# Types of Restoration Filters

**Restoration Filters** are the type of filters that are used for operation of noisy image and estimating the clean and original image. It may consists of processes that are used for blurring or the reverse processes that are used for inverse of blur. Filter used in restoration is different from the filter used in enhancement process.

**Types of Restoration Filters:**

There are three types of Restoration Filters: Inverse Filter, Pseudo Inverse Filter, and Wiener Filter. These are explained as following below.

**1. Inverse Filter:**

Inverse Filtering is the process of receiving the input of a system from its output. It is the simplest approach to restore the original image once the degradation function is known.

It can be define as:

H'(u, v) = 1 / H(u, v) Let, F'(u, v) -> Fourier transform of the restored image G(u, v) -> Fourier transform of the degraded image H(u, v) -> Estimated or derived or known degradation function thenF'(u, v) = G(u, v)/H(u, v)where, G(u, v) = F(u, v).H(u, v) + N(u, v) and F'(u, v) = f(u, v) - N(u, v)/H(u, v)

**Note:** Inverse filtering is not regularly used in its original form.

**2. Pseudo Inverse Filter:**

Pseudo inverse filter is the modified version of the inverse filter and stabilized inverse filter. Pseudo inverse filtering gives more better result than inverse filtering but both inverse and pseudo inverse are sensitive to noise.

Pseudo inverse filtering is defined as:

H'(u, v) = 1/H(u, v), H(u, v)!=0 H'(u, v) = 0, otherwise

**3. Wiener Filter:**

(Minimum Mean Square Error Filter). Wiener filter executes and optimal trade off between filtering and noise smoothing. IT removes the addition noise and inputs in the blurring simultaneously. Weiner filter is real and even.

It minimizes the overall mean square error by:

e^2 = F{(f-f')^2} where, f -> original image f' -> restored image E{.} -> mean value of arguments H(u, v) = H'(u, v)/(|H(u, v)|^2 + (Sn(u, v)/Sf(u, v)) where H(u, v) -> Transform of degradation function Sn(u, v) -> Power spectrum of the noise Sf(u, v) -> Power spectrum of the undergraded original image

**No blur only additive noise:**

H(u, v)=1 W(u, v) = 1 / (1 + Sn(u, v)/Sf(u, v)) W(u, v) = SNR/(1 + SNR) where, SNR = Sf(u, v)/Sn(u, v)

**No noise only blur:**

Sn(u, v)=0 W(u, v) = 1/H(u, v)

**Drawbacks of Restoration Filters:**

- Not effective when images are restored fro the human eye.
- Cannot handle the common cause of non-stationary signals and noise.
- Cannot handle spatially variant blurring point spread function.

## Recommended Posts:

- Frequency Domain Filters and its Types
- Spatial Filtering and its Types
- Vector instruction types
- Types of Models in Object Oriented Modeling and Design
- Introduction of HIP parallel programming language
- C program for Time Complexity plot of Bubble, Insertion and Selection Sort using Gnuplot
- Difference between Opening and Closing in Digital Image Processing
- Node.js | GM swirl() Function
- Node.js | GM rotate() Function
- Nodejs | GM blur() Function
- Nodejs | GM sharpen() Function
- Node.js | GM stroke() Function
- Node.js | GM shade() Function
- Node.js | GM trim() Function

If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.