An abrupt change detection algorithm for buried landmines localization

Potin, D.; Vanheeghe, Philippe; Duflos, Emmanuel; Davy, Manuel

doi:10.1109/tgrs.2005.861413

Cited by 46 publications

(30 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, we adapt the SF-TR matrices proposed in [15] for its potential application to GPR detection of shallowly buried dielectric targets. As discussed in [4,21] and elsewhere, target signature in such cases is hidden in a strong clutter signal, thus poses an important challenge. In our attempt to extract the target signature in the presence of strong clutter, we develop a method based on spatially sliding windows and synthesizing singular vector distributions corresponding to different scattering mechanisms.…”

Section: Introductionmentioning

confidence: 99%

GPR Signal Enhancement Using Sliding-Window Space-Frequency Matrices

Yavuz¹,

Fouda²,

Teixeira³

2014

PIER

View full text Add to dashboard Cite

Abstract-Ground penetrating radar (GPR) has shown to provide useful results for detection of buried objects. However, its performance suffers from strong reflection from ground surface especially for shallowly buried targets. In such cases, the detection problem depends on the separation of the target signal from the ground backscatter such as landmines and unexploded ordnances. In this paper, we discuss and analyze the use of space-frequency time-reversal matrices for the enhancement of ground penetrating radar signals and potential clutter reduction. Through the use of sliding windows, submatrices from a given B-scan (radargram) are utilized to extract localized scattering information of a given detection scenario. Each sub-B-scan is decomposed to its singular vectors and later used to render synthetic aperture time-domain singular vector distributions corresponding to different scattering mechanisms. Later, they are weighted by the singular values and subtracted from the full B-scan to achieve reduced clutter and enhanced target response. The method shows satisfactory results for shallowly buried dielectric targets even in the presence of rough surface profiles.

show abstract

Section: Introductionmentioning

confidence: 99%

GPR Signal Enhancement Using Sliding-Window Space-Frequency Matrices

Yavuz¹,

Fouda²,

Teixeira³

2014

PIER

View full text Add to dashboard Cite

show abstract

“…In the other recent literatures, SVM has been applied to solve various problems related to remote sensing. For example, monitoring of biophyssical parameters [10][11][12], vegetation classification [13][14][15], road extraction [16][17][18], and landmine detection [19].…”

Section: Introductionmentioning

confidence: 99%

Support vector machine as a binary classifier for automated object detection in remotely sensed data

Wardaya

2014

IOP Conf. Ser.: Earth Environ. Sci.

View full text Add to dashboard Cite

“…This approach also shows the advantage of detecting outliers when a test sample is outside all the different SVDD. In multitemporal analysis, novelty detection approaches have been considered for oil slick detection with SAR images using OC-SVM and wavelet decomposition [20], for landmines detection from Ground-Penetrating Radars using OC-SVM [21] or for fire detection using SVDD initialized with CVA [22].…”

Section: Introductionmentioning

confidence: 99%

Semi-Supervised Novelty Detection Using SVM Entire Solution Path

Morsier

Tuia

Borgeaud

et al. 2013

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Abstract-Very often, the only reliable information available to perform change detection is the description of some unchanged regions. Since sometimes these regions do not contain all the relevant information to identify their counterpart (the changes), we consider the use of unlabeled data to perform Semi-Supervised Novelty detection (SSND). SSND can be seen as an unbalanced classification problem solved using the Cost-Sensitive Support Vector Machine (CS-SVM), but this requires a heavy parameter search. We propose here to use entire solution path algorithms for the CS-SVM in order to facilitate and accelerate the parameter selection for SSND. Two algorithms are considered and evaluated. The first one is an extension of the CS-SVM algorithm that returns the entire solution path in a single optimization. This way, the optimization of a separate model for each hyperparameter set is avoided. The second forces the solution to be coherent through the solution path, thus producing classification boundaries that are nested (included in each other). We also present a low density criterion for selecting the optimal classification boundaries, thus avoiding the recourse to cross-validation that usually requires information about the "change" class. Experiments are performed on two multitemporal change detection datasets (flood and fire detection). Both algorithms tracing the solution path provide similar performances than the standard CS-SVM while being significantly faster. The low density criterion proposed achieves results that are close to the ones obtained by cross-validation, but without using information about the changes.

show abstract

An abrupt change detection algorithm for buried landmines localization

Cited by 46 publications

References 14 publications

GPR Signal Enhancement Using Sliding-Window Space-Frequency Matrices

GPR Signal Enhancement Using Sliding-Window Space-Frequency Matrices

Support vector machine as a binary classifier for automated object detection in remotely sensed data

Semi-Supervised Novelty Detection Using SVM Entire Solution Path

Contact Info

Product

Resources

About