Crop Yield Estimation Using Phenological Parameters

Abstract

The assessment of harvest yield is finished utilizing the information got from satellite pictures and the verifiable information about the yield. At first the satellite pictures of the yields are gotten by choosing the time and date of the pass of satellite on the investigation territory and afterward getting the pictures from satellite of harvest in the examination region this should be possible utilizing satellite like LANDSAT, SENTINEL, IKONOS, and so forth The satellite pictures should be preprocessed to eliminate commotion and undesirable regions. With the assistance of GPS (Global Positioning System) the preparation tests or Ground Truth Points are chosen which is needed for Supervised Classification. To characterize the satellite picture the product utilized is GIS (Geographical Information System). QGIS is a high level variant of GIS device which is utilized to chip away at satellite pictures. The satellite pictures are broken down utilizing QGIS and the information, like NDVI (Normalized Difference Vegetation Index), Rainfall, Temperature, and so on, are removed or estimated from satellite pictures. Python is utilized to deal with the information got from the satellite pictures and furthermore on the information of harvest yield from bygone eras to eliminate any inconsistencies or excess information. Presently, to anticipate the yield of harvest for quite a long time 2020 and 2021the Machine Learning Algorithm of Regression Modeling is executed on both the information from satellite pictures and the verifiable yield of yield in the investigation region. The anticipated yield of both satellite information and recorded information is looked at and the general exactness is found.