WO2017048783A1

WO2017048783A1 - Foreign object detection in beef using color analysis

Info

Publication number: WO2017048783A1
Application number: PCT/US2016/051639
Authority: WO
Inventors: Scott T. HOWARD
Original assignee: Cargill, Incorporated
Priority date: 2015-09-14
Filing date: 2016-09-14
Publication date: 2017-03-23

Abstract

A system and method detects foreign objects on at least one surface of beef pieces in a beef packing plant processing line. In the system, an image capture device is configured to capture a plurality of captured images formed from visible light reflected from the surface of the same beef pieces and any permitted background, wherein at least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images. Computer processor software compares at least two of the captured images to determine differences between the compared images indicative of the presence of a foreign object on the beef pieces.

Description

Foreign Object Detection in Beef Using Color Analysis

CROSS-REFERENCE

The present application claims priority to United States Provisional Application No. 62/218,249, filed September 14, 2015, the disclosure of which is incorporated by reference herein in its entirety.

FIELD OF THE INVENTION

The present invention relates to beef processing. More specifically, the present invention relates to the detection of foreign object on beef in beef processing environments.

BACKGROUND OF THE INVENTION

Beef is efficiently provided on a large scale through commercial beef packing plants. Such plants deliver economic benefit and high quality product through the ability to quickly process a number of animals in a rigorously controlled environment. Various approaches have been taken to assure quality of beef being processed, including detection of foreign materials.

For example, US Patent No. 4,208,915 describes a method of determining foreign material such as metal and bone particles in food products. The method uses ultrasonic sound frequencies from a plurality of transducers disposed in a rotatable cylinder having a liquid copulant. In another approach, systems and techniques for detecting the presence of foreign material in food utilizing optical backlighting and/or ultrasonic inspection are described in US Patent No. 6,992,771. EP Patent No.

361625A2 describes detection of undesirable ingredients in food product (e.g. bone) in meat wherein the local intensity of a laser beam that is shined through a food product is determined for detecting undesired ingredients.

Bone detection in deboned meat or poultry products is described in US Patent No. 5,847,382 by using a video camera that receives light transmitted from flattened deboned meat product and projects the image of meat product on a monitor, revealing bone fragments and other defects. US Patent No. 7,460,227 describes detection of bones, bone fragments, and shell fragments on or near the surface of processed animal flesh by detecting their fluorescent emissions. The surface of the flesh is illuminated with IJV or visible light to elicit fluorescence of animal bones or shells, and the fluorescent light emissions are measured at first and second different wavelengths, which wavelengths are selected to yield a substantial difference between the ratio or subtractive difference for bones or shells and the ratio or subtractive difference for animal flesh.

US Patent No. 7,060,981 describes an imaging system for detection of embedded objects in specimens. A computer system receives images from detector systems on a conveyor within an imaging zone and produces image slices through image combination as specimen representations.

SUMMARY OF THE INVENTION

Previous approaches taken to detect foreign objects on beef have either been inefficient, ineffective, or very expensive to implement. For example, plant workers may be stationed to attempt to visually inspect the production line, but can easily overlook small pieces of foreign objects. Alternatively, the worker's attention may be drawn away from the line for a moment, allowing the foreign object to pass by undetected. The cost of using sophisticated detection equipment, such as laser scanners, x-ray and hyperspectral imaging equipment may be prohibitive while at the same time not providing an appropriate solution. For example, X-ray scanning equipment is appropriate for detecting bone and metal objects, but will not detect the presence of less dense materials that do no sink in water - e.g. gloves, plastic, wood.

The present system uses easily obtainable equipment to detect the specific foreign objects that are most likely to cause problems in commercial beef packing plant facilities. Specifically, materials such as glove materials, cotton, plastics from storage containers, and conveyer belt materials may be readily and easily detected in the present system and method, while they are very difficult to detect under production speed conditions using prior art systems.

