Distinguishing abutting food product

• US 8,233,668 B2
• Filed: 08/09/2010
• Issued: 07/31/2012
• Est. Priority Date: 08/09/2010
• Status: Active Grant

• Current Assignee: John Bean Technologies Corporation
• Original Assignee: John Bean Technologies Corporation
• Inventors: Jing, Hansong, Blaine, George R.
• Primary Examiner(s): Lu, Tom Y
• Application Number:
  US12/853,195
• Publication Number:
  US 20120034349A1
• Time in Patent Office: 722 Days
• Field of Search: 382/110, 382/132
• US Class Current: 382/110
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20036 Morphological image processing
• CPC Class Codes: G06T 2207/30128 Food products
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06T 7/155 involving morphological operators
• CPC Class Codes: G06V 10/267 by performing operations on regions, e.g. growing, shrinking or watersheds

Abstract

A method for determining the boundary between abutting food product in a food processing system uses scan data of the food product to identify a perimeter of the abutting food product. Optionally, data from scanning the food product is first tested to determine or predict if more than one article is present. A perimeter of the abutting food product is generated from the scan data, and the perimeter information is manipulated to identify or estimate the boundary between the product. For example, the boundary may be eroded until it separates into two portions. The two portions are then expanded, constrained by the original perimeter, such that a region of interest includes overlapping portion that includes the boundary. The region of interest is analyzed to locate the boundary. In another method, a parametric shape is fit to the image to identify the individual articles. In another method, a shortest line or surface separating the perimeter into two substantial portions is found.

Claims

1. A method for automatically identifying a shared boundary between two abutting food product articles in a food processing system, the method comprising the steps:

i. scanning a pair of food articles comprising a first food article abutting a second food article, wherein a shared boundary is defined between the first and second food articles to produce a scan data set;

ii. identifying from the scan data set a closed contour corresponding to a perimeter of the scanned pair of food articles, wherein the closed contour encloses the first and second food articles;

iii. analyzing the closed contour to estimate the location of the shared boundary between the first and second food articles.

2. The method of claim 1, wherein the scanning is accomplished with at least one of a video camera, an x-ray scanning system, and an ultrasound scanning system.

3. The method of claim 1, wherein the step of analyzing the closed contour comprises:

eroding the closed contour from the outer perimeter until the closed contour separates into two separate closed contours;

growing the two separate contours from the outer perimeter, wherein the two separate contours are constrained by the closed contour identified from the scan data set until the two separate contours define an overlapping region of interest that extends across the shared boundary; and

analyzing the region of interest to determine the location of the shared boundary.

4. The method of claim 3, wherein the step of analyzing the overlapping portion comprises finding the shortest line segment that connects opposite sides of the overlapping portion.

5. The method of claim 3, wherein the step of a eroding the closed contour comprises:

converting the scan image to a binary image having a plurality of pixels, wherein each pixel has a first value if the pixel is outside the closed contour and a second value if the pixel is not outside the closed contour;

for each pixel in the binary image having the second value, setting the pixel value to the first value if any neighboring pixel has the first value; and

repeating the step of setting the pixel value to the first value if any neighboring pixel has the first value until the closed contour separates into two closed contours.

6. The method of claim 3, wherein the step of eroding the closed contour comprises:

at discrete points about the closed contour, calculating a local slope of the closed contour;

for each of the discrete points, moving the discrete point inwardly by a fixed amount in a direction locally perpendicular to the closed contour; and

repeating the step of moving the discrete point inwardly until the closed contour bifurcates.

7. The method of claim 3, wherein the closed contour is a three-dimensional contour, and wherein the step of eroding the closed contour comprises eroding the closed contour in three dimensions.

8. The method of claim 3, wherein the step of eroding the closed contour comprises identifying successive contours corresponding to the perimeter of the scanned pair of food products at increasing elevations until the closed contour separates into two separate closed contours.

9. The method of claim 1, wherein the step of analyzing the closed contour comprises:

selecting a minimum size for the first and second food articles;

identifying the shortest line segment that extends across the closed contour and that divides the closed contour into two portions, wherein each of the two portions is at least as large as the selected minimum size.

