Generating agronomic inferences from extracted individual plant components
Implementations are disclosed for analyzing extracted individual plant components to make agronomic inferences about entire composite plant organs from which the individual plant components were harvested, and for using those agronomic inferences for various purposes. In various implementations, individual plant component(s) may be sampled from multiple plant components removed from composite plant organ(s) that previously included the plant components. Digital image(s) may be captured of the sampled individual plant component(s) and processed based on machine learning model(s) to generate agronomic inference(s) about the composite plant organ(s) that previously included the plurality of plant components. Based on the agronomic inference(s), computing device(s) may render output that includes a diagnosis or recommendation for the grower about the field or the crops, and/or agricultural equipment (e.g., robots) may be operated automatically.
1 . A method implemented using one or more processors and comprising:
sampling a plurality of individual plant components from a plurality of plant components removed from one or more composite plant organs that previously included the plurality of plant components, wherein the one or more composite plant organs are part of crops in a field managed by a grower;
capturing one or more digital images of the sampled one or more plurality of individual plant components;
processing the one or more digital images based on one or more machine learning models, wherein one or more of the machine learning models was trained previously based on training data associating a plurality of reference individual plant components with ground truth agronomic observations about one or more reference composite plant organs that yielded the plurality of reference individual plant components, wherein the processing includes:
generating a plurality of individual plant component embeddings that represent the sampled plurality of individual plant components within the one or more digital images;
determining similarity measures between the plurality of individual plant component embeddings and individual plant component reference embeddings generated previously based on the plurality of reference individual plant components; and
based on the similarity measures and at least some of the ground truth agronomic observations about one or more of the reference composite plant organs that yielded the plurality of reference individual plant components, generating one or more agronomic inferences about the one or more composite plant organs that previously included the plurality of plant components; and
based on the one or more agronomic inferences, causing one or more computing devices to render output that includes a diagnosis or recommendation for the grower about the field or the crops.
2 . The method of claim 1 , wherein the output comprises includes subfield recommendations for the field managed by the grower.
3 . The method of claim 1 , wherein the output includes one or more mid-crop-cycle agronomic recommendations.
4 . The method of claim 1 , wherein the output includes a local environmental zone map for the field.
5 . The method of claim 1 , wherein the plurality of individual plant components are corn kernels, and the one or more composite plant organs are ears of corn.
6 . The method of claim 1 , wherein the plurality of individual plant components are wheat seeds, and the one or more composite plant organs are heads of wheat.
7 . The method of claim 1 , wherein the plurality of individual plant components are sampled by a component onboard a combine harvester operating in the field.
8 . The method of claim 1 , wherein the processing includes segmenting the sampled plurality of individual plant components within the one or more digital images.
9 . The method of claim 1 , wherein the similarity measures are first similarity measures, and further including:
determining second similarity measures between the plurality of individual plant component embeddings; and
based on the second similarity measures, clustering the plurality of individual plant component embeddings into clusters, with each cluster representing an archetype composite plant organ.
10 . A system comprising one or more processors and memory storing instructions that, in response to execution by the one or more processors, cause the one or more processors to:
sample a plurality of individual plant components from a plurality of plant components removed from one or more composite plant organs that previously included the plurality of plant components, wherein the one or more composite plant organs are part of crops in a field managed by a grower;
capture one or more digital images of the sampled one or more plurality of individual plant components;
process the one or more digital images based on machine learning models, wherein one or more of the machine learning models was trained previously based on training data associating a plurality of reference individual plant components with ground truth agronomic observations about one or more reference composite plant organs that yielded the plurality of reference individual plant components, wherein to process includes to:
generate a plurality of individual plant component embeddings that represent the sampled plurality of individual plant components within the one or more digital images;
determine similarity measures between the plurality of individual plant component embeddings and individual plant component reference embeddings generated previously based on the plurality of reference individual plant components; and
based on the similarity measures and at least some of the ground truth agronomic observations about one or more of the reference composite plant organs that yielded the plurality of reference individual plant components, generate one or more agronomic inferences about the one or more composite plant organs that previously included the plurality of plant components; and
based on the one or more agronomic inferences, cause one or more computing devices to render output that includes a diagnosis or recommendation for the grower about the field or the crops.
11 . The system of claim 10 , wherein the output includes subfield recommendations for the field managed by the grower, one or more mid-crop-cycle agronomic recommendations, or a local environmental zone map for the field.
12 . The system of claim 10 , wherein the plurality of individual plant components are corn kernels, and the one or more composite plant organs are ears of corn.
13 . The system of claim 10 , wherein the plurality of individual plant components are wheat seeds, and the one or more composite plant organs are heads of wheat.
14 . The system of claim 10 , wherein the plurality of individual plant components are sampled by a component onboard a combine harvester operating in the field.
15 . The system of claim 10 , wherein the processing includes segmenting the sampled plurality of individual plant components within the one or more digital images.
16 . At least one non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to:
obtain one or more digital images of a plurality of individual plant components that were sampled from a plurality of plant components removed from one or more composite plant organs that previously included the plurality of plant components, wherein the one or more composite plant organs are part of crops in a field managed by a grower;
process the one or more digital images based on one or more machine learning models, wherein one or more of the machine learning models was trained previously based on training data associating a plurality of reference individual plant components with ground truth agronomic observations about one or more reference composite plant organs that yielded the plurality of reference individual plant components, wherein to process includes to:
generate a plurality of individual plant component embeddings that represent the sampled plurality of individual plant components within the one or more digital images:
determine similarity measures between the plurality of individual plant component embeddings and individual plant component reference embeddings generated previously based on the plurality of reference individual plant components; and
based on the similarity measures and at least some of the ground truth agronomic observations about one or more of the reference composite plant organs that yielded the plurality of reference individual plant components, generate one or more agronomic inferences about the one or more composite plant organs that previously included the plurality of plant components; and
based on the one or more agronomic inferences, cause one or more computing devices to render output that includes a diagnosis or recommendation for the grower about the field or the crops.
17 . The non-transitory computer-readable medium of claim 16 , wherein the output includes subfield recommendations for the field managed by the grower, one or more mid-crop-cycle agronomic recommendations, or a local environmental zone map for the field.
18 . The non-transitory computer-readable medium of claim 16 , wherein the plurality of individual plant components are corn kernels, and the one or more composite plant organs are ears of corn.
19 . The non-transitory computer-readable medium of claim 16 , wherein the plurality of individual plant components are wheat seeds, and the one or more composite plant organs are heads of wheat.
20 . The non-transitory computer-readable medium of claim 16 , wherein the plurality of individual plant components are sampled by a component onboard a combine harvester operating in the field.