在哪里看出来是4曾分解

在 CurveLab 工具箱中，perform_curvelet_transform 函数的默认设置是四层分解。可以在函数文档中找到相应的说明：

[C, S] = perform_curvelet_transform(X, options);

Inputs:
   X: array of size (n,n) or (n,n,p) for 2D/3D arrays, n=2^J
   options: optional structure with the following fields:
      J: number of scales (default is log2(n))
      M: number of angles (default is 16)
      filter_type: 'pkva', 'pkvb', 'pkvc', 'pkvd', 'pkharm', 'pkdual', 'h', 'qshift', 'qshift_a', 'qshift_b' (default is 'pkva')
      filter_format: 'fourier', 'fourier_multiresolution', 'fourier_truncated' (default is 'fourier_multiresolution')
      filter_options: optional structure with additional options for the filter bank
      real: if non-zero, enforce real transform (default is 0)
      sparsify: if non-zero, threshold coefficients (default is 0)
      threshold: threshold value (default is 0)
Outputs:
   C: cell of Curvelet coefficients
   S: cell of Curvelet coefficients sizes

其中，J 表示分解的层数，默认值是图像大小的对数。因此，在不指定 J 参数的情况下，perform_curvelet_transform 函数默认使用四层分解。