论文部分内容阅读
本文进一步研究了唯相位重建信号序列的方法。提出在将图象进行变形延拓处理后,应用唯相位重建算法进行图象数据压缩的方法。并证明,对于行场可分的一阶马尔可夫图象场,唯相位重建图象的方法用于图象数据压缩的性能可以做到略优于离散傅里叶变换压缩。从而修正了已有文献给出的结论,即认为就均方误差而言,唯相位重建法用于图象数据压缩所需的比特率远高于傅里叶变换压缩的比特率的结论。而且这种算法能有快速算法。结论还表明,倒余弦变换可作为正向变换用于图象数据压缩,而且(?)一种性能良好的正交变换。
This paper further studies the methods of reconstructing signal sequences with phase only. This paper proposes a method of image data compression using phase-only reconstruction algorithm after the image is transformed and extended. It is also proved that for the first-order Markov image field that can be divided into fields, the method of phase-only image reconstruction can be slightly better than the discrete Fourier transform compression for image data compression. Therefore, we can correct the conclusion of the existing literature that the bit-rate required for the phase-only reconstruction method for image data compression is much higher than that of the Fourier transform compression for the mean square error. And this algorithm can have a fast algorithm. The conclusion also shows that the inverse cosine transform can be used as forward transform for image data compression, and (?) A good orthogonal transform.