In the system, a conveyor is configured to move a plurality of beef pieces in a beef processing operation, and an image capture device is configured to capture a plurality of captured images formed from visible light reflected from the surface of the same beef pieces and any permitted background. At least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm. At least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images. A computer processor is configured to receive the images and comprises software capable of comparing at least two of the captured images to determine differences between the compared images indicative of the presence of a foreign object on the beef pieces.

Methods for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line are also described.

It has surprisingly been found that comparison of images wherein selection of filters so that at least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425 nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images provides surprising contrast and ability to the presence of a foreign object on beef pieces. This is observed despite the fact that the three wavelength ranges are associated with the reflectance of light from beef.

The use of filters to block or absorb light in desired wavelengths before capture of the image provides significant benefit, because the resulting images may be compared and analyzed without the need to resort to complex data processing software. By using a wavelength filter system that operates prior to image capture, both speed and accuracy can be obtained at reasonable computer processor costs.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate several aspects of the invention and together with a description of the embodiments serve to explain the principles of the invention. A brief description of the drawings is as follows:

FIG. 1 is an elevation view of an embodiment of the present system.

DETAILED DESCRD7TION OF PRESENTLY PREFERRED

EMBODIMENTS

The embodiments of the present invention described below are not intended to be exhaustive or to limit the invention to the precise forms disclosed in the following detailed description. Rather a purpose of the embodiments chosen and described is so that the appreciation and understanding by others skilled in the art of the principles and practices of the present invention can be facilitated. The present system detects foreign objects on at least one surface of beef pieces in a beef packing plant processing line. The beef may be inspected for the presence of foreign objects at any stage in processing. In an embodiment, the beef is inspected upstream in processing, i.e. after having been initially divided into carcass halves, primals or subprimals. In an embodiment, the beef is inspected downstream in processing, i.e. after having been separated into individual portions having a size of from about 8 oz to about 25 oz. In an embodiment, the beef to be inspected is trim beef.

The present system uses a conveyor configured to move a plurality of beef pieces in a beef processing operation. This conveyor may be any suitable arrangement for moving beef, including a conveyor belt, a suspension system and the like. In an embodiment, the conveyor belt is prepared from a durable material that is resistant to fracture, spalling or otherwise loss or separation of fragments from the conveyor that could be associated with the beef as a foreign object. In an

embodiment, the conveyor is manufactured from a material having a color that is readily distinguishable from the beef being processed, so that in the event that the conveyor were to lose fragments therefrom, the fragments could be easily identified as a foreign object on the beef. In an embodiment, the conveyor is made from metal materia] In an embodiment, the conveyor is made from a plastic or rubber material. In an embodiment, the conveyor is made from a plastic or rubber material in a color that is in contrast with the color of beef. In an embodiment, the conveyor is made from a material having a color selected from green or blue.

Fig. 1 is an elevation view of a portion of a beef packing plant processing line, wherein an embodiment of the present system 10 is shown. Beef pieces 12, such as trim, are conveyed on a first belt 14, and are allowed to drop to second trim belt 16 past image capture device 20. First belt 14 and second belt 16 have directions of travel as shown. Image capture device 20 is configured to capture a series of images formed from visible light reflected from the surface of beef pieces. In an

embodiment, a plurality of image capture devices are configured to capture a series of images formed from visible light reflected from the surface of beef pieces from different sides of the beef. In an embodiment, a plurality of image capture devices are configured to capture a series of images formed from visible light reflected from the surface of beef pieces so images are captured of all sides of the beef. System 10 additionally comprises a permitted background 22 that is within the frame of images captured by image capture device 20. The permitted background 22 comprises objects known to be present in the background of the image. In an embodiment, permitted background 22 is a single surface having a uniform color to simplify analysis of the beef pieces to determine if a foreign object is present.

