Search results for "Pattern Recognition"
showing 10 items of 2301 documents
Projector operators in clustering
2016
In a recent paper the notion of {\em quantum perceptron} has been introduced in connection with projection operators. Here we extend this idea, using these kind of operators to produce a {\em clustering machine}, i.e. a framework which generates different clusters from a set of input data. Also, we consider what happens when the orthonormal bases first used in the definition of the projectors are replaced by frames, and how these can be useful when trying to connect some noised signal to a given cluster.
An application of neural networks to natural scene segmentation
2006
This paper introduces a method for low level image segmentation. Pixels of the image are classified corresponding to their chromatic features.
A fully adaptive multiresolution scheme for image processing
2007
A nonlinear multiresolution scheme within Harten's framework [A. Harten, Discrete multiresolution analysis and generalized wavelets, J. Appl. Numer. Math. 12 (1993) 153-192; A. Harten, Multiresolution representation of data II, SIAM J. Numer. Anal. 33 (3) (1996) 1205-1256] is presented. It is based on a centered piecewise polynomial interpolation fully adapted to discontinuities. Compression properties of the multiresolution scheme are studied on various numerical experiments on images.
A Parallel Approach for Statistical Texture Parameter Calculation
2014
This chapter focusses on the development of a new image processing technique for the processing of large and complex images, especially SAR images. We propose here a new and effective approach that outperforms the existing methods for the calculation of high order textural parameters. With a single processor, this approach is about \(256^{n-1}\) times faster than the co-occurrence matrix approach considered as classical, where \(n\) is the order of the textural parameter for a 256-gray scales image. In a parallel environment made of N processor, this performance can almost be multiply by the factor N. Our approach is based on a new modeling of textural parameters of a generic order \(n>1\) …
Use of Guided Regularized Random Forest for Biophysical Parameter Retrieval
2018
This paper introduces a feature selection method based on random forest -the Guided Regularized Random Forest (GRRF)- which can be used in classification and regression tasks. The method is based on the regularization of the information gain in the random forest nodes to obtain a subset of relevant and non-redundant features. The proposed method is used as a preliminary step In the process of retrieving biophysical parameters from a hyperspectral image. Preliminary experiments show that we can reduce the RMSE of the retrievals by around 7% for the Leaf Area Index and around 8% for the fraction of vegetation cover when compared to the results using random forest features.
Demultiplexing Visible and Near-Infrared Information in Single-Sensor Multispectral Imaging
2016
In this paper, we study a single-sensor imaging system that uses a multispectral filter array to spectrally sample the scene. Our system captures information in both visible and near-infrared bands of the electromagnetic spectrum. Due to manufacturing limitations, the visible filters in this system also transmit the NIR radiation. Similarly, visible light is transmitted by the NIR filter, leading to inaccurate mixed spectral measurements. We present an algorithm that resolves this issue by separating NIR and visible information. Our method achieves this goal by exploiting the correlation of multispectral images in both spatial and spectral domains. Simulation results show that the mean squa…
Therapeutic Drug Monitoring of Kidney Transplant Recipients Using Profiled Support Vector Machines
2007
This paper proposes a twofold approach for therapeutic drug monitoring (TDM) of kidney recipients using support vector machines (SVMs), for both predicting and detecting Cyclosporine A (CyA) blood concentrations. The final goal is to build useful, robust, and ultimately understandable models for individualizing the dosage of CyA. We compare SVMs with several neural network models, such as the multilayer perceptron (MLP), the Elman recurrent network, finite/infinite impulse response networks, and neural network ARMAX approaches. In addition, we present a profile-dependent SVM (PD-SVM), which incorporates a priori knowledge in both tasks. Models are compared numerically, statistically, and in…
Ranking of Brain Tumour Classifiers Using a Bayesian Approach
2009
This study presents a ranking for classifers using a Bayesian perspective. This ranking framework is able to evaluate the performance of the models to be compared when they are inferred from different sets of data. It also takes into account the performance obtained on samples not used during the training of the classifiers. Besides, this ranking assigns a prior to each model based on a measure of similarity of the training data to a test case. An evaluation consisting of ranking brain tumour classifiers is presented. These multilayer perceptron classifiers are trained with 1H magnetic resonance spectroscopy (MRS) signals following a multiproject multicenter evaluation approach. We demonstr…
Towards farm-level health management of offshore wind farms for maintenance improvements
2015
This paper studies a conceptual architecture for health management of offshore wind farms. To this aim, various necessary enablers of a health management sys- tem are presented to improve reliability and availability while optimizing maintenance costs. The main focus lies on improving existing condition monitoring systems based on concepts of condition-based maintenance and relia- bility centered maintenance. A brief review of the rel- evant state-of-the-art is presented and gaps to be filled towards realization of such health management system are discussed.
FIR Kramers–Kronig transformers for relaxation data conversion
2006
It is shown that relaxation data conversion by the Kramers-Kronig (KK) relations can be treated as a filtering problem of band-unlimited relaxation signals in the Mellin transform domain. Based on this concept, KK relations are implemented in the form of FIR filters with the logarithmic sampling. It is demonstrated that KK transformers have sampling rate dependent impulse and frequency responses and only calculation of the imaginary part from the real part can be implemented by a computationally realisable filter. The performance of different types of transformers is studied.Approximately inversely proportional relationship is established between the error and the frequency range of input s…