TY  - JOUR
TI  - Comparative Analysis on Deep Learning based Pan-Sharpening of Very High-Resolution Satellite Images
AB  - Pan-sharpening is a fundamental task of remote sensing, aiming to produce a synthetic image having high spatial and spectralresolution of original panchromatic and multispectral images. In recent years, as in other tasks of the remote sensing field, deeplearning based approaches have been developed for this task. In this research, a detailed comparative analysis was conducted toevaluate the performance and visual quality of pan-sharpening results from traditional algorithms and deep learning-based models.For this purpose, the deep learning based methods that are CNN based pan-sharpening (PNN), Multiscale and multi-depthconvolutional neural networks (MSDCNN) and Pan-sharpened Generative Adversarial Networks (PSGAN) and traditional methodsthat are Brovey, PCA, HIS, Indusion and PRACS were applied. Analysis was performed on regions with different land covercharacteristics to evaluate the stability of the methods. In addition, effects of the filter size, spectral indices, activation and lossfunctions on the pan-sharpening were investigated. For the accuracy assessment, commonly used with-reference and withoutreference quality metrics were computed in addition to visual quality evaluations. According to results, the deep learning-basedmethods provided promising results in both the reduced resolution and full resolution experiments, while PRACS methodoutperformed other traditional algorithms in most of the experimental configurations.
AU  - Sertel, Elif
AU  - Alganci, Ugur
AU  - WANG, peijuan
DO  - 10.30897/ijegeo.834760
PY  - 2021
JO  - International Journal of Environment and Geoinformatics
VL  - 8
IS  - 2
SN  - 2148-9173
SP  - 150
EP  - 165
DB  - TRDizin
UR  - http://search/yayin/detay/447469
ER  -