SOLVING THE INVERSE PROBLEM OF LOCALIZING IONIZING RADIATION SOURCES USING COMPUTER VISION METHODS: FROM THE HOUGH TRANSFORM TO SUB-PIXEL REFINING
DOI:
https://doi.org/10.32782/geotech2025.39.07Keywords:
UAV, radiation monitoring, Hough transform, Taubin method, source localization, sparse data processing, computer vision.Abstract
The paper considers the current problem of operational search and localization of lost sources of ionizing radiation using unmanned aerial vehicles (UAVs). Traditional methods of radiation reconnaissance face the challenges of processing sparse spatial data and a high level of stochastic noise characteristic of airborne detectors. The aim of the paper is to develop a stable (robust) algorithmic support for automated reconstruction of radiation fields and high-precision identification of anomalies. A hybrid approach is proposed that adapts computer vision methods for radioecological monitoring tasks. The data processing methodology includes: spatial interpolation using the natural neighbor method to restore the continuous topology of the field; adaptive Gaussian filtering to minimize Poisson noise; gradient contour extraction using the Canny detector. The key stage is the application of the Hough transform to detect the centers of radiation anomalies, the parameters of which are additionally refined by the algebraic Taubin method to achieve subpixel accuracy. The validation of the proposed approach was carried out by numerical modeling with simulation of real flight conditions and statistical characteristics of decay. The experimental results demonstrated the high accuracy of the algorithm: the root mean square localization error (RMSE) was 0.99 m. The effectiveness of the method was also confirmed on field data obtained during monitoring of the «Sandy Plateau» area in the Chernobyl Exclusion Zone. It is proven that the developed approach allows to localize sources in quasi-real time, which makes it an effective tool for radiation threat response systems.
References
Canny, J. (1986). A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 8(6), 679–698.
Chen, C. M., Sinclair, L. E., Fortin, R., Coyle, M., & Samson, C. (2020). In-flight performance of the Advanced Radiation Detector for UAV Operations (ARDUO). Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 954, 161609.
Chilès, J. P., & Delfiner, P. (2012). Geostatistics: Modeling spatial uncertainty. John Wiley & Sons.
Czaja, B., Maslanka, K., Skokowski, P., & Kelner, J. (2025). Overview of mutual localization methods between UAVs in a swarm. TransNav: International Journal on Marine Navigation and Safety of Sea Transportation, 19.
Dubois, G. (2008). Advances in automatic interpolation for real-time mapping. Stochastic Environmental Research and Risk Assessment, 22(5), 597–599.
Dubois, P. (2024). Methods for automatization of radiotherapy treatment planning [Doctoral dissertation, Université Paris-Saclay].
Duda, R. O., & Hart, P. E. (1972). Use of the Hough transformation to detect lines and curves in pictures. Communications of the ACM, 15(1), 11–15.
Kasperek, D., & Podpora, M. (2024). Optimizing geospatial data for ML/CV applications: A Python-based approach. Applied Sciences, 14(24).
Khomutinin, Yu. V., Levchuk, S. Ye., Protsak, V. P., & Kashparov, V. O. (2020). Kartohrafuvannia radioaktyvnoho zabrudnennia iz zadanym rivnem doviry [Mapping of radioactive contamination with a given level of confidence]. Nuclear Physics & Atomic Energy, 21(3).
Lin, S., Ning, S., Zhu, H., & others. (2024). Neural network methods for radiation detectors and imaging. Frontiers in Physics, 12, 1334298.
Martin, P. G., Payton, O. D., Fardoulis, J. S., Richards, D. A., Yamashiki, Y., & Scott, T. B. (2016). 3D unmanned aerial vehicle radiation mapping for assessing contaminant distribution and mobility. International Journal of Applied Earth Observation and Geoinformation, 52, 12–19.
Morelande, M. R., & Skvortsov, A. (2009). Radiation field estimation using a Gaussian mixture. In Proceedings of the 12th International Conference on Information Fusion (pp. 2247–2254).
Nosenko, T., & Mashkina, I. (2025). Modeliuvannia protsesiv obrobky ekolohichnykh danykh dlia system mobilnoho monitorynhu na osnovi BPLA ta metodu IDW [Modeling of environmental data processing processes for UAV-based mobile monitoring systems and IDW method]. Cybersecurity: Education, Science, Technique, 2(30), 110–124.
Opanasiuk, B. M., & Mokin, V. B. (2025). Rozrobka informatsiinoi tekhnolohii heostatystychnoho analizu radiatsiinoho zabrudnennia v Yevropi [Development of information technology for geostatistical analysis of radiation pollution in Europe] [Doctoral dissertation, VNTU].
Popov, O., Bondar, O., Ivaschenko, T., & others. (2023). Features of the modern UAV-based complexes use to solve radiation control problems. In Systems, Decision and Control in Energy IV (pp. 35–57). Springer.
Taubin, G. (1991). Estimation of planar curves, surfaces, and nonplanar space curves defined by implicit equations with applications to edge and range image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(11), 1115–1138.
Trysnyuk, V. M., & Nagornyi, Ye. I. (2024). Vykorystannia metodiv interpoliatsii dlia obrobky danykh radiatsiinoi rozvidky [Use of interpolation methods for processing radiation reconnaissance data]. Telekomunikatsiini ta informatsiini tekhnolohii, 1, 73–78.
Trysnyuk, V., Nagornyi, E., Trysnyuk, T., Konetska, O., & Kurilo, A. (2022). Metodyka vyiavlennia radiatsiinoho zabrudnennia mistsevosti ta yoho ryzykiv [Methodology for detecting radiation contamination of the area and its risks]. Systemy upravlinnia, navihatsii ta zviazku, 3(69), 112–115.
Watson, D. F. (1992). Contouring: A guide to the analysis and display of spatial data. Pergamon Press.
West, A., Tsitsimpelis, I., Licata, M., & others. (2021). Use of Gaussian process regression for radiation mapping of a nuclear reactor with a mobile robot. Scientific Reports, 11(1), 13975.
Xu, S., Hou, Z., Wei, C., & others. (2026). Real-time radiation noise suppression and dose detection method based on FPGA with comparative analysis with deep learning methods. Expert Systems with Applications, 295, 128906.
