Method for extracting contours from biomedical images using filters and criteria for their effectiveness
Sergii Pavlov, Iosyph Saldan, Serhii Zlepko, Leonid Tymchenko, Dina Vovkotrub , Nataliia Kokriatska , Dmytro StepaniukThe paper is devoted to the division of the background and the object biomedical image. Consider an analysis on an example of a biomedical image processed using the developed software "Biomedical Imaging Analyzer". Consider the value of the PSNR criterion and the normalized NK correlation for the proposed software. From the results of the efficiency analysis, it should be noted that the value of the PSNR and the normalized NK correlation for the proposed software is higher than for the others, and the MSE mean square deviation and the normalized NMSE deviation are lower, indicating a more precise reproduction of a clear-cut image and a better quality and accurate image reproduction by the proposed method. Thus, the histogram of a biomedical image processed using the Biomedical Image Analyzer software has a clear rupture, unlike histograms of images processed by known methods, in which the spectrum of brightness remained uninterrupted. This indicates a clear separation of the background and object of the biomedical image