Very High Resolution Satellite Images Filtering

Kourgli, Assia; Oukil, Youcef

doi:10.1109/bwcca.2013.81

Cited by 1 publication

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This procedure resulted in a 6-spectral-band image (blue, green, red, near-infrared, SWIR-1, and SWIR-2). Satellite images commonly present fragmentation problems, and hence, a smoothing filter was applied to homogenize pixels and eliminate water noises, which tend to interfere in the image classification process, mainly in the water regions (KOURGLI, 2013).…”

Section: Treatment Of Multispectral Imagesmentioning

confidence: 99%

Bathymetric DEM derived from Lyzenga’s algorithm (multispectral Bathymetry) using Landsat 8 images: a case study of Fort Lauderdale coast

Paula,

Lima,

Cintra

et al. 2023

Rev. Bras. Cartogr.

View full text Add to dashboard Cite

Optical bathymetry can be applied to multispectral images of coastal regions with low spectral and spatial resolutions, and the results obtained are satisfactory in waters up to 15 m. However, the limitations and effectiveness may vary as they are related to physical factors such as water turbidity, specular reflection, and the presence of clouds and shadows during imagery acquisition. Some of these factors also affect the surveys conducted using green Light Detection and Ranging (LiDAR) technology. This study aimed to show that topographic representation by multispectral images is directly related to the physical factors of the water body and the atmospheric conditions during the image acquisition process, good atmospheric conditions and shallow water can generate accurate topographic representation. The experiments carried out in this study using optical bathymetry showed that Landsat 8 satellite images could be used in multispectral bathymetry with a linear regression model by using depth samples obtained from shallow water. Images taken at three different time points were used, with a certain time interval between the data collection performed via the Green LiDAR technology. The depths estimated by this model showed some differences that were possibly caused by water turbidity and temporally-occurring activities. Despite such constraints, the depths obtained by the linear regression model showed a good adaptation to the model derived from the Green LiDAR for depths up to 22 m, with an average error of -0.271 m and a root mean square error of 0.936 m.

show abstract

Section: Treatment Of Multispectral Imagesmentioning

confidence: 99%