Application of satellite images for mapping land use in Ghezel Ozan watershed using fusion techniques and object-oriented processing

The optimal management of natural resources requires timely and accurate information. In this regard, land use plans are one of the most important sources of information on natural resource management. In fact, land use plans include the use of land for different needs. Today, digitization detection methods help to understand the panorama of a landscape using several satellite imagery. The purpose of this study was to classify landslide using satellite imagery and fusion techniques with object image processing approach in Ghezel Ozan watershed. To carry out this research, Landsat 8 satellite image, OLI / TIRS sensor was used. After making the necessary corrections in the preprocessing stage, we made the appearance or fusion of the pictorial image using a panchromatic band. We improved the spatial resolution of the image used from 30 meters to 15 meters. Then, with an object-oriented classification in the eCognition software environment, a map of land cover / land use and final results were presented. Also, the accuracy of the extracted accuracy (total accuracy of 89.41% and kappa coefficient of 0.87) indicate the accuracy of this classification method. Based on the results obtained in this study, it is suggested to use a technique for increasing the spatial resolution of images (fusion) as well as a precise object-oriented classification for the preparation of land-cover / land-use maps and the detection of changes from the image to the present day.