{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T14:17:49Z","timestamp":1770819469839,"version":"3.50.1"},"reference-count":44,"publisher":"MDPI AG","issue":"2","license":[{"start":{"date-parts":[[2026,2,1]],"date-time":"2026-02-01T00:00:00Z","timestamp":1769904000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Future Internet"],"abstract":"<jats:p>Large language models (LLMs) have shown remarkable results when tasked with the analysis and production of texts or images and for captioning images. Aerial images differ from other images since they exhibit many natural objects that have a highly variable color range and no clear contours. This paper reports to what extent an LLM, i.e., Llama-4, can be tasked with the identification and captioning in aerial images of natural objects, such as tree categories, uncultivated land, and some man-made objects, such as roads. This valuable automation is needed to scan large areas and detect the parts for which a sudden maintenance or an emergency intervention is due. Tests on the chosen LLM were performed against a custom image dataset built to overcome the limited availability of such a domain-specific aerial image set. To evaluate the identification and captioning results, the accuracy, precision and recall metrics were computed. The results given by a cutting-edge variant of Llama-4, namely Maverick, reveal its strengths and weaknesses in this context. Although it is remarkable that an out-of-the-box tool can give assistance in such a complex observation and detection task, substantial progress is needed for such a model to improve accuracy and constitute a reliable support, as accuracy is at most 58.6% and recall is at most 56.1%.<\/jats:p>","DOI":"10.3390\/fi18020077","type":"journal-article","created":{"date-parts":[[2026,2,3]],"date-time":"2026-02-03T10:03:29Z","timestamp":1770113009000},"page":"77","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Aerial Image Analysis: When LLMs Assist (And When Not)"],"prefix":"10.3390","volume":"18","author":[{"given":"Salvatore","family":"Calcagno","sequence":"first","affiliation":[{"name":"Dipartimento di Matematica e Informatica, University of Catania, 95125 Catania, Italy"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-3692-6091","authenticated-orcid":false,"given":"Erika","family":"Scaletta","sequence":"additional","affiliation":[{"name":"Dipartimento di Matematica e Informatica, University of Catania, 95125 Catania, Italy"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7169-659X","authenticated-orcid":false,"given":"Emiliano","family":"Tramontana","sequence":"additional","affiliation":[{"name":"Dipartimento di Matematica e Informatica, University of Catania, 95125 Catania, Italy"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-1028-2178","authenticated-orcid":false,"given":"Gabriella","family":"Verga","sequence":"additional","affiliation":[{"name":"Dipartimento di Matematica e Informatica, University of Catania, 95125 Catania, Italy"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,1]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"100211","DOI":"10.1016\/j.hcc.2024.100211","article-title":"A survey on large language model (LLM) security and privacy: The Good, The Bad, and The Ugly","volume":"4","author":"Yao","year":"2024","journal-title":"High-Confid. Comput."},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Giannilias, T., Papadakis, A., Nikolaou, N., and Zahariadis, T. (2025). Classification of Hacker\u2019s Posts Based on Zero-Shot, Few-Shot, and Fine-Tuned LLMs in Environments with Constrained Resources. Future Internet, 17.","DOI":"10.3390\/fi17050207"},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Nam, D., Macvean, A., Hellendoorn, V., Vasilescu, B., and Myers, B. (2024, January 14\u201320). Using an llm to help with code understanding. Proceedings of the IEEE\/ACM 46th International Conference on Software Engineering, Lisbon, Portugal.","DOI":"10.1145\/3597503.3639187"},{"key":"ref_4","unstructured":"Xu, S., Wu, Z., Zhao, H., Shu, P., Liu, Z., Liao, W., Li, S., Sikora, A., Liu, T., and Li, X. (2024). Reasoning before comparison: LLM-enhanced semantic similarity metrics for domain specialized text analysis. arXiv."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Wangsa, K., Karim, S., Gide, E., and Elkhodr, M. (2024). A Systematic Review and Comprehensive Analysis of Pioneering AI Chatbot Models from Education to Healthcare: ChatGPT, Bard, Llama, Ernie and Grok. Future Internet, 16.","DOI":"10.3390\/fi16070219"},{"key":"ref_6","unstructured":"Li, J., Li, D., Savarese, S., and Hoi, S. (2023, January 23\u201329). Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. Proceedings of the International Conference on Machine Learning, PMLR, Honolulu, HI, USA."},{"key":"ref_7","unstructured":"Peng, Z., Wang, W., Dong, L., Hao, Y., Huang, S., Ma, S., and Wei, F. (2023). Kosmos-2: Grounding multimodal large language models to the world. arXiv."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Du, S., Tang, S., Wang, W., Li, X., and Guo, R. (2023). Tree-gpt: Modular large language model expert system for forest remote sensing image understanding and interactive analysis. arXiv.","DOI":"10.5194\/isprs-archives-XLVIII-1-W2-2023-1729-2023"},{"key":"ref_9","unstructured":"He, Y., and Sun, Q. (2023). Towards automatic satellite images captions generation using large language models. arXiv."