In an embodiment, permitted background 22 is provided in a color that is in contrast with the color of beef. In an embodiment, permitted background 22 is made from a material having a color selected from green or blue. In an embodiment, permitted background 22 is spaced from the beef so that no contact with the beef can be made at the time of image capture. In the embodiment where permitted background 22 is spaced from the beef, permitted background 22 may in an embodiment be provided in a color that is identical to at least one of the colors of light reflected by the beef being inspected.

The image captures from image capture device 20 are transmitted to a computer processor (not shown) that is configured to receive the images and comprising software capable of analyzing image capture data to detect one or more foreign objects on the beef pieces. The captured visible light images are correlated with the beef pieces appearing in the frame of the images, so that an action can be performed on beef pieces having a foreign object thereon in subsequent processing steps. On determination of the presence of foreign objects on beef pieces, system 10 in one embodiment can carry out a separation step of separating beef pieces having foreign objects on the surface thereof from beef pieces free of foreign objects. This separation step can be carried out by any appropriate technique such as redirecting pieces of beef by actuating directional arms, a conveyor retraction apparatus, and the like. As shown in Fig. 1, second belt 16 comprises a belt retraction device 18 that can be actuated so mat beef pieces having foreign objects on the surface thereof fall to a reject bin 24 below second beh 16.

On determination of the presence of foreign objects on beef pieces, the system in one embodiment can carry out a removal step of removing the foreign object from the beef pieces having foreign objects on the surface thereof. The removal step may be identification of the beef to a production worker for action, directing a wash solution to the beef piece, or any other such step as appropriate.

The image capture device is a camera capable of digitally capturing images of reflected light. Examples of appropriate image capture devices include multispectral cameras such as the SpectroCam™ by Pixelteq, inc..

In an embodiment, the image capture device is capable of capturing a plurality of images of beef in motion at a flow of from 5 tons to 30 tons per hour, or from 10 tons to 30 tons per hour, or from 15 tons to 30 tons per hour.

Certain predetermined wavelength ranges corresponding to those reflected by beef may be characterized, for example, the wavelengths of light associated with myoglobin in beef. It has been discovered that simple manipulation of the light being received by the image capture device provides significant benefit to ready detection of foreign objects on beef, because by blocking one or more wavelength ranges, foreign objects that reflect light in non-blocked wavelengths may be more readily identified.

In an embodiment, it has further been discovered that certain foreign objects exhibit a different contrast with beef when an image having one region of wavelength of reflected light is blocked as compared to a like image having a different blocked wavelength range or an image where no wavelength range is blocked. As noted above, a computer processor is configured to receive the images and comprises software capable of analyzing image capture data to detect one or more foreign objects on the beef pieces. In an embodiment, this analyzing step comprises comparing images of the same beef piece or beef pieces having different wavelength ranges blocked (or not blocked). This analyzing step may include particular analysis of a wavelength region corresponding to a specific potential foreign object that exhibits heightened contrast in this comparison. For example, a potential foreign object commonly present in beef packing plants is a glove. When an image of beef pieces that also comprises a glove fragment is taken wherein the image is analyzed at a wavelength range between 655nm to 665 nm, the portion of the image containing a portion of glove will reflect a greater amount of light and that portion of the image should appear brighter. When an image of the same beef pieces is taken wherein the image does not contain any pieces of glove, less light will be reflected when the image is analyzed specifically from 655 to 665 nm and the image will be darker. The reflectance of various potential foreign objects as observed when the light is filtered to different wavelengths is shown in Table 1.

In an embodiment, a system is provided for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line comprising: a conveyor configured to move a plurality of beef pieces in a beef processing operation;

an image capture device configured to capture a plurality of captured images formed from visible light reflected from the surface of the same beef pieces and any permitted background, wherein at least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to S8Snm, or c) from 41 Snm to 425nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images; and

a computer processor configured to receive the images and comprising software capable of comparing at least two of the captured images to determine differences between the compared images indicative of the presence of a foreign object on the beef pieces.

