How precise is precise enough? Tree crown segmentation using high resolution close-up multispectral UAV images and its effect on NDVI accuracy in Fraxinus excelsior L. trees

Abstract

Detailed individual tree crown segmentation is highly relevant for the detection and monitoring of Fraxinus excelsior L. trees affected by ash dieback, a major threat to common ash populations across Europe. In this study, both fine and coarse crown segmentation methods were applied to close-range multispectral UAV imagery. The fine tree crown segmentation method utilized a novel unsupervised machine learning approach based on a blended NIR–NDVI image, whereas the coarse segmentation relied on the segment anything model (SAM). Both methods successfully delineated tree crown outlines, however, only the fine segmentation accurately captured internal canopy gaps. Despite these structural differences, mean NDVI values calculated per tree crown revealed no significant differences between the two approaches, indicating that coarse segmentation is sufficient for mean vegetation index assessments. Nevertheless, the fine segmentation revealed increased heterogeneity in NDVI values in more severely damaged trees, underscoring its value for detailed structural and health analyses. Furthermore, the fine segmentation workflow proved transferable to both individual UAV images and orthophotos from broader UAV surveys. For applications focused on structural integrity and spatial variation in canopy health, the fine segmentation approach is recommended.

Keywords

Leaf mass segmentation / Machine learning / Segment anything model / Ash dieback

References

[1] Agostinelli M, Nguyen D, Witzell J, Cleary M. Mycobiome of Fraxinus excelsior with different phenotypic susceptibility to ash dieback. Front Glob Change, 2021, 4. 580514

[2] Argamosa RJL, Paringit EC, Quinton KR, Tandoc FAM, Faelga RAG, Ibañez CAG, Posilero MAV, Zaragosa GP. Fully automated gis-based individual tree crown delineation based on curvature values from a lidar derived canopy height model in a coniferous plantation. Int Arch Photogramm Remote Sens Spatial Inf Sci, 2016, XLI-B8: 563-569.

[3] Balasundaram A, Sharma A, Kumaravelan S, Shaik A, Kavitha MS. An improved normalized difference vegetation index (NDVI) estimation using grounded Dino and segment anything model for plant health classification. IEEE Access, 2024, 12: 75907-75919.

[4] Baral HO, Queloz V, Hosoya T. Hymenoscyphus fraxineus, the correct scientific name for the fungus causing ash dieback in Europe. IMA Fungus, 2014, 5(1): 79-80.

[5] Barbedo JGA. A review on the use of unmanned aerial vehicles and imaging sensors for monitoring and assessing plant stresses. Drones, 2019, 3(2): 40.

[6] Buchner L, Eisen AK, Šikoparija B, Jochner-Oette S. Pollen viability of Fraxinus excelsior in storage experiments and investigations on the potential effect of long-range transport. Forests, 2022, 13(4): 600.

[7] Buchner L, Eisen AK, Jochner-Oette S. Effects of ash dieback on leaf physiology and leaf morphology of Fraxinus excelsior L. Trees, 2024, 38(5): 1205-1221.

[8] Buchner L, Eisen AK, Jochner-Oette S. Identification of damage severity in Fraxinus excelsior L. trees caused by ash dieback using multisensory and multitemporal UAV imagery. For Ecol Manage, 2025, 585. 122660

[9] Chan AHY, Barnes C, Swinfield T, Coomes DA. Monitoring ash dieback (Hymenoscyphus fraxineus) in British forests using hyperspectral remote sensing. Remote Sens Ecol Conserv, 2021, 7(2): 306-320.

[10] Chen XX, Jiang K, Zhu YS, Wang XJ, Yun T. Individual tree crown segmentation directly from UAV-borne LiDAR data using the PointNet of deep learning. Forests, 2021, 12(2): 131.

[11] Cinat P, Di Gennaro SF, Berton A, Matese A. Comparison of unsupervised algorithms for vineyard canopy segmentation from UAV multispectral images. Remote Sens, 2019, 11(9. 1023

[12] Coker TLR, Rozsypálek J, Edwards A, Harwood TP, Butfoy L, Buggs RJA. Estimating mortality rates of European ash (Fraxinus excelsior) under the ash dieback (Hymenoscyphus fraxineus) epidemic. Plants People Planet, 2019, 1(1): 48-58.

[13] Dalponte M, Reyes F, Kandare K, Gianelle D. Delineation of individual tree crowns from ALS and hyperspectral data: a comparison among four methods. Eur J Remote Sens, 2015, 48(1): 365-382.

