A model of digital image preprocessing based on pixel clustering by modified Ward's method
Автор: Khanykov I.G., Kharinov M.V.
Журнал: Вестник Бурятского государственного университета. Математика, информатика @vestnik-bsu-maths
Рубрика: Математическое моделирование и обработка данных
Статья в выпуске: 2, 2017 года.
Бесплатный доступ
The article deals with the model of detecting the dichotomous hierarchy of objects in the image. The model based on Ward's method of pixel clustering is intended for preprocessing images of scenes of any content. It is proposed to test the results for credibility by processing of stereopairs. We discuss the features of pixel clustering by Ward's method and compared with the known solution accepted as a prototype.
Image segmentation, pixel clustering, integral square error, minimization, piecewise constant approximation, hierarchical sequence, convex sequence of values, double scaled model, ward's method, mamford-shah segmentation model
Короткий адрес: https://sciup.org/14835221
IDR: 14835221 | DOI: 10.18101/2304-5728-2017-2-61-70
