Vision system with teat detection

Patent metadata

Field	Value
Publication number	US-9936670-B2
Application number	US-201615239559-A
Country	US
Kind code	B2
Filing date	Aug 17, 2016
Priority date	Aug 17, 2016
Publication date	Apr 10, 2018
Grant date	Apr 10, 2018

Abstract

A system that includes a laser configured to generate a profile signal of at least a portion of a dairy livestock, a memory operable to store a teat detection rule set, and a processor. The processor is configured to obtain the profile signal and detect one or more edge pair candidates in the profile signal, compare the complementary distance gradients of each of the one or more edge pair candidates to a minimum distance gradient length to be considered an edge pair, and identify one or more edge pairs from among the one or more edge pair candidates based on the comparison. The processor is further configured to apply the teat detection rule set to the one or more edge pairs to identify one or more teat candidates from among the one or more edge pairs and determine position information for the one or more teat candidates.

First claim

The invention claimed is:

1. A vision system comprising:
a laser configured to generate a profile signal of at least a portion of a dairy livestock, wherein the profile signal comprises:
information associated with a relative distance between the dairy livestock and the laser;
one or more rising distance gradients indicating an increase in the distance between the dairy livestock and the laser; and
one or more falling distance gradients indicating a decrease in the distance between the dairy livestock and the laser;

a memory operable to store a teat detection rule set; and
a processor operably coupled to the laser and the memory, and configured to:
obtain the profile signal;
detect one or more edge pair candidates in the profile signal, wherein each of the one or more edge pair candidates comprises complementary distance gradients comprising a rising distance gradient and a falling distance gradient;
compare the complementary distance gradients of each of the one or more edge pair candidates to a minimum distance gradient length to be considered an edge pair;
identify one or more edge pairs from among the one or more edge pair candidates based on the comparison, wherein each of the one or more edge pairs comprises complementary distance gradient lengths greater than or equal to the minimum distance gradient length;
apply the teat detection rule set to the one or more edge pairs to identify one or more teat candidates from among the one or more edge pairs; and
determine position information for the one or more teat candidates.

2. The system of claim 1 , wherein applying the teat detection rule set to the one or more edge pairs comprises:
comparing each of the one or more edge pairs to a teat model;
discarding edge pairs that do not correspond with the teat model; and
identifying an edge pair from among the one or more edge pairs that correspond with the teat model as a teat candidate.

3. The system of claim 1 , wherein:
the teat detection rule set indicates a minimum complementary distance gradient spacing to be considered a teat candidate; and
applying the teat detection rule set to the one or more edge pairs comprises:
comparing a complementary distance gradient spacing of each of the one or more edge pairs to the minimum complementary distance gradient spacing to be considered a teat candidate;
discarding edge pairs from the one or more edge pairs with a complementary distance gradient spacing less than the minimum complementary distance gradient spacing to be considered a teat candidate; and
identifying an edge pairs from the one or more edge pairs as a teat candidate when the edge pair has a complementary distance gradient spacing greater than or equal to the minimum complementary distance gradient spacing to be considered a teat candidate.

4. The system of claim 1 , wherein:
the teat detection rule set indicates a maximum complementary distance gradient spacing to be considered a teat candidate; and
applying the teat detection rule set to the one or more edge pairs comprises:
comparing a complementary distance gradient spacing of each of the one or more edge pairs to the maximum complementary distance gradient spacing to be considered a teat candidate;
discarding edge pairs from the one or more edge pairs with a complementary distance gradient spacing greater than the maximum complementary distance gradient spacing to be considered a teat candidate; and
identifying an edge pairs from the one or more edge pairs as a teat candidate when the edge pair has a complementary distance gradient spacing less than or equal to the maximum complementary distance gradient spacing to be considered a teat candidate.

5. The system of claim 1 , wherein:
the teat detection rule set indicates a maximum distance gradient length to be considered a teat candidate; and
applying the teat detection rule set to the one or more edge pairs comprises:
comparing the complementary distance gradient lengths of each of the one or more edge pairs to the maximum distance gradient length to be considered a teat candidate;
discarding edge pairs from the one or more edge pairs with a distance gradient length greater than the maximum distance gradient length to be considered a teat candidate; and
identifying an edge pair from the one or more edge pairs as a teat candidate when the edge pair has a complementary distance gradient length less than or equal to the maximum distance gradient length to be considered a teat candidate.

6. The system of claim 1 , wherein:
the teat detection rule set indicates a minimum distance gradient length to be considered a teat candidate; and
applying the teat detection rule set to the one or more edge pairs comprises:
comparing the complementary distance gradient lengths of each of the one or more edge pairs to the minimum distance gradient length to be considered a teat candidate;
discarding edge pairs from the one or more edge pairs with a distance gradient length less than the minimum distance gradient length to be considered a teat candidate; and
identifying an edge pair from the one or more edge pairs as a teat candidate when the edge pair has a complementary distance gradient length greater than or equal to the minimum distance gradient length to be considered a teat candidate.

7. An apparatus comprising:
a memory operable to store a teat detection rule set; and
a processor operably coupled to the memory, and configured to:
obtain a profile signal of at least a portion of a dairy livestock, wherein the profile signal comprises:
information associated with a relative distance between the dairy livestock and a laser;
one or more rising distance gradients indicating an increase in the distance between the dairy livestock and the laser; and
one or more falling distance gradients indicating a decrease in the distance between the dairy livestock and the laser; and

detect one or more edge pair candidates in the profile signal, wherein each of the one or more edge pair candidates comprises complementary distance gradients comprising a rising distance gradient and a falling distance gradient;
compare the complementary distance gradients of each of the one or more edge pair candidates to a minimum distance gradient length to be considered an edge pair;
identify one or more edge pairs from among the one or more edge pair candidates based on the comparison, wherein each of the one or more edge pairs comprises complementary distance gradient lengths greater than or equal to the minimum distance gradient length;
apply the teat detection rule set to the one or more edge pairs to identify one or more teat candidates from among the one or more edge pairs; and
determine position information for the one or more teat candidates.

8. The apparatus of claim 7 , wherein applying the teat detection rule set to the one or more edge pairs comprises:
comparing each of the one or more edge pairs to a teat model;
discarding edge pairs that do not correspond with the teat model; and
identifying an edge pair from among the one or more edge pairs that correspond with the teat model as a teat candidate.

9. The apparatus of claim 7 , wherein obtaining the profile signal of at least the portion of the dairy livestock comprises performing a laser scan of a surface of the dairy livestock.

10. The apparatus of claim 7 , wherein:
the teat detection rule set indicates a minimum complementary distance gradient spacing to be considered a teat candidate; and
applying the teat detection rule set to the one or more edge pairs comprises:
comparing a complementary distance gradient spacing of each of the one or more edge pairs to the minimum complementary distance gradient spacin

Assignees

• Technologies Holdings Corp

Inventors

• Foresman Mark A
• Prevost Bradley J

Classifications

• G06V20/10

  Terrestrial scenes (scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59) · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06F18/2113

  by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation · CPC title

• G06T7/74

  involving reference images or patches · CPC title

• G06T2207/10028

  Range image; Depth image; 3D point clouds · CPC title