8.3.2 Smooth

The Smooth filter ("blurring") is one of the most applied filters for image transformations. It is applied for defocusing and smoothing the image (because of soothing the transitions between colours), decreasing the noises level (scanner rasters, dust, scratches, etc) and excessive detailing. Sharp borders in the image are soothed, and the image itself becomes blurred. Sometimes the filter is useful by decreasing the size of the image and the following compression.

The filter blurs the image with the use of Gaussian approaching (low frequencies filter). In that case, for all points of the image there happens a recounting (according to the Gaussian number distribution) of the values of primary colour components RGB taking into account the component values for the adjacent points. As the result of the Smooth transformation values of the colour components of the adjacent points are averaged, and the contrast between them decreases.

Gaussian smoothing is usually used before using the filters of finding the borders. The last ones are sensitive to the noise such as, for example, Laplace transformation. Using the Smooth filter in this case leads to the better result.

Filter parameters:

• Size.

The Size determines the radius of smoothing.