[14] Dawson M, Dawson H, Gurnell A, Lewin J, Macklin MG. Ai-assisted interpretation of changes in riparian woodland from archival aerial imagery using Meta’s segment anything model. Earth Surf Process Landf, 2025, 50(1. e6053

[15] DJI (2022) DJI Mavic 3M user manual

[16] Douss R, Farah IR. Extraction of individual trees based on canopy height model to monitor the state of the forest. Trees for People, 2022, 8. 100257

[17] Eisen AK, Fussi B, Šikoparija B, Jochner-Oette S. Aerobiological pollen deposition and transport of Fraxinus excelsior L. at a small spatial scale. Forests, 2022, 13(3. 424

[18] Eisen AK, Semizer-Cuming D, Jochner-Oette S, Fussi B. Pollination success of Fraxinus excelsior L. in the context of ash dieback. Ann for Sci, 2023, 80(1. 22

[19] Eisen AK, Buchner L, Fussi B, Jochner-Oette S. Does ash dieback affect the reproductive ecology of Fraxinus excelsior L.?. J for Res, 2024, 35(1. 16

[20] Enderle R, Nakou A, Thomas K, Metzler B. Susceptibility of autochthonous German Fraxinus excelsior clones to Hymenoscyphus pseudoalbidus is genetically determined. Ann for Sci, 2015, 72(2): 183-193.

[21] Erdem F, Bayrak OC. Evaluating the effects of texture features on Pinus sylvestris classification using high-resolution aerial imagery. Ecol Inform, 2023, 78. 102389

[22] Flynn WRM, Grieve SWD, Henshaw AJ, Owen HJF, Buggs RJA, Metheringham CL, Plumb WJ, Stocks JJ, Lines ER. UAV-derived greenness and within-crown spatial patterning can detect ash dieback in individual trees. Ecol Solut Evid, 2024, 5(2. e12343

[23] Freudenberg M, Magdon P, Nölke N. Individual tree crown delineation in high-resolution remote sensing images based on U-net. Neural Comput Appl, 2022, 34(24): 22197-22207.

[24] Fuchs S, Häuser H, Peters S, Knauf L, Rentschler F, Kahlenberg G, Kätzel R, Evers J, Paar U, Langer GJ. Ash dieback assessments on intensive monitoring plots in Germany: influence of stand, site and time on disease progression. J Plant Dis Prot, 2024, 131(5): 1355-1372.

[25] Gross A, Holdenrieder O, Pautasso M, Queloz V, Sieber TN. Hymenoscyphus pseudoalbidus, the causal agent of European ash dieback. Mol Plant Pathol, 2014, 15(1): 5-21.

[26] Heenkenda MK, Joyce KE, Maier SW. Mangrove tree crown delineation from high-resolution imagery. Photogramm Eng Remote Sens, 2015, 81(6): 471-479.

[27] Huang HY, Li X, Chen CC. Individual tree crown detection and delineation from very-high-resolution UAV images based on bias field and marker-controlled watershed segmentation algorithms. IEEE J Sel Top Appl Earth Obs Remote Sens, 2018, 11(7): 2253-2262.

[28] Huete AR. Vegetation indices, remote sensing and forest monitoring. Geogr Compass, 2012, 6(9): 513-532.

[29] Hultberg T, Sandström J, Felton A, Öhman K, Rönnberg J, Witzell J, Cleary M. Ash dieback risks an extinction cascade. Biol Conserv, 2020, 244. 108516

[30] Kampen M, Lederbauer S, Mund J-P, Immitzer M (2019) UAV-based multispectral data for tree species classification and tree vitality analysis. Dreiländertagung der DGPF, der OVG und der SGPF in Wien, Österreich – Publikationen der DGPF (28), pp 623–639

[31] Kang J, Wang L, Chen F, Niu Z. Identifying tree crown areas in undulating Eucalyptus plantations using JSEG multi-scale segmentation and unmanned aerial vehicle near-infrared imagery. Int J Remote Sens, 2017, 38(8–10): 2296-2312.

[32] Ke YH, Quackenbush LJ. A review of methods for automatic individual tree-crown detection and delineation from passive remote sensing. Int J Remote Sens, 2011, 32(17): 4725-4747.

[33] Kestur R, Angural A, Bashir B, Omkar SN, Anand G, Meenavathi MB. Tree crown detection, delineation and counting in UAV remote sensed images: a neural network based spectral–spatial method. J Indian Soc Remote Sens, 2018, 46(6): 991-1004.

