Cloud-contaminated image reconstruction with contextual spatio-spectral information

Benabdelkader, Souad; Melgani, Farid; Boulemden, Mohammed

doi:10.1109/igarss.2007.4422808

Cited by 7 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our main assumptions are that clouds and shadows can be identified and then replaced using data from other dates (Martinuzzi et al 2007). Indeed, prediction algorithms, based on spatial and spectral correlation (Benabdelkader et al 2007), may be implemented using data available in the cloud-free neighbourhood of contaminated areas (Benabdelkader and Melgani 2008). Visual comparison of multi-band images 'before' and 'after' removal of cloud and cloud shadows (figure 4), along with the accuracy assessment results, indicated that our cloud removal efforts associated with a multi-temporal spectral analysis approach provided a fairly accurate classification map for an area considered difficult to classify due to persistent cloud cover and challenges in spectrally discriminating among land cover types that have similar spectral characteristics during various times of the year.…”

Section: Discussionmentioning

confidence: 99%

Mapping coffee plantations with Landsat imagery: an example from El Salvador

Ortega-Huerta

Komar²,

Price

et al. 2011

International Journal of Remote Sensing

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Mapping coffee plantations with Landsat imagery: an example from El Salvador

Ortega-Huerta

Komar²,

Price

et al. 2011

International Journal of Remote Sensing

View full text Add to dashboard Cite

“…Based on the multitemporal and multispectral characteristics of remote sensing images, a new method, which is different from the previous contextual based methods, 15,16 is developed. For example, a scene of Landsat-5 TM image covers an area of about 185 × 185 km, with about 40 million pixels.…”

Section: Methodsmentioning

confidence: 99%

“…5 These methods are generally rarely consider the relationship information of various spectra. [11][12][13][14][15][16] These methods select the best measurement (the most cloud-free pixel) among a set of measurements acquired over a limited time period to represent the considered multitemporal pixels. 2 They can only process an image without significant heterogeneity.…”

Section: Introductionmentioning

confidence: 99%

Method for reconstructing the pixel missing region on remote sensing images

Fan

2013

J. Appl. Remote Sens

View full text Add to dashboard Cite

Reconstruction of the pixel missing regions on remote sensing images is an important research issue in remote sensing image processing. These regions due to clouds and stripes are generally processed using the traditional synthesis and image inpainting techniques. Such processing can also be implemented with multispectral/multitemporal based methods. However, these methods have some limitations in applications, such as the sharpened borders between the filled regions and their neighbors and the requirement for multiple images. A method is developed for reconstructing pixel missing regions using a corresponding recent term image (reference image) acquired from different sensors with different spectral bands. A case study shows that the new method obviously outperforms the replacement method.

show abstract

“…Shen et al [11] addressed this issue in a broad technical review. Li et al [12], Melgani [13], Benabdelkader et al [14], Benabdelkader et al [15] and Gómez-Chova et al [16] addressed cloud removal with approaches that build cloudfree image composites selecting cloud-free pixels in images collected along a short period. Despite these approaches proving visually high-quality images, they did not meet the high temporal frequency needed for many agricultural applications.…”

Section: Introductionmentioning

confidence: 99%

Pix2pix Conditional Generative Adversarial Network with MLP Loss Function for Cloud Removal in a Cropland Time Series

et al. 2021

View full text Add to dashboard Cite

Clouds are one of the major limitations to crop monitoring using optical satellite images. Despite all efforts to provide decision-makers with high-quality agricultural statistics, there is still a lack of techniques to optimally process satellite image time series in the presence of clouds. In this regard, in this article it was proposed to add a Multi-Layer Perceptron loss function to the pix2pix conditional Generative Adversarial Network (cGAN) objective function. The aim was to enforce the generative model to learn how to deliver synthetic pixels whose values were proxies for the spectral response improving further crop type mapping. Furthermore, it was evaluated the generalization capacity of the generative models in producing pixels with plausible values for images not used in the training. To assess the performance of the proposed approach it was compared real images with synthetic images generated with the proposed approach as well as with the original pix2pix cGAN. The comparative analysis was performed through visual analysis, pixel values analysis, semantic segmentation and similarity metrics. In general, the proposed approach provided slightly better synthetic pixels than the original pix2pix cGAN, removing more noise than the original pix2pix algorithm as well as providing better crop type semantic segmentation; the semantic segmentation of the synthetic image generated with the proposed approach achieved an F1-score of 44.2%, while the real image achieved 44.7%. Regarding the generalization, the models trained utilizing different regions of the same image provided better pixels than models trained using other images in the time series. Besides this, the experiments also showed that the models trained using a pair of images selected every three months along the time series also provided acceptable results on images that do not have cloud-free areas.

show abstract

Cloud-contaminated image reconstruction with contextual spatio-spectral information

Cited by 7 publications

References 9 publications

Mapping coffee plantations with Landsat imagery: an example from El Salvador

Mapping coffee plantations with Landsat imagery: an example from El Salvador

Method for reconstructing the pixel missing region on remote sensing images

Pix2pix Conditional Generative Adversarial Network with MLP Loss Function for Cloud Removal in a Cropland Time Series

Contact Info

Product

Resources

About