Abstract
As herbarium specimens are largely digitized and freely available in online portals, botanists aim to examine their taxonomic aspects to identify the plant specimen regions and generate their morphological data. Nevertheless, different uninformative visual information within the digitized herbarium specimen, such as scale-bar, color pallet, specimen label, envelopes, bar-code, and stamp, represent a source of visual noise. Thus, their identification requires unique detection methods as they are mostly placed at different locations and orientations within the herbarium sheet. Given a collection of digitized herbarium specimen images gathered from the Herbarium Haussknecht of Jena, Germany, we present in this paper a deep learning-based approach for specimen image semantic segmentation. Two different pipelines were involved in this work: (i) coarse segmentation and (ii) fine segmentation. Throughout the whole process, we describe the ground truth annotation used for training our deep learning architecture. The experimental results demonstrate that our proposed model outperforms the other architectures such as SegNet, Squeeze-SegNet, U-Net, and DeepLabv3. Its accuracy achieves 91% compared to 82%, 80%, 86%, and 90% obtained by SegNet, Squeeze-SegNet, U-Net, and DeepLabv3, respectively.
Similar content being viewed by others
References
Badrinarayanan V, Kendall A, Cipolla R (1 Dec. 2017) SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Trans Pattern Anal Mach Intell 39(12):2481–2495. https://doi.org/10.1109/TPAMI.2016.2644615
Beghin T, Cope JS, Remagnino P, Barman S (2010) Shape and Texture Based Plant Leaf Classification. In: Blanc-Talon J, Bone D, Philips W, Popescu D, Scheunders P (eds) Advanced Concepts for Intelligent Vision Systems. ACIVS 2010. Lecture notes in computer science, vol 6475. Springer, Berlin. https://doi.org/10.1007/9783-642-17691-3-32
Borges LM, Reis VC, Izbicki R (2020) Schr¨odinger’s phenotypes: herbarium specimens show two-dimensional images are both good and (not so) bad sources of morphological data. Methods Ecol Evol 11:1296–1308. https://doi.org/10.1111/2041-210X.13450
Borsch T, Stevens A-D, Häffner E, Güntsch A, Berendsohn WG, Appelhans MS, Barilaro C, Beszteri B, Blattner FR, Bossdorf O, Dalitz H, Dressler S, Duque-Thüs R, Esser H-J, Franzke A, Goetze D, Grein M, Grünert U, Hellwig F, … Zizka G (2020) A complete digitization of German herbaria is possible, sensible and should be started now. Res Ideas Outcomes 6:e50675. https://doi.org/10.3897/rio.6.e50675
Canny J, (1986) “A Computational Approach to Edge Detection,” in IEEE Trans Pattern Anal Mach Intell, vol. PAMI-8, no. 6, pp. 679–698, https://doi.org/10.1109/TPAMI.1986.4767851.
Caselles V, Kimmel R, Guillermo Sapiro G (1997) Geodesic Active Contours. Int J Comput Vis 22(61–79):61–79
Cerutti G, Tougne L, Mille J, Vacavant A, Coquin D (2013) Understanding leaves in natural images – a model-based approach for tree species identification. Comput Vis Image Underst 117(10):1482–1501. https://doi.org/10.1016/j.cviu.2013.07.003 ISSN 1077-3142, http://www.sciencedirect.com/science/article/pii/S1077314213001331
Chen L, Papandreou G, Schroff F, Adam H (2017) Rethinking atrous convolution for semantic image segmentation. ArXiv, abs/1706.05587
Figueira R, Lages F (2019) Museum and Herbarium Collections for Biodiversity Research in Angola. Springer International Publishing. https://doi.org/10.1007/978-3-030-03083-4-19
Kommineni VK, Kattge J, Gaikwad J, Baddam P, Tautenhahn S (2020) Understanding intraspecific trait variability using digital herbarium specimen images. Biodivers Inf Sci Stand 4:e59061. https://doi.org/10.3897/biss.4.59061
Kumar N, Belhumeur PN, Biswas A, Jacobs DW, Kress WJ, Lopez IC, Soares JVB (2012) editor=“Fitzgibbon, Andrew and Lazebnik, Svetlana and Perona, Pietro and Sato, Yoichi and Schmid, Cordelia. Leafsnap: A Computer Vision System for Automatic Plant Species Identification. Springer Berlin Heidelberg.
