二维实值与复值离散GABOR变换的比较

A Comparison between the 2D Real-valued and the Complex-valued Discrete Gabor Transforms

本文介绍了作者曾提出的快速二维实值离散 Ｇａｂｏｒ变换（ＲＤＧＴ）方法，以及 ＲＤＧＴ系数与复值离散 Ｇａｂｏｒ变换（ＣＤＧＴ）系数之间的关系，并在不同的等效窗宽条件下，通过对图像的二维ＲＤＧＴ系数 与二维ＣＤＧＴ系数熵的比较，指出二维ＲＤＧＴ比二维ＣＤＧＴ更适合于图像的压缩。

In this paper, we introduced the fast 2D real-valued discrete Gabor transform(RDGT) proposed in our previous work, and the relationship between the coefficients the 2D RDGT and the 2D complex-valued discrete Gabor transform (CDGT).To obtain a numeral comparison of the degree of image data decorrelation,we also performed an entropy computation of both transform coefficients under various effective window width conditions. The comparison results point out that the 2D RDGT is more suitable for image compression than the 2D CDGT.