Wyniki wyszukiwania - Biblioteka Nauki

1

Adaptive generalized vector median

100%

Smołka B.

|

2008

|

tom Vol. 12

157--162

EN

In this paper, a new adaptive filter intended for the attenuation of impulse noise in colour images is proposed. The new filtering design is based on the concept of a peer group of pixels sharing similar chromatic properties. The novel approach adaptively determines the size of the peer group which minimizes the aggregated distance to its members, utilizing the Fisher linear discriminant. The analysis of the obtained noise reduction results leads to the conclusion, that the new filter is capable of reducing even strong impulse noise, while preserving and even enhancing the edges of colour images. This unique property of the proposed filtering design is shown on examples of colour biomedical images.

2

Modified biased anisotropic diffusion for color image processing

100%

Smołka B.

|

tom z. 150

197-206

EN

In this paper a new approach to the problem of noise reduction in color images is presented. The new technique is based on a modification of the biased anisotropic diffusion. In the modified iterative scheme, the input noisy signal is replaced by an image processed by a nonlinear multichannel filter. The experiments revealed, that the proposed solution significantly excels over the standard anisotropic diffusion methods in case of the mixed Gaussian and impulse noise contaminating the color image. The main benefits of the proposed approach are its rapid convergence to the final stable state, low computational complexity and good performance in complex noise scenarios.

PL

W artykule przedstawiono nowe podejście do problemu redukcji szumów w barwnych obrazach cyfrowych. Nowa metoda filtracji oparta jest na modyfikacji obciążonej dyfuzji anizotropowej. W zmodyfikowanym algorytmie iteracyjnym wejściowy obraz zakłócony zastępowany jest przez wyjście wielokanałowego filtru nieliniowego. Przeprowadzone eksperymenty wykazały, że zaproponowana technika znacząco przewyższa standardową metodę dyfuzji anizotropowej w przypadku obrazów zakłóconych przez mieszany szum gaussowski i impulsowy. Główną zaletą zaproponowanej metody jest jej szybka zbieżność do stabilnego stanu końcowego, niska złożoność obliczeniowa i duża efektywność w przypadku szumów o skomplikowanej strukturze.

3

Biomedical image colorization using pixel membership propagation

63%

Popowicz A. , Smołka B.

Studia Informatica

|

2016

|

tom Vol. 37, nr 3A

7--19

EN

In the paper we present a new algorithm of biomedical image colorization based on distance transformation and modified bilateral filtering approach. The method utilizes the scribbles inserted by the user to properly cover the image regions with desirable colors. We present the idea of our algorithm, explain the role of tunable parameters and provide some examples of biomedical image colorization using our approach.

PL

W artykule przedstawiono nową technikę koloryzacji, wykorzystującą transformatę dystansową oraz modyfikację filtru bilateralnego. Proponowana metoda opiera się na wskaźnikach koloru wprowadzanych przez użytkownika w celu zgrubnego początkowego zaznaczenia oczekiwanych kolorów dla poszczególnych elementów obrazów. W artykule wyjaśniono zasadę działania algorytmu, role jego parametrów oraz przedstawiono przykłady barwnych obrazów biomedycznych uzyskanych dzięki proponowanej nowej technice.

4

Image segmentation and matting based on the extended distance transform

63%

Łagodziński P. , Smołka B.

|

2008

|

tom Vol. 12

151--156

EN

In this paper we present a novel semi-automatic method for image segmentation and matting that utilizes the newly proposed generalized distance transform to determine the hybrid distance between the foreground and background regions within the image. This distance values are then used to estimate the alpha matting coefficient needed for composing new image by blending the foreground objects into a new background image scene. The effectiveness of the proposed algorithm allows the user to work interactively and to obtain the desired results promptly after marking the regions by scribbling the image. In the paper we show, that the proposed method yields satisfactory results for gray scale and colour images.

5

Optimal classification method for smiling vs neutral facial display recognition

63%

Nurzyńska K. , Smołka B.

Journal of Medical Informatics & Technologies

|

2014

|

tom Vol. 23

87--94

EN

Human face depicts what happens in the soul, therefore correct recognition of emotion on the basis of facial display is of high importance. This work concentrates on the problem of optimal classification technique selection for solving the issue of smiling versus neutral face recognition. There are compared most frequently applied classification techniques: k-nearest neighbourhood, support vector machines, and template matching. Their performance is evaluated on facial images from several image datasets, but with similar image description methods based on local binary patterns. According to the experiments results the linear support vector machine gives the most satisfactory outcomes for all conditions.

6

Medical image colorization

63%

Łagodziński P. , Smołka B.

|

2007

|

tom Vol. 11

47--57

EN

Colorization is a term used to describe a computerized process for adding colour to black and white pictures, movies or TV programs. This process involves replacing a scalar value that represents pixels' intensity or luminance by a vector in a three dimensional colour space with luminance, saturation and hue or simply RGB. The colorization process is also used to convert the grey scale to colour on medical images. Colour increases the visual appeal of an image and it also makes a medical visualization more attractive. Changes in colour are more easily perceived then changes in shades of grey and therefore this procedure makes the interpretation and understanding of the image easier. Since the mapping between intensity and colour has no inherently "correct" solution, human interaction and/or external information usually plays a large role. In this paper we present a novel colorization method that takes advantage of the morphological distance transformation and image structures to automatically propagate the colour scribbled by the user within the grey scale image. The effectiveness of the algorithm allows the user to work interactively and obtain the desired results promptly after providing the colour. In the paper we show that the proposed method allows for high quality colorization results for still images without precise segmentation.

7

Non-local Mean-Shift filter for the reduction of multiplicative noise in digital images

63%

Kusnik D. , Smołka B.

Journal of Medical Informatics & Technologies

|

2014

|

tom Vol. 23

103--110

EN

In this paper a new method for the reduction of multiplicative noise in digital images is described. The proposed algorithm is a modification of the Mean-Shift (MS) filter which is based on the concept of the Non-Local Means (NLM) denoising. The proposed algorithm does not focus on single pixels only, as in the case of the mean-shift technique, but also on their neighborhoods. The performance of the novel approach is experimentally verified and the obtained results prove that the new design is superior both to the MS and NLM techniques.