Drone noise removal with size, weighted size, and median filters (implementation and evaluation in MATLAB environment)
Drone noise removal with size, weighted size, and median filters (implementation and evaluation in MATLAB environment)
Amin Fakhar1 Behnam Paydar Jahromi2 Hossein Abdollahi3
1) Master of Surveying, Photogrammetry, Arian Amirkola Non-Profit University,
2) Master of Science in Surveying Engineering, Geodesy Major, Islamic Azad University, Shahrood Science and Research Branch,
3) Master of Science in Information Technology Management, Shahid Beheshti University, Tehran,
IOI Code
:
AAHD-ABKHCB
Publication :
National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
Drone image processing, as one of the important fields of remote sensing and machine vision, always faces the challenge of noise in the acquisition, transmission, and storage stages. Noise reduces the visual quality of the image and disrupts the extraction of accurate information. In this study, the performance of three classic spatial filters, including the mean filter, the weighted mean filter, and the median filter, in removing common noise from drone images has been evaluated. First, the drone image was imported into the MATLAB software environment and various types of salt and pepper, Gaussian, and speckle noise were applied to it with specific intensities. Then, the aforementioned filters were used to reduce noise, and the results were compared in terms of visual quality, detail retention, and edges. The findings showed that the median filter has the best performance in removing salt and pepper noise and at the same time preserves edges better. The weighted mean filter preserves image details with less blurring than the conventional mean filter, while the simple mean filter, despite its high speed, reduces edge sharpness. The results of this study can be used as a practical guide for selecting the appropriate filter in preprocessing drone images.
Keywords : denoising drone images weighted average filter median filter image processing in matlab salt and pepper noise gaussian noise