It has been discovered that use of at least one image having a wavelength range of no more than about 5 nm on either side of a central wavelength of 660 nm, or 580 nm, or 420 nm provides significantly improved contrast in reflectance of various potential foreign objects than when the images being compared all have broader ranges of wavelength range data. In an embodiment, at least one image has a wavelength range of no more than about 3 nm on either side of a central wavelength of 660 nm, or 580 nm, or 420 nm. In an embodiment, at least one image has a wavelength range of no more than about 1 nm on either side of a central wavelength of 660 nm, or 580 nm, or 420 nm.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm, or b) from 577nm to 583nm, or c) from 417nm to 423nm.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm and at least one image that is not filtered.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of b) from 575nm to 585nm and at least one image that is not filtered.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm and at least one image that is not filtered.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm and at least one image filtered to contain only data related to a wavelength range of b) from 575nm to 585nm.

In an embodiment, the image capture device is configured to capture at least one image is filtered to contain only data related to a wavelength range of a) from 657nm to 663nm and at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm.

In an embodiment, the image capture device is configured to capture at least one image is filtered to contain only data related to a wavelength range of b) from 575nm to 585nm and at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm.

In an embodiment, the image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm, at least one image filtered to contain only data related to a wavelength range of b) from 575nm to 585nm, and at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm.

In an embodiment, at least one image is filtered to contain only data related to any wavelengths except those corresponding to the reflected light from known potential sources of foreign objects that are typically present in the beef packing plant. In an embodiment, at least one image is filtered to contain only data related to any wavelengths except those corresponding to one or more materials from the group consisting of wood, glove material used in the beef packing plant, cotton material, bone material, plastic container material used in the beef packing plant, and conveyer belt material used in the beef packing plant.

In an embodiment, the Image is physically filtered to block light other than the desired wavelength. In an embodiment, the image capture device is provided with an optical filter that selectively transmits light in a particular range of wavelengths while blocking the remainder.

In an embodiment, the filter is an absorptive filter, wherein the filter absorbs the light wavelengths that are to be blocked from the image capture. In an embodiment, the filter is a reflective or interference filter, which reflects the light wavelengths that are to be blocked from the image capture.

In an embodiment, each image capture device has a single filer mounted thereon to filter the desired range of wavelength of light In an embodiment, an image capture device is provided with a filter wheel on which several filters are mounted. In an embodiment, the same filter wheel may be used to provide filtering to a plurality of image capture devices.

In an embodiment, a plurality of images are captured of each piece of beef. In an embodiment, a plurality of images are captured of a group of pieces of beef,

In an embodiment, a method is provided for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line, using the above described system in all of the variants described above. Specifically, in an embodiment, a method is provided for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line comprising:

moving a plurality of beef pieces with a conveyor system in a beef processing operation;

capturing a series of images formed from visible light reflected from the surface of the same beef pieces and any permitted background, wherein at least one image is filtered to contain only data related to a wavelength range of a) from 6SSnm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images; and

detecting one or more foreign objects on beef pieces by comparing using a computer processor at least two of the captured images to determine differences between the compared images indicative of the presence of a foreign object on the beef pieces.

As used herein, the terms "about" or "approximately" mean within an acceptable range for the particular parameter specified as determined by one of ordinary skill in the art, which will depend in part on how the value is measured or determined, e.g., the limitations of the sample preparation and measurement system. Examples of such limitations include preparing the sample in a wet versus a dry environment, different instruments, variations in sample height, and differing requirements in signal-to-noise ratios. For example, "about" can mean greater or lesser than the value or range of values stated by 1/10 of the stated values, but is not intended to limit any value or range of values to only this broader definition. For instance, a concentration value of about 30% means a concentration between 27% and 33%. Each value or range of values preceded by the term "about" is also intended to encompass the embodiment of the stated absolute value or range of values.

Alternatively, particularly with respect to biological systems or processes, the term can mean within an order of magnitude, preferably within S-fold, and more preferably within 2-fold, of a value.