[34] Khan K, Khan RU, Albattah W, Qamar AM. End-to-end semantic leaf segmentation framework for plants disease classification. Complexity, 2022, 2022: 1168700.

[35] Kirillov A, Mintun E, Ravi N, Mao HZ, Rolland C, Gustafson L, Xiao TT, Whitehead S, Berg AC, Lo WY, Dollár P, Girshick R (2024) Segment anything. In: 2023 IEEE/CVF international conference on computer vision (ICCV). IEEE, Paris, pp 3992–4003

[36] Klesse S, Abegg M, Hopf SE, Gossner MM, Rigling A, Queloz V. Spread and severity of ash dieback in Switzerland–tree characteristics and landscape features explain varying mortality probability. Front for Glob Change, 2021, 4. 645920

[37] Langer G. Collar rots in forests of Northwest Germany affected by ash dieback. Balt for, 2017, 23(1): 4-19

[38] Lassalle G, Ferreira MP, La Rosa LEC, de Souza Filho CR. Deep learning-based individual tree crown delineation in mangrove forests using very-high-resolution satellite imagery. ISPRS J Photogramm Remote Sens, 2022, 189: 220-235.

[39] Liu XY, Wang YX, Kang F, Yue Y, Zheng YJ. Canopy parameter estimation of Citrus grandis var. longanyou based on LiDAR 3D point clouds. Remote Sens, 2021, 13(9. 1859

[40] Lu ZA, Qi LJ, Zhang H, Wan JJ, Zhou JR. Image segmentation of UAV fruit tree canopy in a natural illumination environment. Agriculture (Basel), 2022, 12(7): 1039.

[41] Marques P, Pádua L, Adão T, Hruška J, Peres E, Sousa A, Sousa JJ. UAV-based automatic detection and monitoring of chestnut trees. Remote Sens, 2019, 11(7. 855

[42] Mishra P, Sadeh R, Bino E, Polder G, Boer MP, Rutledge DN, Herrmann I. Complementary chemometrics and deep learning for semantic segmentation of tall and wide visible and near-infrared spectral images of plants. Comput Electron Agric, 2021, 186. 106226

[43] Mitchell RJ, Broome A, Beaton JK, Bellamy PE, Ellis CJ, Hester AJ, Hodgetts NG, Iason GR, Littlewood NA, Newey S, Pozsgai G, Ramsay S, Riach D, Stockan JA, Taylor AF, Woodward S. Challenges in assessing the ecological impacts of tree diseases and mitigation measures: the case of Hymenoscyphus fraxineus and Fraxinus excelsior. Baltic for, 2017, 23(1): 116-140

[44] Mohan M, Silva CA, Klauberg C, Jat P, Catts G, Cardil A, Hudak AT, Dia M. Individual tree detection from unmanned aerial vehicle (UAV) derived canopy height model in an open canopy mixed conifer forest. Forests, 2017, 8(9): 340.

[45] Moussaid A, El Fkihi S, Zennayi Y. Tree crowns segmentation and classification in overlapping orchards based on satellite images and unsupervised learning algorithms. J Imaging, 2021, 7(11. 241

[46] Mu Y, Fujii Y, Takata D, Zheng BY, Noshita K, Honda K, Ninomiya S, Guo W. Characterization of peach tree crown by using high-resolution images from an unmanned aerial vehicle. Hortic Res, 2018, 5: 74.

[47] Murfitt J, He YH, Yang J, Mui A, De Mille K. Ash decline assessment in emerald ash borer infested natural forests using high spatial resolution images. Remote Sens, 2016, 8(3. 256

[48] Neupane K, Baysal-Gurel F. Automatic identification and monitoring of plant diseases using unmanned aerial vehicles: a review. Remote Sens, 2021, 13(19. 3841

[49] Polk SL, Chan AHY, Cui KN, Plemmons RJ, Coomes DA, Murphy JM (2022) Unsupervised detection of ASH dieback disease (Hymenoscyphus fraxineus) using diffusion-based hyperspectral image clustering. In: IGARSS 2022—2022 IEEE international geoscience and remote sensing symposium. Kuala Lumpur, Malaysia. IEEE, pp 2287–2290

[50] Ponce JM, Aquino A, Tejada D, Al-Hadithi BM, Andújar JM. A methodology for the automated delineation of crop tree crowns from UAV-based aerial imagery by means of morphological image analysis. Agronomy, 2022, 12(1): 43.

