Estimating nitrogen content using hyperspectral and multispectral images

A system comprising:

• one or more processors;
• one or more non-transitory computer-readable storage mediums storing one or more instructions which, when executed by the one or more processors, cause the one or more processors to perform steps comprising:
• receiving one or more band values for one or more pixels of one or more remote sensing images and, for each pixel of the one or more pixels, a corresponding nitrogen ground truth value of a nitrogen variable for the pixel;
• training a mapping function based on the one or more band values for the one or more pixels and, for each pixel of the one or more pixel, the corresponding nitrogen ground truth value, wherein the mapping function maps between a set of features related to the one or more band values of the one or more pixels and an estimated value of the nitrogen variable;
• a computer receiving a new remote sensing image, wherein the new remote sensing image specifies, for one or more pixel of the new remote sensing image, the one or more band values for the pixel of the new remote sensing image;
• generating, for each pixel of the one or more pixels of the new remote sensing image, a respective estimated value of the nitrogen variable for the pixel using the mapping function;
• sending, for each pixel of the one or more pixels, the respective estimated value of the nitrogen variable for the pixel to a field manager computing device;
• wherein the field manager computing device, in response to receiving respective estimated values of the nitrogen variable for one or more pixels, displays, on a computer display device, a map in which each pixel is assigned a color shaded based on the respective estimated value of the nitrogen variable generated for a pixel of the one or more pixels;
• wherein dark shadings in the map represent areas where leaf nitrogen content is high;
• wherein light shadings in the map represent areas where leaf nitrogen content is low.

One or more non-transitory computer-readable storage mediums storing one or more instructions which, when executed by one or more processors, cause the one or more processors to perform steps comprising:

• receiving one or more band values for one or more pixels of one or more remote sensing images and, for each pixel of the one or more pixels, a corresponding nitrogen ground truth value of a nitrogen variable for the pixel;
• training a mapping function based on the one or more band values for the one or more pixels and, for each pixel of the one or more pixel, the corresponding nitrogen ground truth value, wherein the mapping function maps between a set of features related to the one or more band values of the one or more pixels and an estimated value of the nitrogen variable;
• a computer receiving a new remote sensing image, wherein the new remote sensing image specifies, for one or more pixel of the new remote sensing image, the one or more band values for the pixel of the new remote sensing image;
• generating, for each pixel of the one or more pixels of the new remote sensing image, a respective estimated value of the nitrogen variable for the pixel using the mapping function;
• sending, for each pixel of the one or more pixels, the respective estimated value of the nitrogen variable for the pixel to a field manager computing device;
• wherein the field manager computing device, in response to receiving respective estimated values of the nitrogen variable for one or more pixels, displays, on a computer display device, a map in which each pixel is assigned a color shaded based on the respective estimated value of the nitrogen variable generated for a pixel of the one or more pixels;
• wherein dark shadings in the map represent areas where leaf nitrogen content is high;
• wherein light shadings in the map represent areas where leaf nitrogen content is low.