From [2016-07-10 Sun] to [2016-07-15 Fri], I attend the 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2016) in Beijing, China. As usual, meeting colleagues and friends is a good way to close an intense academic year. The quality of the conference was good, and I found a lot of interesting papers. I think we should deeply consider convolutional neural nets, it seems to be trendy …
Maïlys Lopes, a PhD student working with me and Stéphane Girard, has presented her work on Grassland monitoring using time series: http://igarss2016.org/Papers/viewpapers.asp?papernum=3617.
For your information, please find the paper I have found interesting for my research purpose (which is a strong filter!):
[2016-07-11 Mon]
“CLODD BASED BAND GROUP SELECTION”
- Do bands grouping
- Ref to a band selection method using proximity measure (ref 5)
- Does not talk about the extracted feature :(
14:50: “OPTIMIZING REMOTE SENSING OF NATIVE GRASSLANDS FOR MONITORING AND CHANGE DETECTION”, Room 3011+B
- Introduction sur pourquoi il est important de gérer les “grasslands”
- Actuellement à la main, c’est pas possible de le faire sur long terme
- Utilisation de l’optique et radar en même temps
Questions:
- best combination of optical and radar ?
- quels sont les données qui couvrent correctment le territoire ?
- N’utilise pas des SITS, mais un couple Radarsat-2/Landsat-5.
- Différencier les prairies des cultures ?
- Classif pixel par SVM, avec differents combination of inputs
15:40: “Synergy of Sentinel-1 and Sentinel-2 imagery for wetland monitoring …”
- In order to use the images stream …
Poster “A COMPREHENSIVE EVALUATION OF FEATURE SELECTION ALGORITHMS IN HYPERSPECTRAL IMAGE CLASSIFICATION”
- Simple FS methods are tested
- Forward strategy is uses with SVM -> very long of course.
- test: stabilty, accuracy and processing time.
- Methods that show high accuracy are not stable and vise versa.
[2016-07-12 Tue]
09:00: “FAST HYPERSPECTRAL IMAGE DENOISING BASED ON LOW RANK AND SPARSE REPRESENTATIONS”, Room3021+B
- Use a subspace model, which is similar to HDDA
- Provide an algorithm to remove noise from HSI
09:20: “AN APPROACH TO CONIFER SPECIES CLASSIFICATION BASED ON CROWN STRUCTURE MODELING IN HIGH DENSITY AIRBORNE LIDAR DATA” , Room 311B
- Provide a model to estimate branches of trees based on PCA
- Derive crown indices from the PCA modeling
- Do classification Sparse SVM (?)
09:40: “LIDAR INFORMATION EXTRACTION BY ATTRIBUTE FILTERS WITH PARTIAL RECONSTRUCTION”, Room 311B
- Partial reconstruction is used to filter LiDar data
- EMP is constructed and used as input to classifier
16:00: “TREE HEIGHT ESTIMATES IN BOREAL FOREST USING GAUSSIAN PROCESS REGRESSION”, Room 301A+B
[2016-07-13 Wed]
IDEAL REGULARIZED KERNEL FOR HYPERSPECTRAL IMAGE CLASSIFICATION
- Work quite well: enlarge the area where the hyperparameter of the kernels works well
- Not clear how to fit the regulatization parameter.
A GAUSSIAN APPROACH TO SUBSPACE BASED CLASSIFICATION OF HYPERSPECTRAL IMAGES
- Each class is modeled by a mixture of Gaussian distribution
- The Gaussian distribution is modeled by random projection
GETTING PIXELS AND REGIONS TO AGREE WITH CONDITIONAL RANDOM FIELDS
- Propose a model to share contextual information between pixel-classifier and region-classifier
Spatial contrains:
- inside a region: penalize if a pixel inside a region is of a different label than the region
- between pixel: potts model
- between regions: potts model between region
FOREST BIODIVERSITY MAPPING USING AIRBORNE LIDAR AND HYPERSPECTRAL DATA
[2016-07-14 Thu]
DETECTION OF COMPOUND STRUCTURES BY REGION GROUP SELECTION FROM HIERARCHICAL SEGMENTATIONS
- Propose a method to detect “compound structure” is urban area
- A markov random field on spatial primitive is proposed
- Gibbs sampler is used to optimize the field
TRANSFORMING THE AUTOCORRELATION FUNCTION OF A TIME SERIES TO DETECT LAND COVER CHANGE
- Detect changes by looking stationarity in the autocorrelation function
- window filter is used when the autocorrelation is computed to remove edges effect. Better results are observed.
[2016-07-15 Fri]
A DATA-DRIVEN IDENTIFICATION OF GROWTH-MODEL CLASSES FOR THE ADAPTIVE ESTIMATION OF SINGLE-TREE STEM DIAMETER IN LIDAR DATA
- high intensity lidar for individual tree caracterization
extraction of several parameters modeling
- the crown structure
- the forest density
- the topography
SPARSE DISTRIBUTED HYPERSPECTRAL UNMIXING
- unmixing of hyperspectral image with ADMM
- can be used on large data set
- but the solution is not identical when the method is distributed
Comments
comments powered by Disqus