Bigand A.,LISIC ULCO |
Colot O.,CNRS Lille Research Center in Informatics, Signal and Automatic control
Fuzzy Sets and Systems | Year: 2015
Models based on interval-valued fuzzy sets make it possible to manage numerical and spatial uncertainty in grey-scale values of image pixels. In a recent paper, we proposed a method that links the ultrafuzziness index (that makes it possible to take into account some uncertainty, like noise, and inherent to image capture) with impulse noise removal. However, computing with interval-valued fuzzy sets requires assigning their membership functions (MFs). The present article proposes a novel method for the generation of membership functions, based on image histogram, to remedy that drawback and it complements our previous study. The performance of the method is evaluated by applying this technique to the particular case of Gaussian noise detection and reduction, which remains a crucial issue for image processing. Experimental results have demonstrated that the proposed method leads to interval-valued fuzzy filters that are comparable with some well-known conventional and fuzzy filters, especially in the case of iterative filtering methods. Image details are preserved while reducing Gaussian noise, and the link between image noise and interval-valued fuzzy sets is thus verified. The main advantage of the proposed technique is to use basic image information, namely an image histogram, which is easy to obtain. © 2015 Elsevier B.V.
Constantin J.,Lebanese University |
Bigand Andre.,LISIC ULCO |
Constantin I.,Lebanese University |
Hamad D.,LISIC ULCO
Neurocomputing | Year: 2015
Global illumination methods based on stochastic techniques provide photo-realistic images. However, they are prone to stochastic perceptual noise that can be reduced by increasing the number of paths as proved by Monte Carlo theory. The problem of finding the required number of paths in order to ensure that human observers cannot perceive any noise is still open. Until now, we do not know precisely which features are considered by the human visual system (HVS) for the evaluation of the image quality. This paper proposes a relevant model to predict which image highlights perceptual noise by using fast relevance vector machine (FRVM). This model can then be used in any progressive stochastic global illumination method in order to find the visual convergence threshold of different parts of any image. A comparative study of this model with experimental psycho-visual scores demonstrates the good consistency between these scores and the model quality measures. The proposed model has also been compared with SVM model and gives competitive performances. © 2015 Elsevier B.V.
Nasser A.,Lebanese University |
Hamad D.,LISIC ULCO
2015 5th International Conference on Digital Information and Communication Technology and Its Applications, DICTAP 2015 | Year: 2015
We apply, in this article, a new method to identify outliers from a dataset. It consists to use the K-means clustering algorithm on the smallest principal components provided by the kernel principal components analysis. Two leading methods commonly used in the domain namely the standard deviation and the Tukey boxplot are tested and compared to our method. The experiments on artificial and real datasets show that our approach better detects outliers than the two classical methods. © 2015 IEEE.