DETECTION OF LOCAL NOISE IN THE TEXTURAL IMAGES

Journal Title: Journal of Science And Arts - Year 2008, Vol 8, Issue 1

Abstract

The paper presents a classification and identification method for the local noise in the textural images. To identify local noise in the textural images, a reference database with the textural images is used. For the region recognition with local noise, a decision theoretic method is used. The features are of the statistic type and derive from the medium co-occurrence matrices: contrast, energy, entropy, homogeneity, and variance. The algorithm is implemented in a mathematical software and allows the simultaneously display of both the investigated regions pairs, and the Euclidian distance between them. Our experimental results indicate that the five selected features for texture classification give good results to identify the local noise. The results also confirm that the distances between the similar regions are relatively small and the distances between regions from different textured images with local noise are relatively great.

Authors and Affiliations

NICOLETA ANGELESCU, IULIAN UDROIU, CRISTIAN FLUIERARU, CACIULA ION

Keywords

Related Articles

ON AN INEQUALITY FOR THE MEDIANS OF A TRIANGLE

In this paper, we give two new simpler proofs of a sharp inequality for the medians of a triangle. We also establish two new inequalities by using this sharp inequality. Some related conjectures checked by the computer a...

WAXD INVESTIGATION OF THE SECOND ORDER RESIDUAL STRESS INTO THREE HIGH ALLOYED STEELS

Mechanical processing like milling, polishing, grinding, etc generate internal residual stresses of the first order (macrostresses) at bulk level and of the second order (σ11) at the crystallite level. The paper address...

ON VIBRATION EXPOSURE MONITORING AT INDUSTRIAL INTENSIVE POLLUTANT AREAS

Main objective of this study consisted by identification, evaluation and characterization of a direct linkage between the vibration levels at the intensive pollutant industrial sources and the vibration exposure of the p...

APPLICATIONS OF Itô STOCHASTICALLY INTEGRAL IN THE EVALUATION OF THE EXPECTATION VALUES OF BROWNIAN MOTION

In the paper below are evaluated the expectation values of the natural powers of Brownian motion. It is given the definition of these evaluations and will be demonstrated the recurrent relation between the terms of the s...

PROPAGATORS AND DILATIONS ON PSEUDO-HILBERT SPACES

In this paper we shall try to transpose the conditions of the existence of propagators for kernel on *-semigroups, notion introduced by P. Masani [6]. We will start with a few additional observations concerning *-represe...

Download PDF file
  • EP ID EP102398
  • DOI -
  • Views 247
  • Downloads 0

How To Cite

NICOLETA ANGELESCU, IULIAN UDROIU, CRISTIAN FLUIERARU, CACIULA ION (2008). DETECTION OF LOCAL NOISE IN THE TEXTURAL IMAGES. Journal of Science And Arts, 8(1), 5-9. https://www.europub.co.uk/articles/-A-102398