Publication:
An adaptive spectrally weighted structure  tensor applied to tensor anisotropic nonlinear  diffusion for hyperspectral images  

dc.contributor.advisor Vélez-Reyes, Miguel
dc.contributor.author Marin-Quintero, Maider J.
dc.contributor.college College of Engineering en_US
dc.contributor.committee Hunt, Shawn
dc.contributor.committee Manien, Vidya
dc.contributor.committee Rivera, Wilson
dc.contributor.department Department of Electrical and Computer Engineering en_US
dc.contributor.representative Maldonado, Francisco
dc.date.accessioned 2019-02-12T15:30:47Z
dc.date.available 2019-02-12T15:30:47Z
dc.date.issued 2012
dc.description.abstract The structure tensor for vector valued images is most often defined as the average of the scalar structure tensors in each band. The problem with this definition is the assumption that all bands provide the same amount of edge information giving them the same weights. As a result non-edge pixels can be reinforced and edges can be weakened resulting in a poor performance by processes that depend on the structure tensor. Iterative processes, in particular, are vulnerable to this phenomenon. In this work, a structure tensor for Hyperspectral Images (HSI) is proposed. The initial matrix field is calculated using a weighted smoothed gradient. The weights are based on the Heat Operator. This definition is motivated by the fact that in HSI, neighboring spectral bands are highly correlated, as are the bands of its gradient. To use the heat operator, the smoothed gradient is modeled as the initial heat distribution on a compact manifold M. A Tensor Anisotropic Nonlinear Diffusion (TAND) method using the spectrally weighted structure tensor is proposed to do two kind of processing: Image regularization known as Edge Enhancing Diffusion (EED) and structure enhancement known as Coherence Enhancing Diffusion (CED). Diffusion tensor and a stopping criteria were also developed in this work. Comparisons between methods show that the structure tensor with weights based on the heat operator better discriminates edges that need to be persistent during the iterative process with EED and produces more complete edges with CED. Remotely sensed and biological HSI are used in the experiments. en_US
dc.description.abstract El tensor de estructura para imágenes vectoriales es comúnmente definido como el promedio de los tensores de estructura que ha sido previamente calculado para cada banda de la imagen. El problema con esta definición es que ella asume que todas las bandas proveen la misma cantidad de información de los bordes. Por lo tanto le da el mismo peso a cada una de las bandas. Como resultado, pixeles que no son bordes son reforzados y los bordes pueden ser debilitados. Esto hace que otros procesos que dependan del tensor de estructura den resultados mediocres. Los procesos iterativos son los más vulnerables a este fenómeno. En este trabajo se propone un tensor de estructura para imágenes HiperEspectrales (IHE). El campo matricial inicial es calculado usando un gradiente suavizado ponderado. Los pesos son basados en el operador de calor. Esta definición es motivada por una propiedad de las IHE y de su gradiente. Esta es: las bandas espectrales que están cercanas son altamente correlacionadas. Para poder hacer uso del operador del calor, el gradiente suavizado es modelado como la distribucion inicial de calor en una variedad compacta denotada por M. Este tensor de estructura será aplicado a la Difusión Anisotrópica No Lineal basada en Tensores (DANT) para hacer Difusión que Preserva Bordes (DPB) y Difusión que Realza Coherencia (DRC). Comparación entre los métodos muestran que el tensor de estructura ponderado con pesos basados en el operador de calor discrimina mejor los bordes con DPB y produce bordes mas completos con DRC. Estos métodos han sido aplicados a IHE sensadas remotamente como tambien imágenes biológicas adquiridas con microscopios hiperespectrales. en_US
dc.description.graduationSemester Summer en_US
dc.description.graduationYear 2012 en_US
dc.description.sponsorship This work was supported by U.S. Department of Homeland Security under Award Number 2008-ST-061-ML0002 and partial support also was provided under DHS Award Number 2008-ST-061-ED0001. This work used facilities of the Bernard M. Gordon Center of Subsurface Sensing and Imaging sponsored by the NSF ERC program under award EEC-9986821. My last year was supported by NASA under Award Number NNX10AM80H. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied of the U.S. Department of Homeland Security or the National Science Foundation or NASA. en_US
dc.identifier.uri https://hdl.handle.net/20.500.11801/1800
dc.language.iso English en_US
dc.rights.holder (c) 2012 Maider J. Marin Quintero en_US
dc.rights.license All rights reserved en_US
dc.subject Hyperspectral images en_US
dc.subject Anisotropic linear diffusion en_US
dc.title An adaptive spectrally weighted structure  tensor applied to tensor anisotropic nonlinear  diffusion for hyperspectral images   en_US
dc.type Dissertation en_US
dspace.entity.type Publication
thesis.degree.discipline Computing and Information Sciences and Engineering en_US
thesis.degree.level Ph.D. en_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
CIIC_MarinQuinteroM_2012.pdf
Size:
6.3 MB
Format:
Adobe Portable Document Format
Description: