Open AccessDIGITAL IMAGING BASED ON FRACTAL THEORY AND ITS SPATIAL DIMENSIONALITY
Author(s) -
Yuan Tian,
Gaoyuan Cui,
Harry Morris
Publication year - 2020
Publication title -
fractals
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.654
H-Index - 44
eISSN - 1793-6543
pISSN - 0218-348X
DOI - 10.1142/s0218348x20400149
Subject(s) - fractal dimension , fractal , digital geometry , dimension (graph theory) , artificial intelligence , point (geometry) , fractional brownian motion , digital image , curse of dimensionality , image processing , computer science , computer vision , fractal analysis , digital image processing , edge detection , mathematics , geometry , image (mathematics) , brownian motion , mathematical analysis , statistics , pure mathematics
Due to the complexity of digital imaging targets and imaging conditions, fractal theory techniques in existing digital imaging systems still have various shortcomings. In this paper, a digital imaging processing method based on fractal theory is proposed for the first time. For X-ray images, the rapid calculation method of H-parameters is derived based on the fractional Brownian random field model. The H-parameters of X-ray images are calculated point by point. After that, all the singular points are connected, which is the edge of the defect in the image. We apply this method to analyze and process the X-ray images with defects such as missing joints, skins and hollows. Secondly, by means of fractal geometry, the contour slice measurement of the digital imaging space of this fractal is studied. The approximate index value is the digital imaging section profile dimension (D1 dimension) and the section shadow dimension (D2 dimension), so that the dimension determines the complexity of the form and detail of digital imaging. Finally, it can be seen from the experimental results that this method is effective and explores a new way for the development of digital imaging technology. At the same time, it is of great significance to the automatic pattern recognition of the application.