An Enhanced Woelfel Image Noise Filter

The purpose of this research was to investigate image processing methods in MATLAB by using FIR filters in two dimensions. The basic image signal was a BMP scan of the brain. Each pixel of the underlying image was given a random amount of noise. Two FIR filters were put to the test in order to compare and contrast their performance in terms of SNR enhancement and image quality output. The PSD of the uniquely filtered image did not ring, but it did enable a little amount of information to pass through in a uniform shape from the edges of the PSD. This PSD was considerably closer to the original image than the lowpass filtered PSD. In every regard, the novel filter outperformed the ideal lowpass filter. The SNR was improved, and more information was recovered from the original BMP scan. This unique filter design was more complicated and required more tuning to ensure a good performance for this image, but the lowpass filter was not precise enough to remove the noise without also blurring the information.