Estimating soil properties within a field using hyperspectral remote sensing

Patent metadata

Field	Value
Publication number	US-11714211-B2
Application number	US-202117243447-A
Country	US
Kind code	B2
Filing date	Apr 28, 2021
Priority date	Sep 12, 2014
Publication date	Aug 1, 2023
Grant date	Aug 1, 2023

Abstract

A method for building and using soil models that determine soil properties from soil spectrum data is provided. In an embodiment, building soil model may be accomplished using soil spectrum data received via hyperspectral sensors from a land unit. A processor updates the soil spectrum data by removing interference signals from the soil spectrum data. Multiple ground sampling locations within the land unit are then determined based on the updated soil spectrum data. Soil property data are obtained from ground sampling at the ground sampling locations. Soil models that correlate the updated soil spectrum data with the soil property data are created based on the updated soil spectrum data and the soil property data. The soil models are sent to a storage for future use.

First claim

What is claimed is:

1. A computer-implemented method for building and using soil models that determine soil properties from soil spectrum data, the method comprising:
selecting, by a processor of a server computer system, a subset of a set of spectral bands of remotely sensed, soil spectrum data for an agricultural field;
generating, by the processor, based on, at least in part, the subset of the set of spectral bands and soil property data obtained for the agricultural field, multiple soil models that correlate the subset of the set of spectral bands with the soil property data;
selecting, by the processor, one or more soil models, from the multiple soil models, based on cross validation of the one or more soil models satisfying a defined quality threshold;
based on the selected one or more soil models, generating one or more scripts for controlling one or more operating parameters of one or more agricultural machines; and
transmitting the one or more scripts to the one or more agricultural machines to cause the one or more agricultural machines to use the one or more scripts to control the one or more operating parameters of the one or more agricultural machines to allow the one or more agricultural machines to perform agricultural operations at one or more location of the agricultural field.

2. The computer-implemented method of claim 1 , further comprising receiving, by the processor, the remotely sensed, soil spectrum data from sensors affixed to movable equipment; and
wherein the soil spectrum data are associated with different locations within the agricultural field.

3. The computer-implemented method of claim 2 , further comprising:
prior to generating the multiple soil models, removing interference signals from the soil spectrum data, using a preprocessing technique selected from a plurality of preprocessing techniques, based on the soil spectrum data, a type of the sensors used to collect the soil spectrum data, or elevation at which the soil spectrum data are received.

4. The computer-implemented method of claim 3 , wherein the plurality of preprocessing techniques comprises: data smoothing, spectral derivatives, standard normal variate preprocessing, or absorbance.

5. The computer-implemented method of claim 1 , further comprising:
determining one or more ground sampling locations within the agricultural field based on the soil spectrum data; and
using spatial sampling techniques on the soil spectrum data.

6. The computer-implemented method of claim 5 , wherein the spatial sampling techniques include a conditional Latin Hypercube Sampling.

7. The computer-implemented method of claim 1 , wherein generating multiple soil models comprises: discovering and calibrating latent variables during a multivariate regression analysis.

8. The computer-implemented method of claim 7 , wherein the discovering and calibrating latent variables comprises: using one or more signature spectral band determination techniques of a partial least-square regression algorithm, a random forest algorithm, principal component regression, partial least squares, ridge regression, lasso regression, or decision tree statistical procedures.

9. The computer-implemented method of claim 1 , further comprising: predicting soil properties for a specific geo-location within the agricultural field using the one or more soil models.

10. The computer-implemented method of claim 1 , further comprising: determining optimal locations for planting, nutrient applications, scouting, or implementing sentinel seed technology using the one or more soil models.

11. The computer-implemented method of claim 10 , further comprising: causing displaying the optimal locations in a soil map.

12. One or more non-transitory storage media storing instructions which, when executed by one or more computing devices, cause performance, by the one or more computing devices, of:
selecting a subset of a set of spectral bands of remotely sensed, soil spectrum data for an agricultural field;
generating, based on, at least in part, the subset of the set of spectral bands and soil property data obtained for the agricultural field, multiple soil models that correlate the subset of the set of spectral bands with the soil property data;
selecting one or more soil models, from the multiple soil models, based on cross validation of the one or more soil models satisfying a defined quality threshold;
based on the selected one or more soil models, generating one or more scripts for controlling one or more operating parameters of one or more agricultural machines; and
transmitting the one or more scripts to the one or more agricultural machines to cause the one or more agricultural machines to use the one or more scripts to control the one or more operating parameters of the one or more agricultural machines to allow the one or more agricultural machines to perform agricultural operations at one or more location of the agricultural field.

13. The one or more non-transitory storage media of claim 12 , wherein the instructions, when executed by the one or more computing devices, cause performance, by the one or more computing devices, of receiving the remotely sensed, soil spectrum data from sensors that are affixed to movable equipment; and
wherein the soil spectrum data are associated with different locations within the agricultural field.

14. The one or more non-transitory storage media of claim 13 , wherein the instructions, when executed by the one or more computing devices, further cause performance, by the one or more computing devices, of removing interference signals from the soil spectrum data through a preprocessing technique of a plurality of preprocessing techniques, based on the soil spectrum data, a type of the sensor used to collect the soil spectrum data, or elevation at which the soil spectrum data are received.

15. The one or more non-transitory storage media of claim 14 , wherein the plurality of preprocessing techniques comprises: data smoothing, spectral derivatives, standard normal variate preprocessing, or absorbance.

16. The one or more non-transitory storage media of claim 12 , wherein the instructions, when executed by the one or more computing devices, further cause performance, by the one or more computing devices, of determining ground sampling locations within the agricultural field based on updated soil spectrum data; and using spatial sampling techniques on the updated soil spectrum data.

17. The one or more non-transitory storage media of claim 16 , wherein the spatial sampling techniques include conditional Latin Hypercube Sampling.

18. The one or more non-transitory storage media of claim 12 , wherein the instructions, when executed by the one or more computing devices, cause performance, by the one or more computing devices, in connection with generating the multiple models, of discovering and calibrating latent variables during a multivariate regression analysis.

19. The one or more non-transitory storage media of claim 18 , wherein the instructions, when executed by the one or more computing devices, cause performance, by the one or more computing devices, in connection with discovering and calibrating latent variables, of using one or more signature spectral band determination techniques of a partial least-square regression algorithm, a random forest algorithm, principal component regression, partial least squares, ridge regression, lasso regression, or decision tree statistical procedures.

Assignees

• Climate Llc

Inventors

• Xiang Haitao
• Yang Xianyuan
• Koshnick Nick
• Cisek Nick

Classifications

• G01V99/005 Primary

  Physics · mapped topic

• A01B76/00

  Parts, details or accessories of agricultural machines or implements, not provided for in groups A01B51/00 - A01B75/00 · CPC title

• G01N21/31

  Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry {(G01N21/72 takes precedence)} · CPC title

• G01N21/359

  using near infrared light · CPC title

• G01N33/0098

  Plants or trees (wood G01N33/46) · CPC title