Application of modified conjugate directional filtering for remote sensing image processing

In image processing, enhancement of conjugate linear features on an image is required sometimes. This enhancement can be achieved by using two directional filters in two specific conjugate directions. Traditionally the filtered data from the two conjugate directions are presented in two separate images. This is inconvenient for making a direct comparison between linear features in the two conjugate directions. We recently proposed a new method named the modified conjugate directional filtering (MCDF) to combine two direction-filtered results in two conjugate directions into one image that retains the maximum information of the linear features in the two conjugate directions. Since conjugated features are often seen on remote sensing images, in this paper we present a few cases in using the MCDF to enhance the conjugated features on some remote sensing images. These examples show that the MCDF is a useful tool for linear feature enhancement not only in the two selected directions, but in other directions as well.