An Introduction to the Application of Tensorial Manifold Learning Methods in the Digital Image Processing and Computer Vision

International Journal of Mathematical Modelling & Computations

مقاله 3، دوره 12، 1 (WINTER) - شماره پیاپی 45، خرداد 2022، صفحه 27-35   

نوع مقاله: Full Length Article

شناسه دیجیتال (DOI): 10.30495/ijm2c.2022.1936139.1224

نویسندگان

Hamid Reza Yazdani^{* 1}؛ Ali Reza Shojaeifard²

¹Department of Mathematics, Payame Noor University (PNU), P.O. Box 19395-3697, Tehran, Iran

²Department of Mathematics, Faculty of Sciences, Imam Hossein Comprehensive University, Tehran, Iran.

چکیده

Tensors as vector fields structures and manifolds as great geometrical-topological structures have many applications in the fields of big data analysis. Types of norms, metrics and scalable structures have been defined from various aspects. Nowadays, the hybrid methods between tensorial algorithms and manifold learning (MaL) methods have been attracted some attention. In image and signal processing, from image recovery to face recognition, these methods have appeared very excellent. According to our experiments by MATLAB R2021a, the hybrid algorithms are powerful other than algorithms based on the efficient popular parameters.

کلیدواژه‌ها

Image recovery؛ Image recognition؛ Manifold learning؛ Tensors؛ Tensor completion