":25,"publisher":"MDPI AG","issue":"21","license":[{"start":{"date-parts":[[2022,10,31]],"date-time":"2022-10-31T00:00:00Z","timestamp":1667174400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Artificial Intelligence National Laboratory","award":["RRF-2.3.1-21-2022-00004"],"award-info":[{"award-number":["RRF-2.3.1-21-2022-00004"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>A hologram, measured by using appropriate coherent illumination, records all substantial volumetric information of the measured sample. It is encoded in its interference patterns and, from these, the image of the sample objects can be reconstructed in different depths by using standard techniques of digital holography. We claim that a 2D convolutional network (CNN) cannot be efficient in decoding this volumetric information spread across the whole image as it inherently operates on local spatial features. Therefore, we propose a method, where we extract the volumetric information of the hologram by mapping it to a volume\u2014using a standard wavefield propagation algorithm\u2014and then feed it to a 3D-CNN-based architecture. We apply this method to a challenging real-life classification problem and compare its performance with an equivalent 2D-CNN counterpart. Furthermore, we inspect the robustness of the methods to slightly defocused inputs and find that the 3D method is inherently more robust in such cases. Additionally, we introduce a hologram-specific augmentation technique, called hologram defocus augmentation, that improves the performance of both methods for slightly defocused inputs. The proposed 3D-model outperforms the standard 2D method in classification accuracy both for in-focus and defocused input samples. Our results confirm and support our fundamental hypothesis that a 2D-CNN-based architecture is limited in the extraction of volumetric information globally encoded in the reconstructed hologram image.<\/jats:p>","DOI":"10.3390\/s22218366","type":"journal-article","created":{"date-parts":[[2022,11,2]],"date-time":"2022-11-02T06:49:02Z","timestamp":1667371742000},"page":"8366","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Classification of Holograms with 3D-CNN"],"prefix":"10.3390","volume":"22","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3548-4685","authenticated-orcid":false,"given":"D\u00e1niel","family":"Terbe","sequence":"first","affiliation":[{"name":"Institute for Computer Science and Control, H-1111 Budapest, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"L\u00e1szl\u00f3","family":"Orz\u00f3","sequence":"additional","affiliation":[{"name":"Institute for Computer Science and Control, H-1111 Budapest, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"\u00c1kos","family":"Zar\u00e1ndy","sequence":"additional","affiliation":[{"name":"Institute for Computer Science and Control, H-1111 Budapest, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2022,10,31]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"243902","DOI":"10.1103\/PhysRevLett.121.243902","article-title":"Low photon count phase retrieval using deep learning","volume":"121","author":"Goy","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"17141","DOI":"10.1038\/lsa.2017.141","article-title":"Phase recovery and holographic image reconstruction using deep learning in neural networks","volume":"7","author":"Rivenson","year":"2018","journal-title":"Light Sci. Appl."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"1117","DOI":"10.1364\/OPTICA.4.001117","article-title":"Lensless computational imaging through deep learning","volume":"4","author":"Sinha","year":"2017","journal-title":"Optica"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"704","DOI":"10.1364\/OPTICA.5.000704","article-title":"Extended depth-of-field in holographic imaging using deep-learning-based autofocusing and phase recovery","volume":"5","author":"Wu","year":"2018","journal-title":"Optica"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"136","DOI":"10.1007\/s10043-019-00574-8","article-title":"Analysis of non-iterative phase retrieval based on machine learning","volume":"27","author":"Nishizaki","year":"2020","journal-title":"Opt. Rev."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Dardikman, G., and Shaked, N.T. (2018, January 25\u201328). Phase unwrapping using residual neural networks. Proceedings of the Computational Optical Sensing and Imaging. Optical Society of America, Orlando, FL, USA.","DOI":"10.1364\/COSI.2018.CW3B.5"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"54","DOI":"10.1109\/LSP.2018.2879184","article-title":"PhaseNet: A deep convolutional neural network for two-dimensional phase unwrapping","volume":"26","author":"Spoorthi","year":"2018","journal-title":"IEEE Signal Process. Lett."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"294","DOI":"10.1021\/acsphotonics.8b01479","article-title":"Deep learning enables high-throughput analysis of particle-aggregation-based biosensors imaged using holography","volume":"6","author":"Wu","year":"2018","journal-title":"ACS Photonics"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"4617","DOI":"10.1021\/acsphotonics.8b01109","article-title":"Label-free bioaerosol sensing using mobile microscopy and deep learning","volume":"5","author":"Wu","year":"2018","journal-title":"ACS Photonics"},{"key":"ref_10","first-page":"1","article-title":"A deep learning-enabled portable imaging flow cytometer for cost-effective, high-throughput, and label-free analysis of natural water samples","volume":"7","author":"Tamamitsu","year":"2018","journal-title":"Light Sci. Appl."