Imaging Vector Fields Using Line Integral Convolution

I found this paper to be too mathematical, and a bit hard to understand, given that I have no prior knowledge of this field and application of vector fields. From what I grasp, I think the paper suggests that this approach of using a LIC approach to imaging vector fields is quite a generic and useful method for 2D and 3D vector fields. Though the maths part is a bit complicated, I think the algorithm takes one image and a vector field as input, and gives an output image by using the vector fields to generate a number of filters to process the image. The author also gives the advantages of this method over other currently available techniques like spatial resolution and DDA convolution which suffer from drawbacks like loss of control and accuracy.