Arrangement and method for determining a body condition score of an animal

• US 9,684,956 B2
• Filed: 11/03/2009
• Issued: 06/20/2017
• Est. Priority Date: 12/03/2008
• Status: Active Grant

• Current Assignee: Delavan Limited (Enpro Industries Incorporated)
• Original Assignee: Delavan Limited (Enpro Industries Incorporated)
• Inventors: Liao, Bohao, Krukowski, Marilyn
• Primary Examiner(s): Kelley, Christopher S
• Assistant Examiner(s): KWAN, MATTHEW K
• Application Number:
  US13/132,107
• Publication Number:
  US 20110279650A1
• Time in Patent Office: 2,786 Days
• Field of Search: 348 46, 119 5102
• CPC Class Codes: A01K 29/00 Other apparatus for animal husbandry
• CPC Class Codes: A01K 29/005 Monitoring or measuring activity, e.g. detecting heat or mating
• CPC Class Codes: G06T 2200/04 involving 3D image data
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/20036 Morphological image processing
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/40 Analysis of texture depth or shape recovery from texture G06T7/529
• CPC Class Codes: G06T 7/41 based on statistical description of texture
• CPC Class Codes: G06V 20/64 Three-dimensional objects
• CPC Class Codes: G06V 2201/121 using special illumination
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands

Abstract

An arrangement for determining a body condition score of an animal comprises a three-dimensional camera system directed towards the animal and provided for recording at least one three-dimensional image of the animal; and an image processing device connected to the three-dimensional camera system and provided for forming a three-dimensional surface representation of a portion of the animal from the three- dimensional image recorded by the three-dimensional camera system; for statistically analyzing the surface of the three-dimensional surface representation; and for determining the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation.

Claims

1. An arrangement for determining a body condition score of an animal, comprising:

a three-dimensional camera system provided for being directed towards the animal and for recording at least one three-dimensional image of the animal; and

an image processing device connected to the three-dimensional camera system and provided for;

forming a three-dimensional surface representation of a portion of the animal from the three-dimensional image recorded by the three-dimensional camera system;

statistically analyzing the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation,

wherein the statistical analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the pin or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the pin or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the pin or hook bones.

2. The arrangement of claim 1, wherein said at least one three-dimensional image of the animal comprises a two-dimensional image of the animal and, for each pixel of the two-dimensional image, the three-dimensional image further includes a distance value reflecting a distance between the three-dimensional camera system and an object image point associated with the respective pixel.

3. The arrangement of claim 1, wherein said statistical analyzing includes analyzing the unevenness of the surface of the three-dimensional surface representation.

4. The arrangement of claim 1, wherein the histogram is of a feature parameter of the surface of the three-dimensional surface representation, wherein the feature parameter is one of surface gradients, surface normals, or range or depth values.

5. The arrangement of claim 1, wherein said statistical analyzing includes analyzing spectral measures of the unevenness of the surface of the three-dimensional surface representation.

6. The arrangement of claim 1, wherein said statistical analyzing includes analyzing surface curvatures or spin-images of the surface of the three-dimensional surface representation.

7. The arrangement of claim 1, wherein said statistical analyzing includes analyzing feature parameters of the vertebrae of the animal.

8. The arrangement of claim 1, wherein said statistical analyzing includes analyzing feature parameters of the hook bones of the animal.

9. The arrangement of claim 1, wherein said statistical analyzing includes analyzing feature parameters of the ribs of the animal.

10. The arrangement of claim 1, wherein said statistical analyzing includes analyzing feature parameters of cavities between the tail head and the pin bones of the animal.

11. The arrangement of claim 1, wherein the portion of the animal is a region of at least one of the back or loin of the animal.

12. The arrangement of claim 11 wherein the portion of the animal has an area, which area is smaller than the area of the back of the animal, and which area is located in front of the tail head and narrow enough to not include the hip or pin bones of the animal.

13. The arrangement of claim 1, wherein:

said image processing device is further provided for processing the three-dimensional image prior to forming the three-dimensional surface representation;

said processing of the three-dimensional image comprises filtering of the three-dimensional image; and

said filtering comprises removing image data points depending on at least one of an intensity value, or distance information associated with the image data points.

14. The arrangement of claim 13, wherein said processing of the three-dimensional image comprises segmenting the three-dimensional image.

15. The arrangement of claim 1, wherein said three-dimensional camera system is a time-of-flight range camera or an active wavefront sampling camera.

16. The arrangement of claim 1, wherein the three-dimensional camera system is located above and behind the animal, and is directed downwards and forward, towards the back of the animal, to provide image information regarding the dorsal and posterior portions of the animal.

17. The arrangement of claim 1, wherein the three-dimensional camera system is located at a side of the animal and above the animal, and is directed sideways and downwards, to provide a lateral/dorsal image of the animal showing the area between the pin bones and the hook bones.

18. An animal feeding, milking, or resting station provided with the animal identification device and the arrangement of claim 1.

