Deep Learning framework for crop identification
Government agencies, FPOs and industry stakeholders require information on the spatial distribution and area of cultivated crops for planning purposes. Stakeholders can plan the import and export of food products based on such information. Although some ministries of agriculture and food security annually commission their staff to map different crop types, these ground surveys are expensive and yet cover only a sample of farms. Remote sensing data when used with deep learning, enable the determination of the spatial distribution of crops at varying spatial scales with relatively lesser financial resources. Open datasets from NASA and Sentinel are very useful in this regard.
GEO SPATIAL TECHNOLOGIES FOR CROP IDENTIFICATION
We have taken up the challenge of crop identification and to analyze the difference between paddy and sugarcane. So as to get accurate and faster estimation of crop area which is very essential for area-based subsidies and helps in deciding agriculture policies .The outcomes of the analysis made the team understand various parameters to identify crop.
The android application “kriti demo” is given to us. which collects the GPS coordinates of specific field crop.LANDSAT-8(OLI) image of Anakapalli (March, April, and May 2018) is taken.
NDVI & MSAVI – indexes indicating vegetation with in a threshold values using raster calculater.
We apply threshold value to identify the chlorophyll content. For example , the ndvi vale is 0.09479 to 0.32475 then the range given to apply threshold is >0.18.It converts in to binary as 0 and 1.
0 = black (condition unsatisfied); 1 = white(condition satisfied).
The graph showing the variations in range of MSAVI and NDVI for paddy along three months march, April and May.The paddy is cultivated in Rabi and Kharif seasons. Rabi crop is harvested in march-may and sowed in Oct-nov. Kharif crop is harvested in sept-oct and sowed in june-july.According to the graph,the values of Msavi and NDVI are maximum in march and april decrease in may.It shows that as the paddy ripening the chlorophyll content decreases.
The graph showing the variations in range of MSAVI and NDVI for sugarcane along three months march,april and may. Sugarcane is a commercial crop.Sugarcane Seed is grown from march/april/may/june to December.According to the graph the values of MSAVI and NDVI are minimum in march and gradually increase towards may.It shows that the chlorophyll content increase.
This analysis is useful to get accurate and faster estimation of crop area which is very essential for area-based subsidies and helps in deciding agriculture policies. This can be useful to agricultural department.
Being a girl born from an agricultural family I am interested in agriculture related problems.So I have chosen to use spatial data for crop identification to improve accuracy and speed in estimation of crop area which is very essential for area-based subsidies in India.