What is bicubic interpolation in image processing

Comparison of different interpolation methods

The result of the step interpolation in 1% steps is particularly frightening (i.e. a bicubic interpolation is carried out and the image is enlarged by 1% per step). Obviously errors add up here (probably rounding errors). On the other hand, the step interpolations in 5% and 10% steps bring very good results - in my opinion the best. So it is particularly important here to choose the right step size. Steps that are too small primarily lead to an increase in image defects, while steps that are too large do not provide any advantages over a one-time bicubic interpolation.

The best results can be achieved with bicubic or step interpolation (based on bicubic interpolation), but we strongly recommend that you sharpen the result before printing. In the examples shown, the differences between Setp-5, Step-10 and bicubic are very small.

A comparison with other interpolation methods may not simply be possible because the example deliberately did not need to be sharpened. "Closed source" interpolations, which may use sharpening algorithms themselves, can therefore provide visually better results. But the results shown here can also be improved by resharpening.
When resharpening, however, it should be noted that there are no general rules for this, but the type and coordination of the sharpening algorithm must be selected depending on the output medium and size.

Archived page

This page has been archived, i.e. it is no longer actively maintained and the information may no longer be up-to-date.