Multiscale Decomposition for Local Edge-Preservation using Median Filter as Preprocessor to LEP Filter

Edge detection is mainly used in image processing, and is used for feature selection and feature extraction. Edge preservation is important in filtering design to avoid the halo artifacts. Edge preserving decomposition of an image can be achieved using a Local Edge Preserving (LEP) filter. This technique is suitable for High Dynamic Range (HDR) images. Here, multi scale decomposition process is used, in which the image is divided into base layer and detail layer. The output of LEP provides better visual quality to the images. The problem of LEP filter is that it is not capable of eliminating the noise at the output completely and the smoothening performance of the LEP filter is poor compared with the previously established filters. Therefore a median filter is used as the preprocessor. The median filter has the ability to preserve the edges, and the capability to remove the noise present at the input. Thus, the LEP filter performance gets improved by using the median filter as the preprocessor.