Integration of Landscape Metrics and Object-Oriented Remote Sensing in order to determine the Crop Type and Arrangement of Agricultural Land

This study has been carried out to identify the crop type and spatial pattern of agricultural lands in the Segzi Hydrological Unit in Isfahan Province, Iran.

 Considering the vegetative calendar and phonological cycles of the major crops in the study area including wheat, alfalfa, fruit trees and vegetables, as well as agricultural land size, it was used for the study 3 Landsat satellite images (OLI) of the year 2015. After corrections and preliminary preprocesses were used from the NDVI and multi-resolution segmentation algorithm, taking into account three criteria of color, scale and shape, to determine the agricultural land area. Then, with the development of a decision tree based on the NDVI index, major crops were identified, mapped and evaluated for their accuracy. Then, using landscape metrics including number of patch (NP), mean patch size (MPS), Mean Shape Index (MSI), Perimeter-area ratio (PARA) and Mean Nearest Neighbor (MNN) to study the structure and agricultural arrangement.

The results of this study showed that a large area of agricultural land (about 46%) in the region is dedicated to wheat cultivation and less than 8% to vegetable cultivation. The results also showed that all lands in the region have a regular geometric shape with a minimum amount of area per area.

 The output of this study shows the movement of agricultural land in the region towards a monoculture. Also, the lack of water resources in recent years has formed a picture of a fragmented land.