No abstract

OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. AbstractThe SPOT 6-7 satellite ground segment includes a systematic and automatic cloud detection step in order to feed a catalogue with a binary cloud mask and an appropriate confidence measure. However, current approaches for cloud detection, that are mostly based on machine learning and hand crafted features, have shown lack of robustness. In other tasks such as image recognition, deep learning methods have shown outstanding results outperforming many state-of-the-art methods. These methods are known to produce a powerful representation that can capture texture, shape and contextual information. This paper studies the potential of deep learning methods for cloud detection in order to achieve state-of-the-art performance. A comparison between deep learning methods used with classical handcrafted features and classical convolutional neural networks is performed for cloud detection. Experiments are conducted on a SPOT 6 image database with various landscapes and cloud coverage and show promising results.

Bathymetric estimation can be obtained from multispectral satellite images for shallow waters. The method is based on the rotation of a pair of spectral bands. One of the resulting images is depth-dependent. Therefore several pixels corresponding to different depths are required to numerically evaluate the linear relation between the pixel values and the real depth for a training area. The aim of this study is to compare, for one bathymetric estimation method and one mesotrophic site, the results of depth estimation with a large panel of satellite and aerial images: CASI, QUICKBIRD, CHRIS PROBA, ETM, HYPERION and MeRIS. For each image the pair of spectral bands chosen to compute the bathymetry has been optimized. Error on depth estimation has been computed on two regions of the image: the training area and a validation area. This comparison is discussed to identify the influence of image parameters (spectral bands, S/N ratio, spatial resolution, and quantization) on the bathymetric results and to propose the most adapted image parameters for bathymetric estimation. For validation purposes, we compared the results obtained with a CASI image matching the optimized parameters in an oligotrophic site in the Red Sea.

No abstract

No abstract