[vtkusers] about vtkEncodedGradientEstimator

[yhq at sjtu.edu.cn]

  can anyone tell me how vtkEncodedGradientEstimator works?
  I only find that I offer vtkEncodedGradientEstimator a vtkimagedata as input and vtkEncodedGradientEstimator produce
a output of encodednormals and then I use this encodenormals to look into the lookuptable of 
redspecular ,greenspecular and bluespecular ,reddiffuse,greendiffuse and bluediffuse,
  but I do not know clearly how this vtkEncodedGradientEstimator works,can you tell me ?