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"7892","DOI":"10.1109\/JSTARS.2025.3547144","article-title":"MiniCPM-V LLaMA Model for Image Recognition: A Case Study on Satellite Datasets","volume":"18","year":"2025","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_11","first-page":"4409318","article-title":"LLaMA-Unidetector: A LLaMA-Based Universal Framework for Open-Vocabulary Object Detection in Remote Sensing Imagery","volume":"63","author":"Xie","year":"2025","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_12","unstructured":"Lam, D., Kuzma, R., McGee, K., Dooley, S., Laielli, M., Klaric, M., Bulatov, Y., and McCord, B. (2018). xView: Objects in context in overhead imagery. arXiv."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"296","DOI":"10.1016\/j.isprsjprs.2019.11.023","article-title":"Object detection in optical remote sensing images: A survey and a new benchmark","volume":"159","author":"Li","year":"2020","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_14","unstructured":"Irvin, J., Sheng, H., Ramachandran, N., Johnson-Yu, S., Zhou, S., Story, K., Rustowicz, R., Elsworth, C., Austin, K., and Ng, A.Y. (2020). Forestnet: Classifying drivers of deforestation in indonesia using deep learning on satellite imagery. arXiv."},{"key":"ref_15","unstructured":"Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.A., Lacroix, T., Rozi\u00e8re, B., Goyal, N., Hambro, E., and Azhar, F. (2023). Llama: Open and efficient foundation language models. arXiv."},{"key":"ref_16","first-page":"34892","article-title":"Visual instruction tuning","volume":"36","author":"Liu","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"7543","DOI":"10.1109\/TPAMI.2025.3571946","article-title":"Otter: A multi-modal model with in-context instruction tuning","volume":"47","author":"Li","year":"2025","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_18","unstructured":"Su, Y., Lan, T., Li, H., Xu, J., Wang, Y., and Cai, D. (2023). Pandagpt: One model to instruction-follow them all. arXiv."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"166","DOI":"10.1016\/j.isprsjprs.2019.04.015","article-title":"Deep learning in remote sensing applications: A meta-analysis and review","volume":"152","author":"Ma","year":"2019","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1109\/MGRS.2024.3393010","article-title":"Deep learning for satellite image time-series analysis: A review","volume":"12","author":"Miller","year":"2024","journal-title":"IEEE Geosci. Remote Sens. Mag."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Sheehan, A., Beddows, A., Green, D.C., and Beevers, S. (2023). City scale traffic monitoring using worldview satellite imagery and deep learning: A case study of Barcelona. Remote Sens., 15.","DOI":"10.3390\/rs15245709"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"93","DOI":"10.1186\/s40537-023-00772-x","article-title":"Review of deep learning methods for remote sensing satellite images classification: Experimental survey and comparative analysis","volume":"10","author":"Adegun","year":"2023","journal-title":"J. Big Data"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"21","DOI":"10.54254\/2977-3903\/2025.20517","article-title":"Multi-temporal analysis of land use change using GIS and satellite imagery: Implications for sustainable urban planning","volume":"15","author":"Li","year":"2025","journal-title":"Adv. Eng. Innov."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Ou, R., Yan, H., Wu, M., and Zhang, C. (2023). A method of efficient synthesizing post-disaster remote sensing image with diffusion model and LLM. Proceedings of the Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC), Taipei, Taiwan, 31 October\u20133 November 2023, IEEE.","DOI":"10.1109\/APSIPAASC58517.2023.10317383"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Osco, L.P., Lemos, E.L.d., Gon\u00e7alves, W.N., Ramos, A.P.M., and Marcato Junior, J. (2023). The potential of visual chatgpt for remote sensing. Remote Sens., 15.","DOI":"10.20944\/preprints202304.0926.v1"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Kang, Y., Zheng, B., and Shen, W. (2025). Research on Oriented Object Detection in Aerial Images Based on Architecture Search with Decoupled Detection Heads. Appl. Sci., 15.","DOI":"10.3390\/app15158370"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"87","DOI":"10.30897\/ijegeo.1010741","article-title":"Comparison of YOLO versions for object detection from aerial images","volume":"9","author":"Atik","year":"2022","journal-title":"Int. J. Environ. Geoinform."