برآورد سطح زیرکشت محصولات کشاورزی با استفاده از تصاویر ماهواره لندست ۸ (مطالعه موردی: شهرستان شوشتر)

Satellite images have a high capability for estimating the area under agricultural crops. The aim of this study was to identify the area under dominant crops such as in Shushtar Province using Landsat ۸ satellite images during the growing season during ۲۰۱۹. With Maximum Probability technique and Support Vector Machine in the first approach and using NDVI index in the second approach, crops in different growing seasons and according to their calendar, a cropping pattern map was drawn. In order to evaluate the accuracy of the results, the generated maps with reference data were examined. Agricultural Jihad statistics of Khuzestan were also used. The results showed that Kappa coefficient and overall accuracy were calculated as ۹۰% and ۸۰% in the Maximum Probability technique, ۹۲% and ۹۰% in the Support Vector Machine and ۹۵% and ۹۳% in the NDVI, respectively. Based on the results, the cultivation area of wheat, barley, rice, and corn, in the Maximum Probability technique, in comparison with the statistics of Agricultural Jihad, had an error of ۱۲.۶, ۱۶.۴, ۸.۷ and ۶.۶%, respectively and in the Support Vector Machine had an error of ۱۰.۱, ۸.۳, ۵.۱ and ۷.۲%, respectively. However, using the NDVI index as the best approach for estimating the cultivation area in this region, in comparison with the statistics of Agricultural Jihad, has an error of ۲.۴, ۱.۵, ۴.۳ and ۴.۶%, respectively, which indicates the high capability of vegetation indices to estimate the Cultivation Area, According to their phenological stage.

کلیدواژه ها:

الکوی کشت ، تصاویر ماهواره ای ، طبقه بندی ، NDVI ، شوشتر