/   Home   /   Newsroom   /   Research News

Multiscale Adaptive Sparse Representation to Reconstruct Missing Remote Sensing Information Proposed

May 18, 2018     Email"> PrintText Size

Reconstructing missing Remote Sensing (RS) information caused by poor atmospheric conditions and defective sensors may improve the availability of RS data. A research team led by Prof. YANG Xiaomei at the Institute of Geographic Sciences and Natural Resources Research (IGSNRR) of the Chinese Academy of Sciences found that sparse dictionary learning was beneficial for clouds removal from high-spatial resolution RS images.  

Studies published in Sensors and Remote Sensing Letters showed that multiscale sparse representation could recover missing surface information for RS data, yielding consistent textures and structures with surrounding information. 

Many studies have been dedicated to coping with the inverse problem of missing RS information reconstruction. However, some approaches cannot work when suffering from significant spectral difference, severe overlapping of missing information, or rapid land-cover change. Moreover, most existing methods based on spatial complementation will introduce artifacts and discard details. 

Image features usually tend to repeat themselves both within the same scale and across different scales. Meanwhile, different features prefer different scales for optimal representation. Inspired by the idea of exemplar-based inpainting and latest progress on sparse representation, researchers recovered missing RS data via multiscale adaptive sparse representation. 

Their main contributions lie in scale adaptive patch inpainting and introduction of adaptive neighborhood consistency into the optimization model of patch inpainting. They trained multiscale feature dictionaries using different sizes of exemplars from the known image region, and utilized them to infer missing information patch-by-patch via sparse representation afterwards.  

"Structure sparsity was incorporated to encourage filling-in of missing patch on image structure. Neighboring information was also employed to restrain the appearance of estimated patch, to yield semantically consistent inpainting result", said Prof. YANG. 

They performed experiments on cloud removal, gaps filling and quantitative products reconstruction, and the results demonstrated the effectiveness of proposed method. 

These works were supported by the National Key Research and Development Program of China and the National Natural Science Foundation of China. 


Illustration of multiscale sparse representation: (a) multiscale adaptive patch inpainting; (b) possible representation of a patch under multiscale dictionary. (Image by YANG Xiaomei) 


(Editor: LI Yuan)

Related Articles

remote sensing;Mekong River;sustainable development;AOGEOSS

Remote Sensing Observation to Help Sustainable Development along Mekong

May 23, 2018

China will coordinate with countries from Asia and Oceania to conduct satellite remote sensing observation along the Mekong River, according to the website of Chinese Academy of Sciences (CAS). The first international conference of the Asia-Oceania Glo...

remote sensing;earthquake

Wenchuan Shows Effects of Ecological Restoration

May 07, 2018

Remote sensing images showed that the environment returning to normal in the Wenchuan earthquake zone ten years after the event, according to a report by the Chinese Academy of Sciences (CAS). The survey showed that landslide areas have been covered by...

remote sensing;Roman ruin;Digital Belt and Road

Scientists Find Roman Ruins Using Orbiting Eye

May 03, 2018

Chinese scientists recently discovered 10 new archaeological sites in Tunisia dating to ancient Rome using Chinese space-based remote sensing technology. The archaeological sites are located in three Tunisian provinces - Gafsa, Tataouine and Medenine -...

Contact Us

Copyright © 2002 - Chinese Academy of Sciences