Archive \ Volume.7 2018 Issue 4

Classification of Satellite Imagery Data to Estimate Cultivated Areas of Grape Farms

Mahdi Saghari , Mehran Dizbadi , Majid Taghipour
Abstract

Agricultural sector is of great importance in economy due to climate variations and imbalanced population growth. Having accurate information of cultivated area can be helpful in formulation and implementation of agricultural programs. Analysis of satellite data by Remote Sensing techniques is a useful method for investigation and monitoring of the cultivated areas. Using high resolution imagery data can be an economical technique to delineate the cultivated area and to reveal temporal changes. The objective of this research is to estimate the area of cultivated farms of grape using Landsat 8 and Sentinel 2 for July 2016 in Shahroud County. We have initially taken 50 grape gardens for training samples to conduct different classification methods for each of the Landsat and Sentinel images. The results of the methods on the two images have been compared based on Kappa coefficient, accuracy values, and cultivated areas. Based on the results of the analysis, it can be concluded that just in Mahalanobis Distance classification method we obtained relatively the same value of 3600 hectare in both the images for the cultivated area. The value is also close to the value 3550 hectares measured directly by Jihad Agricultural Organization of Shahroud for the year 2016. Thus, the Mahalanobis can be the best classification method for measurement of cultivated areas of grape farms.


[PDF]