Performance Comparison of Noise Reduction Methods on Gokturk-2 Satellite Images

Optical satellite images may suffer from quality degradation due to various types of noise. This study aims to eliminate thermal noise, identify appropriate noise reduction methods, and compare them using specific metrics. The tests were conducted on high-spatial-resolution Gokturk-2 satellite images. Among the methods applied, EPLL, DnCNN, and FFDNet were found to be effective in preserving structural accuracy and fine details. WNNM and NCSR offer a balanced performance by effectively reducing noise while maintaining structural integrity. Wavelet and Bilateral filters are useful for enhancing edge contrast, but should be applied with caution due to the potential for distortion. The results indicate that each algorithm has distinct strengths and weaknesses, emphasizing the importance of selecting the appropriate method based on task requirements.