Watershed transformation frequently produces over-segmented images. The authors propose a solution to this problem. It utilizes hierarchical cluster analysis for grouping watersheds which are treated as objects characterized by a number of attributes. Initially the watershed merging method was meant only for gray-scale images, but later it was adapted for colour images. This paper presents further extension of the method that allows it to either automatically select the numberof classes or to provide a hint as to which numbers in a specified range should be considered first.Segmentation quality assessment functions for colour images are presented. The results obtained using an extended watershed merging method are discussed. The examples of segmentations selected by the method, along with the graphs of assessment functions, are shown.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
W pracy przedstawiono przegląd najskuteczniejszych metod detekcji (wykrywania) krawędzi obiektów w obrazach kolorowych. Zaprezentowano skuteczność wektorowych operatorów porządkowych i wektorowych operatorów różnicowych oraz ich odmiany rozszerzone o różnego typu filtry. Detektory uzyskane w ten sposób są w różnym stopniu skomplikowane obliczeniowo oraz odporne na szum impulsowy i gaussowski. W przypadku operatorów DV dodanie członów filtrujących istotnie podnosi ich odporność na obecność szumu. Najlepsze wyniki ekstrakcji krawędzi w zaszumionych obrazach kolorowych uzyskano, stosując operator DV_&aL-trim.
EN
The paper offers review of most effective methods of detecting edges in noised colour images. Presented has been effectiveness of vector order operators and vector difference operators, as well as variants thereof enriched with filters of different types. Detectors gained in this way are complicated to different degrees from the viewpoint of computations, and resistant to impulsive and Gaussian noises. In the case of DV operators, addition of filtering elements significantly increases resistance thereof to the presence of noise. Best effects of edge extraction in noised images have been gained using the DV_ Alpha-trim operator.
Methods used for colour images filtering are not always just simple extensions of the procedures known for greyscale images. Filtration of colour images considering only luminance information not always leads to desired results. In many cases proper choice of colour space used during filtration plays an important role. In that case pixels are represented by vectors of usually three components in given colour space. In the paper some results obtained by conditional ordering techniques for median filtering are compared. Additionally using varying colour representations and varying weights in adaptive median filtering is considered.
4
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
A discrete two-dimensional Fourier transform is presented, based on quaternion numbers (a generalisation of the complex numbers with three imaginary components). This transform allows a colour image in, say, RGB format, to be transformed as a whole, rather than as separate color components. Quaternion numbers are briefly reviewed, the transform is defined, and its two-dimensional cosinusoidal basis functions. The paper concludes with a discussion of a known problem with the transform : that a reversed image does not have a conjugate transform.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.