Wyniki wyszukiwania - BazTech

1

Could k-NN classifier be Useful in tree leaves recognition?

Horaisová K.

Archives of Control Sciences

|

2014

|

Vol. 24, no. 2

177--192

EN

This paper presents a method for affine invariant recognition of two-dimensional binary objects based on 2D Fourier power spectrum. Such function is translation invariant and their moments of second order enable construction of affine invariant spectrum except of the rotation effect. Harmonic analysis of samples on circular paths generates Fourier coefficients whose absolute values are affine invariant descriptors. Affine invariancy is approximately saved also for large digital binary images as demonstrated in the experimental part. The proposed method is tested on artificial data set first and consequently on a large set of 2D binary digital images of tree leaves. High dimensionality of feature vectors is reduced via the kernel PCA technique with Gaussian kernel and the k-NN classifier is used for image classification. The results are summarized as k-NN classifier sensitivity after dimensionality reduction. The resulting descriptors after dimensionality reduction are able to distinguish real contours of tree leaves with acceptable classification error. The general methodology is directly applicable to any set of large binary images. All calculations were performed in the MATLAB environment.

2

An outline of the system for computer-assisted diagnosis based on binary erythrocytes shapes

Frejlichowski D.

Metody Informatyki Stosowanej

|

2008

|

nr 4 (Tom 17)

63--69

EN

The paper presents some initial results of the work on the system for Computer-Assisted Diagnosis (CAD) of diseases involving erythrocytes (e.g. anaemia). The approach is based on binary shapes of red blood cells (RBCs), extracted from digital microscopic images. This comes from the fact that some diseases have their origin in deformation of RBCs shapes, what makes the proper delivering of oxygen to body tissues impossible. In result the blood circulation is non-regulated. The approach under development was examined and implemented in form of the prototype in Matlab environment.

3

A Co-processor for Computing the Euler Number of a Binary Image using Divide-and-Conquer Strategy

Dey S., Bhattacharya Bhargab.B., Kundu M.K., Bishnu A., Acharya T.

Fundamenta Informaticae

|

2007

|

Vol. 76, nr 1-2

75-89

EN

Euler number is a fundamental topological feature of an image. The efficiency of computation of topological features of an image is critical for many digital imaging applications such as image matching, database retrieval, and computer vision that require real time response. In this paper, a novel algorithm for computing the Euler number of a binary image based on divide-and-conquer paradigm, is proposed, which outperforms significantly the conventional techniques used in image processing tools. The algorithm can be easily parallelized for computing the Euler number of an N ×N image in O(N) time, with O(N) processors. Using a simple architecture, the proposed method can be implemented as a special purpose VLSI chip to be used as a co-processor.

4

Wybrane metody rozwiązywania problemów pojawiających się w trakcie detekcji granic ziarn.

Szala J.

Inżynieria Materiałowa

|

2002

|

R. XXIII, nr 3

89-96

PL

Podstawowym warunkiem stosowania do ilościowej oceny struktury komputerowych metod pomiaru jest poprawne odwzorowanie w finalnym obrazie binarnym mierzonych elementów struktury. Z praktyki wiadomo, że spełnienie tego warunku jest w wielu przypadkach niezwykle trudne. Dotyczy to w szczególności granic ziarn w tworzywach jednofazowych. W artykule przedstawiono niektóre problemy pojawiające się w trakcie detekcji granic ziarn w tych tworzywach oraz metody ich eliminowania. Mimo spełnienia zaleceń podanych w systemach eksperckich używanych w praktyce, na powierzchni zgładów może pozostać pewna liczba drobnych rys. Gdy rysy są przypadkowo rozmieszczone na powierzchni zgładu, dobre rezultaty dają procedury oparte na tzw. przekształceniu Hougha. W artykule przedstawiono również metodę selekcji granic bliźniaków. Na podstawie wyników obserwacji szeregu struktur rzeczywistych stwierdzono, że w zdecydowanej większości przypadków granica ziarna nie zmienia lub zmienia tylko w niewielkim stopniu swój przebieg w punkcie potrójnym, w którym styka się z granicą bliźniaka. Gdy kąt beta, jaki tworzą ze sobą wektory o początku w punkcie potrójnym styczne do granic ziarn w tym punkcie, jest zbliżony do 180 stopni, trzecia granica wychodząca z tego punktu potrójnego jest granicą bliźniaka. Kąt beta przyjęto zatem jako kryterium detekcji tych granic. Stwierdzono, że zaproponowana metodyka jest najbardziej skuteczna wtedy, gdy granice bliźniaków uzna się te, dla których kąt beta zawarty jest w przedziale 150/200 stopni. Zapewnia to prawidłową detekcję około 85% granic bliźniaków oraz około 70% granic ziarn.

EN

The basic condition to apply computer measurement methods for a quantitative evaluation of a structure is correct mapping of the measured structure elements in the final binary image. It is known from practice that the fulfillment of this condition is, in many cases, extremly difficult. This refers in particular to grain boundaries in single-phase materials. This article presents some of the problems that occur when detecting grain boundaries in those materials as well as methods of their elimination. In spite of implementing the recommendations given in the expert systems used in preparation, a number of fine scratches can be left on the microsection surfaces. Where scratches are located on the microsection surface at random, good results are obtained when applying procedures based on the so-called Hough's transformation. The article also presents the method of twins' boundaries selection. Based on the results of observation of a number of real structures it was found that in the great majority of cases the grain boundary does not change, or changes only to a small degree, its course in the triple point where it contacts the twin boundary. When angle beta, which is created by the vectors having their origins in the triple point, tangent to the grain boundaries in that point, is close to 180 degrees, the third boundary coming out from that triple point is the boundary of the twin. The beta angle is thus assumed to be the criterion of those boundaries detection. It was found that the methodology proposed is most effective when as the twins boundaries are regarded those for which angle beta is contained in the range 150/200 degrees. This ensures correct detection of ca. 85% of twin boundaries and ca. 70% of grain boundaries.

5

Automatic selection of structuring element for object classification through morphology

Chanda B., Bhattacharyya R.

Image Processing & Communications

|

1999

|

Vol. 5, no 1

67-79

EN

How to select a structuring element for a given task is one of the most frequently asked questions in morphology. Present work tries to give a solution for a restricted class of problems, namely shape classification. In this work an algorithm that extracts distinctive structure of each of a given set of objects is proposed. The proposed algorithm is based on a new approach for computing distance transform.