Search results for "Convolution"
showing 10 items of 334 documents
Convolutional Neural Networks for Multispectral Image Cloud Masking
2020
Convolutional neural networks (CNN) have proven to be state of the art methods for many image classification tasks and their use is rapidly increasing in remote sensing problems. One of their major strengths is that, when enough data is available, CNN perform an end-to-end learning without the need of custom feature extraction methods. In this work, we study the use of different CNN architectures for cloud masking of Proba-V multispectral images. We compare such methods with the more classical machine learning approach based on feature extraction plus supervised classification. Experimental results suggest that CNN are a promising alternative for solving cloud masking problems.
Correction of the tip convolution effects in the imaging of nanostructures studied through scanning force microscopy
2014
AFM images are always affected by artifacts arising from tip convolution effects, resulting in a decrease in the lateral resolution of this technique. The magnitude of such effects is described by means of geometrical considerations, thereby providing better understanding of the convolution phenomenon. We demonstrate that for a constant tip radius, the convolution error is increased with the object height, mainly for the narrowest motifs. Certain influence of the object shape is observed between rectangular and elliptical objects with the same height. Such moderate differences are essentially expected among elongated objects; in contrast they are reduced as the object aspect ratio is increa…
Production and Characterization of Oxyhydroxyapatites
2018
The amount and alignment of hydroxyl ions influence the bioactivity of hydroxyapatite. Hydroxyl ions in hydroxyapatite are the most mobile and upon heating are the first to leave the lattice to form oxyhydroxyapatite. This work describes a method for producing hydroxyapatite with different amounts of hydroxyl ions, and reports on the changes in Fourier transform infrared absorption at increasing level of dehydroxylation. Detailed analysis of spectra in the 500 – 700 cm-1 range showed a peak shift for the hydroxyl ion absorption line at 632 cm-1 to 637 cm-1 and an increase in the wavenumber for the phosphate line at 575 cm-1.
Hydrothermal Processing for Increasing the Hydroxyl Ion Concentration in Hydroxyl Depleted Hydroxyapatite
2018
Thermal spraying is commercially used to produce hydroxyapatite coatings, but the high temperature depletes hydroxyl ions in the structure. To return hydroxyapatite to its original state, it is necessary to restore the hydroxyl ion content in the structure. In this work, the effect of hydrothermal treatment on the hydroxyl ions was investigated in hydroxyapatite powder and thermally sprayed hydroxyapatite coatings. Samples were hydrothermally treated at 200 °C for 24 h and 48 h. Chemical phases were determined by X-ray diffraction, functional groups and hydroxyl ion concentration was examined by Fourier transform infra-red spectroscopy. Results showed that hydrothermal processing of hydroxy…
Optical sectioning with a Wiener-like filter in Fourier integral imaging microscopy
2018
Non-scanning, single-shot, 3D integral microscopy with optical sectioning is presented. The method is based on the combination of Fourier-mode integral microscopy with a 3D deconvolution technique. Specifically, the refocused volume provided by a regular back-projection algorithm is 3D deconvolved with a synthetic 3D impulse response function that takes into account the number and positions of the elemental images. The use of this hybrid technique provides a stack of true-color depth-refocused images with significant gain of optical sectioning. The stack can be used, among other applications, to inspect inside the thick microscope specimen, to calculate collections of perspective views wit…
Partially Coherent Optical Diffraction Tomography Toward Practical Cell Study
2021
Optical diffraction tomography (ODT) is a computational imaging technique based on refractive index (RI) contrast. Its application for microscopic imaging of weakly absorbing and scattering samples has been demonstrated by using a specially designed holographic microscope with angular scanning of the coherent sample illumination direction. Recently, an alternative low cost technique based on partially coherent sample illumination (PC-ODT), which is compatible with the conventional wide-field transmission microscope, has been established. In this case, the 3D refractive index distribution of the sample is obtained by deconvolution from a single stack of through-focus intensity images. The pe…
Deconvolution of the Effects of Binary Associations and Collective Assemblies on the Rheological Properties of Entangled Side-Chain Supramolecular Po…
2019
The properties and function of supramolecular polymer networks are determined not only by pairwise interchain transient associations but also by chain entanglement and nanoscopic phase separation of the associative groups. To unravel the impact and interplay of these different factors, we devise a set of model supramolecular polymer networks in which the number of entanglements and the density of associative groups are systematically varied. Rheological data show that by increasing the density of associative groups, the plateau modulus grows to a steady level and extends over a distinct frequency range. This is credited to the presence of binary associations with unique partner exchange tim…
Study of As and Tl high-frequency electrodeless lamps for Zeeman absorption spectroscopy
2020
High-frequency electrodeless light sources are known as bright radiators with the line spectrum, characterized by high intensities and narrow line shapes. The lamp balloons are mostly made of quartz and filled with a metal and buffer gas. These light sources must be optimized for each application in accordance with the specific requirements of radiation quality, lifetime, and stability. This work is devoted to the diagnostic of high-frequency electrodeless light sources for their use in high precision atomic absorption analyzers and other applications like disinfection. The spectral line intensities and profiles were studied in discharges with arsenic and thallium fillings. Special attentio…
A general framework for a class of non-linear approximations with applications to image restoration
2018
Este artículo se encuentra disponible en la página web de la revista en la siguiente URL: https://www.sciencedirect.com/science/article/abs/pii/S0377042717301188 Este es el pre-print del siguiente artículo: Candela, V., Falcó, A. & Romero, PD. (2018). A general framework for a class of non-linear approximations with applications to image restoration. Journal of Computational and Applied Mathematics, vol. 330 (mar.), pp. 982-994, que se ha publicado de forma definitiva en https://doi.org/10.1016/j.cam.2017.03.008 This is the pre-peer reviewed version of the following article: Candela, V., Falcó, A. & Romero, PD. (2018). A general framework for a class of non-linear approximations with applic…
Matrix Shuffle- Exchange Networks for Hard 2D Tasks
2021
Convolutional neural networks have become the main tools for processing two-dimensional data. They work well for images, yet convolutions have a limited receptive field that prevents its applications to more complex 2D tasks. We propose a new neural model, called Matrix Shuffle-Exchange network, that can efficiently exploit long-range dependencies in 2D data and has comparable speed to a convolutional neural network. It is derived from Neural Shuffle-Exchange network and has O(log N) layers and O(N ^ 2 log N) total time and O(N^2) space complexity for processing a NxN data matrix. We show that the Matrix Shuffle-Exchange network is well-suited for algorithmic and logical reasoning tasks on …