Lee JW, Chan Yoon Y (2019) Fine-Grained Plant Identification using wide and deep learning model 1. 2019 international conference on platform technology and service (PlatCon), Jeju, pp 1–5. https://doi.org/10.1109/PlatCon.2019.8669407
Long J, Shelhamer E, Darrell T (2015) Fully convolutional networks for semantic segmentation. 2015 IEEE conference on computer vision and pattern recognition (CVPR), Boston, pp 3431–3440. https://doi.org/10.1109/CVPR.2015.7298965
Jiang X, Lou S, Scott PJ (2012) Morphological method for surface metrology and dimensional metrology based on the alpha shape. Meas Sci Technol 23:015003
McAllister CA, McKain MR, Li M, Bookout B, Kellogg EA (2018) Specimen based analysis of morphology and the environment in ecologically dominant grasses: the power of the herbarium. Philos Trans R Soc B 374:20170403. https://doi.org/10.1098/rstb.2017.0403
Nanfack G, Elhassouny A, Thami RO (2018) Squeeze-SegNet: a new fast deep convolutional neural network for semantic segmentation. International Conference on Machine Vision
Noh H, Hong S, Han B (2015) Learning Deconvolution Network for Semantic Segmentation. 2015 IEEE international conference on computer vision (ICCV), Santiago, pp 1520–1528. https://doi.org/10.1109/ICCV.2015.178
Olsen A, Konovalov DA, Philippa B, Ridd P, Wood JC, Johns J, Banks W, Girgenti B, Kenny O, Whinney J, Calvert B, Azghadi MR, White RD (2019) DeepWeeds: a multiclass weed species image dataset for deep learning. Sci Rep 9:2058. https://doi.org/10.1038/s41598-018-38343-3
Ronneberger O., Fischer P., Brox T. (2015) U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Navab N., Hornegger J., Wells W., Frangi A. (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture notes in computer science. vol 9351, Springer, Cham https://doi.org/10.1007/978-3-319-24574-4-28
Simonyan K, Zisserman A (2015) Very Deep Convolutional Networks for LargeScale Image Recognition. CoRR abs/1409.1556: n. pag
Singh V, Varsha, Misra AK (2015) Detection of unhealthy region of plant leaves using image processing and genetic algorithm. 2015 International Conference on Advances in Computer Engineering and Applications, Ghaziabad, pp 1028–1032. https://doi.org/10.1109/ICACEA.2015.7164858
Suzuki S, Abe K (1985) Topological structural analysis of digitized binary images by border following. Comput Vision Graph Image Process 30(1):32–46. https://doi.org/10.1016/0734-189X(85)90016-7 ISSN 0734-189X, http://www.sciencedirect.com/science/article/pii/0734189X85900167
Valliammal N, Geethalakshmi S (2012) Plant leaf segmentation using non linear K means clustering
Vukadinovic D, Polder G (2015) Watershed and supervised classification based fully automated method for separate leaf segmentation
Wäldchen J, Rzanny M, Seeland M, Mäder P (2018) Automated plant species identification—trends and future directions. PLoS Comput Biol 14(4):1–19
Wang Q, Gao J, Yuan Y (2018) A Joint Convolutional Neural Networks and Context Transfer for Street Scenes Labeling. IEEE Trans Intell Transp Syst 19(5):1457–1470. https://doi.org/10.1109/TITS.2017.2726546
Wang Q, Gao J, Li X (2019) Weakly Supervised Adversarial Domain Adaptation for Semantic Segmentation in Urban Scenes. IEEE Trans Image Process 28(9):4376–4386. https://doi.org/10.1109/TIP.2019.2910667
Weaver WN, Ng J, Laport RG (2020) LeafMachine: Using machine learning to automate leaf trait extraction from digitized herbarium specimens. Appl Plant Sci 8(6):e11367. https://doi.org/10.1002/aps3.11367
Weaver WN, Ng J, Laport RG (2020) LeafMachine: Using machine learning to automate leaf trait extraction from digitized herbarium specimens. Appl Plant Sci 8(6):e11367. Published 2020 Jul 1. https://doi.org/10.1002/aps3.11367
Wilf P, Zhang S, Chikkerur S, Little SA, Wing SL, Serre T (Mar 2016) Computer vision cracks the leaf code. Proc Natl Acad Sci 113(12):3305–3310. https://doi.org/10.1073/pnas.1524473113
Yu F, Koltun V (2016) Multi-Scale Context Aggregation by Dilated Convolutions. CoRR abs/1511.07122: n. pags
Zhao C, Chan SSF, Cham W-K, Chu LM (2015) Plant identification using leaf shapes—A pattern counting approach. Pattern Recogn 48(10):3203–3215. https://doi.org/10.1016/j.patcog.2015.04.004 ISSN 0031-3203, http://www.sciencedirect.com/science/article/pii/S0031320315001284
Acknowledgements
This work is part of the MAMUDS project (Management Multimedia Data for Science) funded by the German ministry of education and research (BMBF Project No. 01D16009) and the Tunisian ministry of higher education and research. We thank the herbarium Haussknecht of Jena for cooperating with them and providing the dataset to train our model. We also thank FSU Jena and Sfax University. The authors would like to thank Prof. Dr. Birgitta Konig-Ries, Prof.Dr.Frank Hellwig and Dr. Jorn Hentschel for supporting this project.
Funding
Supported by the German ministry of education and research (BMBF) and Tunisian ministry of higher education and research (MESR).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interests
The authors declare that there is no conflict of interests regarding the publication of this article.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Triki, A., Bouaziz, B., Mahdi, W. et al. Deep learning based approach for digitized herbarium specimen segmentation. Multimed Tools Appl 81, 28689–28707 (2022). https://doi.org/10.1007/s11042-022-12935-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-12935-8