Document Type : Research Paper
Authors
1 MSc, Agricultural and Natural Resources Research Center, Esfahan, Iran,
2 Scientific Board, Najaf abad Unit, Islamic Azad University, Iran,
3 Scientific Board, Agricultural and Natural Resources Research Center, Esfahan, Iran
Abstract
Nowadays, application of remote sensing is common in soil conservation and erosion projects and most of them are used aerial photogarphs and desipte of the numerous advantages, there are some limitations. In this study, replacemen of moderate-resolution satellite data instead of aerial photographs has been investigated with emphasis on processing of satellite images and using image filters and the purpose of soil erosion figures detection. Finaly, image processing of area are prepared with maximum capability vitual interpretation of effects associated with erosion, after the study design, atmospheric and geometric corrections and perform pre-processing and initial processing on ETM+ and ASTER images. In order to evaluate the efficiency of filters for detection and identification of soil erosion were used low-pass and high-pass filters such as filters, median, mean, High Pass, Gaussian, Laplacian, Sobel and linear filters on ASTER and ETM images in the part of the Esfahan Dastkan basin. According to the results, using of low pass filters are not recommended because of the resolution reduction of soil erosion. High-Pass filters can be partialy appeard the edages of image and can be help to visual interpretation (especially in combination with the original image). Use of filters is recommended in combination with the image and as an aid to vitual interpretation, due to problems of application the fiters for separate soil erosion units. Regard to the dimension of soil erosion (rill and gully) and compared to the spatial resolution of the used sensor in this study, the erosions are not clearly appeared in the picture and application various image filters to automatically discover and exploit of these effects can not be efficiently. Resolution images of ASTER and ETM+ are not responsive for mapping erosion by using automated method and should be looked for a solution of semi-automated and visual interpretation. Also, because of the similarity between the spectral and geometrical picture was not found a major difference in the results. Therefore, it is recommended to produce a susceptibility map of erosion or erosion risk map instead the variouse types of regional soil erosion.
Keywords
)