中文摘要：

提出一种基于指数构建的圆谐傅里叶矩——指数矩.在笛卡尔坐标系下直接计算指数矩,并利用指数矩进行图像的分解和重建.用指数矩能完整、无冗余的重建原图像,若选取有限数量的低阶指数矩与同级次的指数函数的乘积叠加求和,则可以近似的重建原图像.仿真实验结果表明,指数矩有很好的图像描述能力,使用10阶以上的指数矩就可以很好的重建原图像,而且计算速度比较快.

英文摘要：

One kind of moments based on the Circular Harmonic-Fourier,i.e.Moments Exponential-Fourier Moments,has been proposed.We calculated Exponential-Fourier Moments in the Cartesian coordinate system,and decomposed and rebuilt the images using the Exponential-Fourier Moments.The Exponential-Fourier Moments could reconstruct original image precisely.The approximate reconstructed original image could be made by multiplying the lower order of limited Exponential-Fourier Moments and the same class of Exponential-Fourier Moments.The result of simulation experiments proved that Exponential-Fourier Moments have a good ability for image description,even very little amount of independent Exponential-Fourier Moments can reconstruct the original image.