A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
SNR Characterization in RapidEye Satellite Images
This report outlines an entirely new method of automatically detecting homo-geneous regions in images for the purpose of noise characterization. The methodwas developed with the support of BlackBridge AG in Berlin, Germany. The aim ofthe method was to characterize the signal-to-noise ratio as a function of radianceof the multi-spectral image sensors aboard the RapidEye satellite constellation.The method uses the random nature of noise in order to detect homogeneous im-age regions. The method works by dividing an image into multiple square tiles.Each tile is then corrupted with additive Poisson noise (Gaussian noise with zeromean and a standard deviation equal to the tile mean). The Pearson CorrelationCoefficient between the corrupted tile and the original tile is then used as a ho-mogeneity criterion. It was found that a Pearson Correlation Coefficient of lessthan 0.7 identifies homogeneous regions. When applied to RapidEye images, themethod correctly identified homogeneous regions and allowed the characteriza-tion of the signal-to-noise ratio of the RapidEye image sensors across their dy-namic range. Three case studies of Level 3A RapidEye image products are pre-sented herein. These clearly demonstrate the high quality of RapidEye images aswell as the effectiveness of the described method. ; Validerat; 20141003 (global_studentproject_submitter)
SNR Characterization in RapidEye Satellite Images
This report outlines an entirely new method of automatically detecting homo-geneous regions in images for the purpose of noise characterization. The methodwas developed with the support of BlackBridge AG in Berlin, Germany. The aim ofthe method was to characterize the signal-to-noise ratio as a function of radianceof the multi-spectral image sensors aboard the RapidEye satellite constellation.The method uses the random nature of noise in order to detect homogeneous im-age regions. The method works by dividing an image into multiple square tiles.Each tile is then corrupted with additive Poisson noise (Gaussian noise with zeromean and a standard deviation equal to the tile mean). The Pearson CorrelationCoefficient between the corrupted tile and the original tile is then used as a ho-mogeneity criterion. It was found that a Pearson Correlation Coefficient of lessthan 0.7 identifies homogeneous regions. When applied to RapidEye images, themethod correctly identified homogeneous regions and allowed the characteriza-tion of the signal-to-noise ratio of the RapidEye image sensors across their dy-namic range. Three case studies of Level 3A RapidEye image products are pre-sented herein. These clearly demonstrate the high quality of RapidEye images aswell as the effectiveness of the described method. ; Validerat; 20141003 (global_studentproject_submitter)
SNR Characterization in RapidEye Satellite Images
Moufid, Taha (author)
2014-01-01
Local a5ef9067-143d-4ddd-bf92-a698cb7fcede
Theses
Electronic Resource
English
Images , Satellite , Teknik , RapidEye , Technology , SNR , Estimation , Noise , Signal-to-noise
DDC:
710
Urban vegetation classification: Benefits of multitemporal RapidEye satellite data
| Online Contents | 2013
Arctic Vegetation Type Mapping with RapidEye Data.
| TIBKAT | 2011