Gokhan KilicMehmet Suleyman Ünlütürk2025-10-062016009039730090-39731945-755310.1520/JTE20150057https://www.scopus.com/inward/record.uri?eid=2-s2.0-84988878444&doi=10.1520%2FJTE20150057&partnerID=40&md5=0d966f616a2c5366779c05ef24176423https://gcris.yasar.edu.tr/handle/123456789/9775Ground penetrating radar (GPR) uses data collected with the aid of electromagnetic waves transmitted into a structure by antenna to assess and monitor the structural health of many different kinds of civil infrastructure. With GPR technology promoting their system with promises of the achievement of in excess of 1000 sample points per scan this research demonstrated on the basis of the Nyquist theorem that 256 sample points per scan provided equally reliable inspection results. Furthermore 256 sample points per scan GPR data were further analyzed by order statistic filtering with neural networks to locate cracks within concrete materials. The results showed that the neural network order statistic filters are effective in their use of detecting cracks in noisy environments using 256 sample points per scan GPR data. © 2017 Elsevier B.V. All rights reserved.EnglishCrack, Gpr, Neural Network, Nyquist Theorem, Structural Health, Cracks, Electromagnetic Waves, Geological Surveys, Ground Penetrating Radar Systems, Neural Networks, Civil Infrastructures, Concrete Materials, Ground Penetrating Radar (gpr), Noisy Environment, Nyquist Theorem, Order Statistic Filter, Order Statistic Filtering, Structural Health, Crack DetectionCracks, Electromagnetic waves, Geological surveys, Ground penetrating radar systems, Neural networks, Civil infrastructures, Concrete materials, Ground penetrating radar (GPR), Noisy environment, Nyquist theorem, Order statistic filter, Order statistic filtering, Structural health, Crack detectionArticle