Vision enhanced drones for precision farming

Patent metadata

Field	Value
Publication number	US-9898688-B2
Application number	US-201615169972-A
Country	US
Kind code	B2
Filing date	Jun 1, 2016
Priority date	Jun 1, 2016
Publication date	Feb 20, 2018
Grant date	Feb 20, 2018

Abstract

Methods, apparatuses and systems may provide for a neural network that analyzes and classifies agricultural conditions based on depth data and color data recorded by one or more drones, and generates an annotated three dimensional (3D) map with the agricultural conditions. Additionally, an object recognition model may be trained for use by a drone controller to trigger drones to conduct a collection of depth data at an increased proximity to crop-related objects based on agricultural conditions.

First claim

I claim:

1. An apparatus comprising:
a scene perceptor, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to generate a three-dimensional (3D) map of a terrain based on depth data associated with one or more drones;
a condition analyzer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to identify one or more agricultural conditions based the depth data and color data associated with at least one of the one or more drones, wherein the condition analyzer includes an object recognizer to recognize one or more crop-related objects, and a neural network to conduct an analysis of the one or more crop-related objects;
a camera locator, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to generate pose data based on the depth data and inertia data associated with the one or more drones;
a map enhancer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, communicatively coupled to the scene perceptor and the condition analyzer, the map enhancer to annotate the 3D map with the one or more agricultural conditions and the pose data; and
a drone controller, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to trigger a collection of the depth data at an increased proximity to the one or more crop-related objects based on the one or more agricultural conditions, wherein one or more images associated with the collection are to correspond to a canopied area of an interior space.

2. The apparatus of claim 1 , further including:
a classifier trainer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to train one or more classifiers of the neural network based on one or more training images of the terrain; and
an object recognition trainer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to train an object recognition model based on the one or more training images.

3. The apparatus of claim 1 , wherein the one or more agricultural conditions are to include one or more of crop disease, animal-imposed damage or water-imposed damage.

4. An apparatus comprising:
a scene perceptor, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to generate a three-dimensional (3D) map of a terrain based on depth data associated with one or more drones;
a condition analyzer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to identify one or more agricultural conditions based the depth data and color data associated with at least one of the one or more drones;
a camera locator, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to generate pose data based on the depth data and inertia data associated with the one or more drones; and
a map enhancer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, communicatively coupled to the scene perceptor and the condition analyzer, the map enhancer to annotate the 3D map with the one or more agricultural conditions and the pose data.

5. The apparatus of claim 4 , wherein the condition analyzer includes:
an object recognizer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to recognize one or more crop-related objects; and
a neural network, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to conduct an analysis of the one or more crop-related objects.

6. The apparatus of claim 5 , further including:
a classifier trainer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to train one or more classifiers of the neural network based on one or more training images of the terrain; and
an object recognition trainer, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to train an object recognition model based on the one or more training images.

7. The apparatus of claim 4 , further including a drone controller, implemented at least partially in one or more of configurable logic or fixed functionality logic hardware, to trigger a collection of the depth data at an increased proximity to one or more crop-related objects based on the one or more agricultural conditions.

8. The apparatus of claim 7 , wherein one or more images associated with the collection are to correspond to a canopied area of an interior space.

9. The apparatus of claim 4 , wherein the one or more agricultural conditions are to include one or more of crop disease, animal-imposed damage or water-imposed damage.

10. A method comprising:
generating a three-dimensional (3D) map of a terrain based on depth data associated with one or more drones;
identifying one or more agricultural conditions based on the depth data and color data associated with at least one of the one or more drones;
generating pose data based on the depth data and inertia data associated with the one or more drones; and
annotating the 3D map with the one or more agricultural conditions and the pose data.

11. The method of claim 10 , further including:
recognizing one or more crop-related objects; and
conducting an analysis of the one or more crop-related objects based on a neural network, wherein the one or more agricultural conditions are identified based on the analysis.

12. The method of claim 11 , further including:
training one or more classifiers of the neural network based on one or more training images of the terrain; and
training an object recognition model based on the one or more training images.

13. The method of claim 10 , further including triggering a collection of the depth data at an increased proximity to one more crop-related objects based on the one or more agricultural conditions.

14. The method of claim 13 , wherein one or more images associated with the collection correspond to a canopied area of an interior space.

15. The method of claim 10 , wherein the one or more agricultural conditions include one or more of crop disease, animal-imposed damage or water-imposed damage.

16. At least one non-transitory computer readable storage medium comprising a set of instructions, which when executed, cause a computing device to:
generate a three-dimensional (3D) map of a terrain based on depth data associated with one or more drones;
identify one or more agricultural conditions based on the depth data and color data associated with at least one of the one or more drones;
generate pose data based on the depth data and inertia data associated with the one or more drones; and
annotate the 3D map with the one or more agricultural conditions and the pose data.

17. The at least one non-transitory computer readable storage medium of claim 16 , wherein the instructions, when executed, cause a computing device to:
recognize one or more crop-related objects; and
conduct an analysis of the one or more crop-related objects based on a neural network, wherein the one or more agricultural conditions are to be identified based the analysis.

18. The at least one non-transitory computer readable storage medium of claim 17 , wherein the instructions, when exe

Assignees

• Intel Corp

Inventors

• Bleiweiss Amit

Classifications

• B64U2201/10

  autonomous, i.e. by navigating independently from ground or air stations, e.g. by using inertial navigation systems [INS] · CPC title

• B64U2201/20

  Remote controls · CPC title

• B64U2101/30

  for imaging, photography or videography · CPC title

• G06T2207/30188

  Vegetation; Agriculture · CPC title

• G06T2207/10024

  Color image · CPC title