10. The method of claim 1, wherein the step of analyzing the closed contour comprises:

selecting an ideal shape characterizing at least a portion of the first and second food products;

identifying from the scan data a first portion of the closed contour that matches at least a portion of the selected ideal shape, and fitting the ideal shape to the first portion of the closed curve to define a first ideal model;

identifying from the scan data a second portion of the closed contour that matches at least a portion of the selected ideal shape, and fitting the ideal shape to the second portion of the closed curve to define a second ideal model; and

calculating the shared boundary location from the first and second ideal models in the region of the shared boundary.

11. The method of claim 10, wherein the location of the shared boundary is calculated from the midpoint between the first and second ideal models in the region of the shared boundary.

12. The method of claim 1, comprising the steps of searching for the shortest line segment that divides the closed contour into two portions, wherein the ratio of the size of the two portions is within a specified range.

13. The method of claim 1, wherein the step of scanning a pair of food articles further comprises scanning a third food article, wherein the third food article abuts at least one of the first and second food articles.

14. A method for discriminating between abutting food product disposed on a conveyor comprising:

i. scanning food product disposed on a conveyor as the food product passes by a scanning system;

ii. predicting when a scanned food product comprises a plurality of articles, wherein the plurality of articles are touching such that at least one shared boundary is formed between the plurality of articles;

iii. if a scanned food product is predicted to comprise a plurality of articles, then;

(a) generating a closed contour corresponding to a perimeter of the plurality of articles; and

(b) analyzing the closed contour to locate the shared boundary between the plurality of articles.

15. The method of claim 14, wherein the step of predicting when a scanned food product comprises a plurality of touching articles comprises determining at least one physical attribute of the scanned food product and, if the determined at least one attribute exceeds a predefined value, then predicting that a plurality of touching articles are scanned.

16. The method of claim 15, wherein the physical attribute of the scanned food product comprises at least one of the length, width, area, volume, and weight of the scanned food product.

17. The method of claim 15, wherein the scanning is accomplished with at least one of a video camera, an x-ray scanning system, and an ultrasound scanning system.

18. The method of claim 14, wherein the step of analyzing the closed contour comprises:

eroding the closed contour from the outer perimeter until the closed contour separates into two separate closed contours;

expanding the two separate closed contours, wherein the two separate contours are constrained by the generated closed contour, until the two separate contours define an overlapping region of interest; and

analyzing the region of interest to estimate the location of the shared boundary.

19. The method of claim 18, wherein the step of analyzing the overlapping region of interest comprises finding the shortest line segment that connects opposite sides of the overlapping region of interest.

20. The method of claim 18, wherein the step of eroding the closed contour comprises:

converting the closed contour to a binary image having a plurality of pixels, wherein each pixel has a first value if the pixel is outside of the closed contour and a second value if the pixel is not outside the closed contour;

for each pixel in the binary image having the second value, setting the pixel value to the first value if any neighboring pixel has the first value; and

repeating the step of setting the pixel value to the first value if any neighboring pixel has the first value until the closed contour separates into two separate closed contours.

21. The method of claim 18, wherein the step of eroding the closed contour comprises:

at discrete points about the closed contour, calculating a local slope of the closed contour;

for each of the discrete points, moving the discrete point inwardly by a fixed amount in a direction locally perpendicular to the closed contour; and

repeating the step of moving the discrete point inwardly until the closed contour bifurcates.

22. The method of claim 18, wherein the closed contour is a three-dimensional contour, and wherein the step of eroding the closed contour comprises eroding the closed contour in three dimensions.

23. The method of claim 18, wherein the step of eroding the closed contour comprises identifying successive contours corresponding to the perimeter of the scanned pair of food products at increasing elevations until the closed contour separates into two separate closed contours.

24. The method of claim 14, wherein the step of analyzing the closed contour comprises:

selecting a minimum size for the first and second food articles;

identifying the shortest line segment that extends across the closed contour and that divides the closed contour into two portions, wherein each of the two portions is at least as large as the selected minimum size.

25. The method of claim 14, wherein the step of analyzing the closed contour comprises:

selecting an ideal shape characterizing at least a portion of the first and second food products;

identifying from the scan data a first portion of the closed contour that matches at least a portion of the selected ideal shape, and fitting the ideal shape to the first portion of the closed curve to define a first ideal model;

identifying from the scan data a second portion of the closed contour that matches at least a portion of the selected ideal shape, and fitting the ideal shape to the second portion of the closed curve to define a second ideal model; and

calculating the shared boundary location from the first and second ideal models in the region of the shared boundary.