Improvement of diagnostic efficiency of pulmonary diseases is currently a very urgent task. For building of automated systems of recognition and classification of radiographic images it is required to perform the process of localization of pathological formations on photofluorograms of chest. The correct choice of the right method for segmentation of examined radiographic images ensures high quality and improves the accuracy of diagnostic decisions to be taken. The article contains a structural diagram of the device wherein the method of automatic segmenting of chest photofluorogram is implemented. To build algorithms for detection of pathological formations there has been used a histogram of image brightness created in a dedicated window. Graphic primitives are determined to approximate the fluorogram brightness histogram in the optimum size window for the analysis. Based on the information obtained the vector or informative features is formed. A decision to refer the vector of informative features to a particular class is effected by a trained double alternative classifier. When the analysis window is moving across the image the binary image is getting finally formed.


    Access

    Download


    Export, share and cite



    Title :

    METHOD OF AUTOMATIC SEGMENTATION OF CHEST X-RAY TESTS FINDINGS ON THE BASIS OF PRESENTATION OF IMAGE BRIGHTNESS IN A SLIDING WINDOW


    Contributors:


    Publication date :

    2017




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    Unknown




    Sliding window

    HICK ROBERT | European Patent Office | 2018

    Free access

    SLIDING WINDOW

    HICK ROBERT | European Patent Office | 2016

    Free access

    SYSTEM FOR CONTROLLING BRIGHTNESS OF VEHICLE WINDOW

    CHOI JIN HO | European Patent Office | 2016

    Free access

    SLIDING WINDOW ASSEMBLY

    LAHNALA DAVID W / BENNETT DANIEL | European Patent Office | 2020

    Free access

    SLIDING WINDOW ASSEMBLY

    LAHNALA DAVID W / BENNETT DANIEL / BACH JAMES W | European Patent Office | 2016

    Free access