Estimation of Agricultural Crop Yield Using Sentinel-2 Images (Case Study: Zanjan City)

The basis for proper planning and management is to have accurate and timely statistics and information. One of the most important statistics of the agricultural sector is the annual production rate of each crop, which also depends on the area under cultivation of crop and its efficiency. One of the tools that can calculate the area under cultivation in the least time, with low cost and with high accuracy is remote sensing science and technology. In this study, two classification methods including artificial neural networks and support vector machine with different kernels are used and the area under cultivation of major crops in the region consisting of 8 classes is estimated. According to the results, the overall accuracy of the artificial neural networks and support vector machine was 97.74% and 97.96% with a kappa coefficient of 0.97 for both methods, indicating that both methods are good for separation and classification of agricultural lands in the area. Based on the overall accuracy, it can be concluded that the two methods of classification have almost the same results in the region. Also, based on the results of the user's accuracy for the four crops including Alfalfa, Rice, Onion and Melon, the support vector machine method performs better than the neural network method and also for dry and water Wheat, Sorghum, and Pasture, the neural network method out performs the support vector machine in the region.