[51] Pontius J, Martin M, Plourde L, Hallett R. Ash decline assessment in emerald ash borer-infested regions: a test of tree-level, hyperspectral technologies. Remote Sens Environ, 2008, 112(5): 2665-2676.

[52] Qiu L, Jing LH, Hu BX, Li H, Tang YW. A new individual tree crown delineation method for high resolution multispectral imagery. Remote Sens, 2020, 12(3): 585.

[53] Safonova A, Guirado E, Maglinets Y, Alcaraz-Segura D, Tabik S. Olive tree biovolume from UAV multi-resolution image segmentation with mask R-CNN. Sensors, 2021, 21(5): 1617.

[54] Sahin HM, Miftahushudur T, Grieve B, Yin HJ. Segmentation of weeds and crops using multispectral imaging and CRF-enhanced U-net. Comput Electron Agric, 2023, 211. 107956

[55] Sarabia R, Aquino A, Ponce JM, López G, Andújar JM. Automated identification of crop tree crowns from UAV multispectral imagery by means of morphological image analysis. Remote Sens, 2020, 12(5): 748.

[56] Speckenwirth S, Brandmeier M, Paczkowski S. Treeseg—a toolbox for fully automated tree crown segmentation based on high-resolution multispectral UAV data. Remote Sens, 2024, 16(19. 3660

[57] Peters S, Langer G, Kätzel R (eds) (2021) Eschentriebsterben. Kriterien zur Schadensbonitur an Eschen, 1. Auflage. Fachagentur Nachwachsende Rohstoffe (FNR), Gülzow-Prüzen

[58] Sun CX, Huang CW, Zhang HQ, Chen BQ, An F, Wang LW, Yun T. Individual tree crown segmentation and crown width extraction from a heightmap derived from aerial laser scanning data using a deep learning framework. Front Plant Sci, 2022, 13. 914974

[59] Tahar KN, Asmadin MA, Sulaiman SAH, Khalid N, Idris AN, Razali MH. Individual tree crown detection using UAV orthomosaic. Eng Technol Appl Sci Res, 2021, 11(2): 7047-7053.

[60] Timmermann V, Børja I, Hietala AM, Kirisits T, Solheim H. Ash dieback: pathogen spread and diurnal patterns of ascospore dispersal, with special emphasis on Norway. EPPO Bull, 2011, 41(1): 14-20.

[61] Timmermann V, Nagy N, Hietala A, Børja I, Solheim H. Progression of ash dieback in Norway related to tree age, disease history and regional aspects. Balt for, 2017, 23: 150-158

[62] Torresan C, Berton A, Carotenuto F, Di Gennaro SF, Gioli B, Matese A, Miglietta F, Vagnoli C, Zaldei A, Wallace L. Forestry applications of UAVs in Europe: a review. Int J Remote Sens, 2017, 38(8–10): 2427-2447.

[63] Ulku I, Akagündüz E, Ghamisi P. Deep semantic segmentation of trees using multispectral images. IEEE J Sel Top Appl Earth Obs Remote Sens, 2022, 15: 7589-7604.

[64] Wacker K, Kim C, van Iersel MW, Sidore B, Pham T, Haidekker M, Seymour L, Ferrarezi RS. Development of an automated low-cost multispectral imaging system to quantify canopy size and pigmentation. Sensors, 2024, 24(17): 5515.

[65] Waser L, Küchler M, Jütte K, Stampfer T (2014) Evaluating the potential of worldview-2 data to classify tree species and different levels of ash mortality. Remote Sens 6(5):4515–4545. https://doi.org/10.3390/rs6054515

[66] Wu L, Zhang Y, Gao Y, Zhang YI. Tree crown detection and delineation in high resolution RS image. IEEE Int Geosci Remote Sens Sympos, 2004, 2004: 3841-3844.

[67] Zhang JC, Huang YB, Pu RL, Gonzalez-Moreno P, Yuan L, Wu KH, Huang WJ. Monitoring plant diseases and pests through remote sensing technology: a review. Comput Electron Agric, 2019, 165. 104943

[68] Zhang N, Wang YT, Zhang XL. Extraction of tree crowns damaged by Dendrolimus tabulaeformis Tsai et Liu via spectral-spatial classification using UAV-based hyperspectral images. Plant Methods, 2020, 16: 135.

[69] Zheng JP, Yuan S, Li WJ, Fu HH, Yu L, Huang JX (2025) A review of individual tree crown detection and delineation from optical remote sensing images: current progress and future

Funding

Katholische Universität Eichstätt-Ingolstadt (3115)

RIGHTS & PERMISSIONS

The Author(s)