median filter menu

Median Filtering is a method for removing noise from an image. Image noise generally looks like someone sprinkled salt and pepper over an image. Median filtering lets you remove such noise, without damaging other image features.

The Median Filter works as follows:

For each pixel in the image, the system defines a neighborhood of that pixel. A 3 by 3 neighborhood, for example, would include the pixel and its eight immediate neighbors.
The system calculates the median intensity value among the pixels in the neighborhood. The median value is the one that splits the population in half, so that half the pixels in the neighborhood have a larger intensity value, and half are smaller.
The pixel's intensity is replaced by the median value.

The procedure for running the median filter is:

You decide how big a neighborhood you want to use. A 3 by 3 neighborhood is typical. Indicate the dimensions of the neighborhood on the menu. The first number is the width of the neighborhood in pixels, and the second is the height. The larger the neighborhood, the longer it takes.
You click on the FILTER menu field. This starts the filtering procedure.
To see the result image, click on the DISPLAY menu field.

You can restrict the filter to work on only a portion of the image by specifying an Area of Interest (AOI). The AOI can be either rectangular or polygonal. To specify a rectangular AOI, first display the image, then click on USE RECTANGULAR AREA. The system will prompt you to indicate the rectangular area on the image. Do so with two clicks. When you then apply the filter, only the pixels in the rectangular area will be affected. Similarly, you can make an irregular AOI by first clicking on USE POLYGONAL AREA and then outlining the AOI with a series of clicks on mouse button 1. Terminate the drawing of the AOI by clicking with button 2. If you have restricted processing to an AOI and you want to make it apply to the entire image again, click on USE ENTIRE IMAGE.