Fast Interactive Noise Removal

Summary:

This project for PSYCH 267, was an implementation of an image noise removal algorithm, that used both frequency and time-domain information [1].

The goal is to be able to quickly remove small unnatural defects to images (such as a scratch, or some localized blur). In addition, the interactivity is to be minimal; in this case it is simply specifying sample (not-distorted) image regions to be used for reference when correcting defective subimages.

Reference:

[1] Hirani, A.N., and Totsuka, T., "Combining Frequency and Spatial Domain Information for Fast Interactive Image Noise Removal", Proc. Siggraph 96, pp. 269 - 276.