US 2018/0018517 A1Published application

Generating pixel maps from non-image data and difference metrics for pixel maps

AssigneeClimate LLC

InventorsHao Zhong, Ying Xu

Life of this patent

Estimating the term…

Loading patent…

Abstract

Systems and methods for scalable comparisons between two pixel maps are provided. In an embodiment, an agricultural intelligence computer system generates pixel maps from non-image data by transforming a plurality of values and location values into pixel values and pixel locations. The non-image data may include data relating to a particular agricultural field, such as nutrient content in the soil, pH values, soil moisture, elevation, temperature, and/or measured crop yields. The agricultural intelligence computer system converts each pixel map into a vector of values. The agricultural intelligence computer system also generates a matrix of metric coefficients where each value in the matrix of metric coefficients is computed using a spatial distance between to pixel locations in one of the pixel maps. Using the vectors of values and the matrix of metric coefficients, the agricultural intelligence computer system generates a difference metric identifying a difference between the two pixel maps. In an embodiment, the difference metric is normalized so that the difference metric is scalable to pixel maps of different sizes. The difference metric may then be used to select particular images that best match a measured yield, identify relationships between field values and measured crop yields, identify and/or select management zones, investigate management practices, and/or strengthen agronomic models of predicted yield.

Claims21

Dependent claims

A computing device comprising:

a memory;
one or more processors communicatively coupled to the memory;
one or more instructions stored in the memory, executed by the one or more processors, and configured to cause the one or more processors to perform:
obtaining a first pixel map for a predicted agronomic yield of a particular field wherein each pixel of the first pixel map represents an agronomic yield of a crop at a physical location within the particular field;
obtaining a second pixel map for a measured agronomic yield of the particular field wherein each pixel of the second pixel map represents an agronomic yield of a crop at a physical location within the particular field;
generating, from the first pixel map, a first vector of values;
generating, from the second pixel map, a second vector of values;
generating a matrix of metric coefficients wherein each row of the matrix of metric coefficients corresponds to a first pixel location of the first pixel map, each column of the matrix of metric coefficients corresponds to a second pixel location of the first pixel map, and each value in the matrix of metric coefficients is computed using a spatial distance between the first pixel location and the second pixel location for the value;
computing a particular difference metric that specifies a difference between the first pixel map and the second pixel map based on the first vector of values, the second vector of values, and the matrix of metric coefficients;
computing, from the particular difference metric, a particular normalized difference metric comprising a quotient of the difference metric with a sum of the metric coefficients in the matrix of metric coefficients;
generating and sending, to a field manager computing device, one or more field management recommendations based, at least in part, on the particular normalized difference metric.

8 more elements

A method comprising:

obtaining a first pixel map for a first physical property at a plurality of locations in a particular region;
obtaining a second pixel map for a second physical property at the plurality of locations in a particular region, wherein the second pixel map is an equal size as the first pixel map;
generating, from the first pixel map, a first vector of values;
generating, from the second pixel map, a second vector of values;
generating a matrix of metric coefficients wherein each row of the matrix of metric coefficients corresponds to a first pixel location of the first pixel map and each column of the matrix of metric coefficients corresponds to a second pixel location of the first pixel map and each value in the matrix of metric coefficients is computed using a spatial distance between the first pixel location and the second pixel location for the value;
computing a difference metric identifying a difference between the first pixel map and the second pixel map based on the first vector of values, the second vector of values, and the matrix of metric coefficients.

3 more elements

A method comprising:

receiving first digital data comprising yield values for a plurality of locations;
receiving a particular image comprising a plurality of image values for the plurality of locations;
identifying a first spatial resolution of the yield values;
identifying a second spatial resolution of the image values of the particular image;
aggregating the yield values and the image values of the particular image to a common grid based, at least in part, on the first spatial resolution and the second spatial resolution;
generating a first empirical cumulative distribution function of the yield values;
computing a first quantile transformation for the first empirical cumulative distribution function;
transforming each yield value to a first normal distribution using the first quantile transformation;
generating a first pixel map using the transformed yield values;
generating a second empirical cumulative distribution function of the image values of the particular image;
computing a second quantile transformation for the second empirical cumulative distribution function;
transforming each image value of the particular image to a normal distribution using the second quantile transformation;
generating a second pixel map using the transformed image values of the particular image;
computing a particular difference metric indicating a difference between the yield values and the particular image using the first pixel map and the second pixel map.

11 more elements

Dependencies are inferred from the claim text.

Facts

Application: 15/209,628Filed 13 Jul 2016
Published: 18 Jan 2018
Term ends: Not estimatedThe source records don’t support an estimate.

Classification7 codes

G06V 20/188
Terrestrial scenes: vegetation
G06F 18/22
Analysing: matching criteria, e.g. proximity measures
G06V 10/761
Image or video pattern matching; Proximity measures in feature spaces: proximity, similarity or dissimilarity measures
G06F 18/217
Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation: validation; Performance…