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"3926","DOI":"10.1038\/s41598-019-40554-1","article-title":"Deep learning-based super-resolution in coherent imaging systems","volume":"9","author":"Liu","year":"2019","journal-title":"Sci. Rep."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Shimobaba, T., Kakue, T., and Ito, T. (2018, January 13\u201315). Convolutional neural network-based regression for depth prediction in digital holography. Proceedings of the 2018 IEEE 27th International Symposium on Industrial Electronics (ISIE), Cairns, Australia.","DOI":"10.1109\/ISIE.2018.8433651"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/s41377-019-0139-9","article-title":"Bright-field holography: Cross-modality deep learning enables snapshot 3D imaging with bright-field contrast using a single hologram","volume":"8","author":"Wu","year":"2019","journal-title":"Light Sci. Appl."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"5567","DOI":"10.1364\/OL.440900","article-title":"Deep-learning-based bright-field image generation from a single hologram using an unpaired dataset","volume":"46","author":"Terbe","year":"2021","journal-title":"Opt. Lett."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"17003","DOI":"10.1038\/s41598-018-35274-x","article-title":"Deep transfer learning-based hologram classification for molecular diagnostics","volume":"8","author":"Kim","year":"2018","journal-title":"Sci. Rep."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Zhu, Y., Yeung, C.H., and Lam, E.Y. (2020, January 20\u201323). Holographic classifier: Deep learning in digital holography for automatic micro-objects classification. Proceedings of the 2020 IEEE 18th International Conference on Industrial Informatics (INDIN), Warwick, UK.","DOI":"10.1109\/INDIN45582.2020.9442146"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"024013","DOI":"10.1088\/2515-7647\/abf250","article-title":"Microplastic pollution monitoring with holographic classification and deep learning","volume":"3","author":"Zhu","year":"2021","journal-title":"J. Phys. Photonics"},{"key":"ref_18","first-page":"1","article-title":"Plankton classification with high-throughput submersible holographic microscopy and transfer learning","volume":"21","author":"MacNeil","year":"2021","journal-title":"BMC Ecol. Evol."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Alakwaa, W., Nassef, M., and Badr, A. (2017). Lung cancer detection and classification with 3D convolutional neural network (3D-CNN). Int. J. Adv. Comput. Sci. Appl., 8.","DOI":"10.14569\/IJACSA.2017.080853"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Singh, S.P., Wang, L., Gupta, S., Goli, H., Padmanabhan, P., and Guly\u00e1s, B. (2020). 3D deep learning on medical images: A review. Sensors, 20.","DOI":"10.3390\/s20185097"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"3146","DOI":"10.3390\/rs14133146","article-title":"2D&3DHNet for 3D Object Classification in LiDAR Point Cloud","volume":"14","author":"Song","year":"2022","journal-title":"Remote Sens."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"109297","DOI":"10.1016\/j.measurement.2021.109297","article-title":"A 3D convolutional neural network based near-field acoustical holography method with sparse sampling rate on measuring surface","volume":"177","author":"Wang","year":"2021","journal-title":"Measurement"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"111790","DOI":"10.1016\/j.measurement.2022.111790","article-title":"Research on joint training strategy for 3D convolutional neural network based near-field acoustical holography with optimized hyperparameters","volume":"202","author":"Wang","year":"2022","journal-title":"Measurement"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"19662","DOI":"10.1364\/OE.17.019662","article-title":"Band-limited angular spectrum method for numerical simulation of free-space propagation in far and near fields","volume":"17","author":"Matsushima","year":"2009","journal-title":"Opt. Express"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Chen, H., Huang, L., Liu, T., and Ozcan, A. (2022). Fourier Imager Network (FIN): A deep neural network for hologram reconstruction with superior external generalization. arXiv.","DOI":"10.1117\/12.2648180"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/22\/21\/8366\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T01:07:04Z","timestamp":1760144824000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/22\/21\/8366"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,31]]},"references-count":25,"journal-issue":{"issue":"21","published-online":{"date-parts":[[2022,11]]}},"alternative-id":["s22218366"],"URL":"https:\/\/doi.org\/10.3390\/s22218366","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,31]]}}}