A RGB-D target recognition method based on quaternion generalized discriminant analysis

Abstract

The invention relates to a RGB-D target recognition method based on quaternion generalized discriminant analysis, including quaternion-based RGB-D image representation method, define quaternion generalized discriminant analysis, two-way quaternion generalized discriminant analysis recognition method base on average row and average column; wherein the quaternion-based RGB-D image representation method solves that problems of data redundancy and extra computational overhead when the existing quaternion color image representation method adopts four-dimensional quaternion to represent three-dimensional color images, The kernel function is introduced into the field of quaternion subspace analysis, and the quaternion generalized discriminant analysis is defined, which solves the problem that theexisting quaternion subspace analysis algorithms are not effective in processing quaternion nonlinear signals. Finally, the bidirectional quaternion generalized discriminant analysis based on averagerows and average columns is used to eliminate the influence that aiming at the RGB-D recognition process, the computational complexity of quaternion kernel matrix eigendecomposition is too large, which improves the recognition effect of the target recognition method.

Description

technical field [0001] The invention relates to the field of civil engineering, in particular to a RGB-D object recognition method based on quaternion generalized discriminant analysis. Background technique [0002] The color image quaternion representation method (QR) is widely used in processing color images: each pixel of a color image is represented by a pure quaternion, and three channels are used as its three imaginary parts [Subakan O N, Vemuri B C .Aquaternion framework for color image smoothing and segmentation[J].International Journal of Computer Vision,2011,91(3):233-250.]. Its advantage is that the color image is regarded as a vector whole and the internal relationship between the three channels is considered. However, when a four-dimensional quaternion is used to represent a three-dimensional color image, its fourth dimension will generate storage redundancy and additional computational costs [Asefa D, Mansinha L, Tiampo K F, et al.The trinion Fourier transform...

AI Technical Summary

Problems solved by technology

However, there are still two problems to be solved in this work. On the one hand, this work uses traditional QR, which does not solve the data redundancy problem of QR representation of color images mentioned above. The processing effect of arity nonlinear signals is not very ideal

Images