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Hu, M., Li, Z., Yu, J., Wan, X., Tan, H., and Lin, Z. (2023). Efficient-lightweight yolo: Improving small object detection in yolo for aerial images. Sensors, 23.","DOI":"10.3390\/s23146423"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Marletta, D., Midolo, A., and Tramontana, E. (2023). Detecting Photovoltaic Panels in Aerial Images by Means of Characterising Colours. Technologies, 11.","DOI":"10.3390\/technologies11060174"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Calcagno, S., Scaletta, E., Tramontana, E., and Verga, G. (2025). YOLO-based Recognition of some Crop Categories from Real-World Aerial Images. Proceedings of the International Conference on Enabling Technologies: Infrastructure for Collaborative Enterprises (WETICE), Catania, Italy, 23\u201325 July 2025, IEEE.","DOI":"10.1109\/WETICE67341.2025.11092212"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Ju, H., Park, I., Nalcakan, Y., Jin, Y., Yeo, S., and Kim, S. (2025). Exploring the Limits of Large Language Models\u2019 Ability to Distinguish Between Objects. Appl. Sci., 15.","DOI":"10.3390\/app15094620"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Pritt, M., and Chern, G. (2017). Satellite image classification with deep learning. Proceedings of the Applied Imagery Pattern Recognition Workshop (AIPR), Washington, DC, USA, 10\u201312 October 2017, IEEE.","DOI":"10.1109\/AIPR.2017.8457969"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Marletta, D., Midolo, A., and Tramontana, E. (2024). Automatic Land Use and Land Cover Classification by Means of Characterising Colours. Proceedings of the International Conference on Enabling Technologies: Infrastructure for Collaborative Enterprises (WETICE), Reggio Emilia, Italy, 26\u201328 June 2024, IEEE.","DOI":"10.1109\/WETICE64632.2024.00034"},{"key":"ref_34","first-page":"812","article-title":"High resolution satellite imagery segmentation based on adaptively integrated multiple features","volume":"Volume 6786","author":"Wang","year":"2007","journal-title":"Proceedings of the Automatic Target Recognition and Image Analysis, and Multispectral Image Acquisition (MIPPR), Wuhan, China, 15\u201317 November 2007"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"5227","DOI":"10.1109\/TPAMI.2024.3362475","article-title":"SpectralGPT: Spectral Remote Sensing Foundation Model","volume":"46","author":"Hong","year":"2024","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"106341","DOI":"10.1016\/j.still.2024.106341","article-title":"Assessing field-scale rill erosion mitigation by cover crops in arable land using drone image analysis","volume":"246","author":"Futerman","year":"2025","journal-title":"Soil Tillage Res."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"128953","DOI":"10.1016\/j.ufug.2025.128953","article-title":"Drone-based assessment of urban green space structure and cooling capacity","volume":"112","author":"Shen","year":"2025","journal-title":"Urban For. Urban Green."},{"key":"ref_38","first-page":"51","article-title":"Wireless Surveillance with Human Detection Using Artificial Intelligence and Drone","volume":"17","author":"Sulaiman","year":"2025","journal-title":"J. Telecommun. Electron. Comput. Eng. (JTEC)"},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Effendi, M.R., Al-Falah, R.S., and Ismail, N. (2021). IoT-Based Battery Monitoring System in Solar Power Plants with Secure Copy Protocol (SCP). Proceedings of the 2021 7th International Conference on Wireless and Telematics (ICWT), Bandung, Indonesia, 19\u201320 August 2021, IEEE.","DOI":"10.1109\/ICWT52862.2021.9678210"},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"111284","DOI":"10.1016\/j.patcog.2024.111284","article-title":"PSSCL: A progressive sample selection framework with contrastive loss designed for noisy labels","volume":"161","author":"Zhang","year":"2025","journal-title":"Pattern Recognit."},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Alexandrova, S., Tatlock, Z., and Cakmak, M. (2015). RoboFlow: A flow-based visual programming language for mobile manipulation tasks. Proceedings of the International Conference on Robotics and Automation (ICRA), Seattle, WA, USA, 26\u201330 May 2015, IEEE.","DOI":"10.1109\/ICRA.2015.7139973"},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"420208","DOI":"10.55003\/ETH.420208","article-title":"Development of an Image Processing System for Defect Detection in Nam Dok Mai Golden Mangoes","volume":"42","author":"Mungkan","year":"2025","journal-title":"Eng. Technol. Horiz."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"42","DOI":"10.1145\/3703155","article-title":"A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions","volume":"43","author":"Huang","year":"2025","journal-title":"ACM Trans. Inf. Syst."},{"key":"ref_44","unstructured":"Bai, Z., Wang, P., Xiao, T., He, T., Han, Z., Zhang, Z., and Shou, M.Z. (2024). Hallucination of multimodal large language models: A survey. arXiv."}],"container-title":["Future Internet"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1999-5903\/18\/2\/77\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T13:37:23Z","timestamp":1770817043000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1999-5903\/18\/2\/77"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,1]]},"references-count":44,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["fi18020077"],"URL":"https:\/\/doi.org\/10.3390\/fi18020077","relation":{},"ISSN":["1999-5903"],"issn-type":[{"value":"1999-5903","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,1]]}}}