19. An animal arrangement comprising a herd management system and the arrangement of claim 1 operatively connected to the herd management system and provided to supply the herd management system with the determined body condition score of the animal.

20. A method for determining a body condition score of an animal, comprising:

directing a three-dimensional camera system towards the animal and recording at least one three-dimensional image of the animal;

forming a three-dimensional surface representation of a portion of the animal from the three-dimensional image recorded by the three-dimensional camera system;

statistically analyzing the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation,

wherein the statistical analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

21. The method of claim 20, wherein the at least one three-dimensional image of the animal comprises a two-dimensional image of the animal and, for each pixel of the two-dimensional image, the three-dimensional image further includes a distance value reflecting a distance between the three-dimensional camera system and an object image point on the animal.

22. An arrangement for determining a body condition score of an animal comprising:

a three-dimensional camera system provided for being directed towards the animal and for recording at least one three-dimensional image of a portion of the animal; and

an image processing device connected to the three-dimensional camera system and provided for;

forming a three-dimensional surface representation from the three-dimensional image;

statistically analyzing the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the normalized three-dimensional surface representation,

wherein the statistical analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

23. A method for determining a body condition score of an animal, comprising:

directing a three-dimensional camera system towards the animal and recording at least one three-dimensional image of a portion of the animal;

forming a three-dimensional surface representation from the three-dimensional image;

statistically analyzing a surface of the three-dimensional surface representation by calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the normalized three-dimensional surface representation,

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

24. An arrangement for determining a body condition score of an animal, comprising:

a three-dimensional camera system provided for being directed towards the animal and for recording at least one three-dimensional image of a portion of the animal; and

an image processing device connected to the three-dimensional camera system and provided for;

forming a three-dimensional surface representation from the three-dimensional image;

recognizing anatomical features in said three-dimensional surface representation;

statistically analyzing a surface of the three-dimensional surface representation by calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation,

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

25. A method for determining a body condition score of an animal, comprising:

directing a three-dimensional camera system towards the animal and recording at least one three-dimensional image of a portion of the animal;

forming a three-dimensional surface representation from the three-dimensional image;

recognizing anatomical features in said three-dimensional surface representation;

statistically analyzing a surface of the three-dimensional surface representation by calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation;

determining the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation,

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

26. A method for managing animals using body condition score, comprising:

obtaining at least one three-dimensional image of the animal using a camera system;

forming a three-dimensional surface representation of at least a portion of the animal based on the three-dimensional image;

analyzing the three-dimensional surface representation to determine a body condition score for the animal representative of a fatness or thinness of the animal;

automatically triggering an alarm or an action to be taken with respect to the animal, based on the determined body condition score, wherein

the analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over a surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

searching, prior normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalizing the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

27. An arrangement for managing animals using body condition score, comprising:

a three-dimensional camera system arranged to obtain at least one three-dimensional image of the animal; and

an image processing device configured to;

form a three-dimensional surface representation of at least a portion of the animal based on the three-dimensional image;

analyze the three-dimensional surface representation to determine a body condition score for the animal representative of a fatness or thinness of the animal;

automatically trigger an alarm or an action to be taken with respect to the animal, based on the determined body condition score,

wherein the analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over a surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

search, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the in or hook bones of the animal in the three-dimensional surface representation; and

normalize the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the in or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the in or hook bones.

28. An arrangement for determining a body condition score of an animal, comprising:

a three-dimensional camera system configured to be directed towards the animal and for recording at least one three-dimensional image of the animal; and

an image processing device connected to the three-dimensional camera system and configured to;

form a three-dimensional surface representation of a portion of the animal from the three-dimensional image recorded by the three-dimensional camera system,

wherein the portion of the animal is a region of at least one of the back or loin of the animal, and

wherein the portion of the animal has an area, which area is smaller than the area of the back of the animal, and which area is located in front of the tail head and narrow enough to not include the hip or pin bones of the animal;

statistically analyze the surface of the three-dimensional surface representation;

determine the body condition score of the animal based on the statistically analyzed surface of the three-dimensional surface representation,

wherein the statistical analyzing includes calculating statistical properties of a histogram that compares how measurement data may vary over the surface of the three-dimensional surface representation, and

wherein the statistical properties comprise at least one of mean value, standard deviation, smoothness, skewness, uniformity, entropy, the width of the histogram, the width at the half maximum value, or parameters of a curve fitted to the histogram;

search, prior to normalizing the three-dimensional surface representation, for reference points in the three-dimensional surface representation of the portion of the animal serving as control points for the normalization, said reference points including the spinal ridge and the pin or hook bones of the animal in the three-dimensional surface representation; and

normalize the three-dimensional surface representation of the portion of the animal by rotation, translation, and scaling based on the location of the reference points such that the spinal ridge is essentially parallel with a first axis of a Cartesian coordinate system, a connection line between the tips of the pin or hook bones is essentially parallel with a first axis of a Cartesian coordinate system, and the scaling of the three-dimensional surface representation is made depending on the distance between the tips of the pin or hook bones.