When the XIL library uses bilinear (or first-order) interpolation to determine the value of an image at noninteger coordinates, it calculates the value by looking at the values of the four pixels surrounding the point of interest and then using a bilinear equation. This type of interpolation yields better results than nearest neighbor interpolation, but can itself have an undesirable smoothing effect on an image. To alleviate this problem, you can use bicubic interpolation.
Process which extracts the best value for a result pixel based on a weighted average of the four pixels surrounding the corresponding source pixel.
Two-dimensional linear interpolation of pixel values based on the four pixels in a 2 x 2 pixel neighborhood. See also bicubic interpolation, nearest-neighbor interpolation.