Throughout this specification and claims, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and

"comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integer or step. When used herein "consisting of excludes any element, step, or ingredient not specified in the claim element. When used herein, "consisting essentially of does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim. In the present disclosure of various embodiments, any of the terms "comprising", "consisting essentially of and

"consisting of used in the description of an embodiment may be replaced with either of the other two terms. All patents, patent applications (including provisional applications), and publications cited herein are incorporated by reference as if individually incorporated for all purposes. Unless otherwise indicated, all parts and percentages are by weight and all molecular weights are weight average molecular weights. The foregoing detailed description has been given for clarity of understanding only. No unnecessary limitations are to be understood therefrom. The invention is not limited to the exact details shown and described, for variations obvious to one skilled in the art will be included within the invention defined by the claims.

Claims

WHAT IS CLAIMED IS::

1. A system for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line comprising:

a conveyor configured to move a plurality of beef pieces in a beef processing operation;

an image capture device configured to capture a plurality of captured images formed from visible light reflected from the surface of the same beef pieces and any permitted background, wherein at least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665 nm, or b) from 575nm to

585nm, or c) from 415nm to 425nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images; and

2. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from

657nm to 663nm, or b) from 577nm to 583nm, or c) from 417nm to 423nm.

3. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm.

4. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm and at least one image that is not filtered.

5. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of b) from 575nm to 585nm and at least one image that is not filtered.

6. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm and at least one image that is not filtered.

7. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 6S7nm to 663 nm and at least one image filtered to contain only data related to a wavelength range of b) from 575nm to 585nm.

8. The system of claim 1, wherein image capture device is configured to capture at least one image is filtered to contain only data related to a wavelength range of a) from 657nm to 663 nm and at least one image filtered to contain only data related to a wavelength range c) from 4l5nm to 425nm.

9. The system of claim 1, wherein image capture device is configured to capture at least one image is filtered to contain only data related to a wavelength range of b) from 575nm to 585nm and at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm.

10. The system of claim 1, wherein image capture device is configured to capture at least one image filtered to contain only data related to a wavelength range of a) from 657nm to 663nm, at least one image filtered to contain only data related to a wavelength range of b) from S7Snm to 585nm, and at least one image filtered to contain only data related to a wavelength range c) from 415nm to 425nm.

1 1. The system of any of claims 1-10, wherein at least one image does not contain data related to any wavelengths except those corresponding to the reflected light from identified potential sources of foreign objects that are present in the beef packing plant.

12. The system of any of claims 1-10, wherein at least one image does not contain data related to any wavelengths except those corresponding to one or more materials from the group consisting of wood, glove material used in the beef packing plant, cotton material, bone material, plastic container material used in the beef packing plant, and conveyer belt material used in the beef packing plant.

13. The system of any of claims 1- 12, wherein the image capture device is configured to capture a plurality of images of each individual piece of beef for detection of one or more foreign objects on each individual piece of beef.

14. The system of any of claims 1-12, wherein the image capture device is configured to capture a plurality of images of a group of pieces of beef for detection of one or more foreign objects on at least one of the pieces of beef in the group.

15. The system of any of claims 1-14, further comprising a separation device for separating beef pieces having foreign objects on the surface thereof from beef pieces free of foreign objects.

16. The system of claim 15, wherein the computer processor controls the separation step.

17. The system of any of claims 1-16, wherein the detection step further comprises correlating in the computer processor the captured visible light images with the imaged beef pieces image so that an action may be performed on the correlated beef pieces in subsequent processing steps.

18. A method for detecting foreign objects on at least one surface of beef pieces in a beef packing plant processing line comprising:

capturing a series of images formed from visible light reflected from the surface of the same beef pieces and any permitted background, wherein at least one image is filtered to contain only data related to a wavelength range of a) from 655nm to 665nm, or b) from 575nm to 585nm, or c) from 415nm to 425nm, and wherein at least one of the captured images does not contain data related to a wavelength range that is present in at least one of the other captured images; and detecting one or more foreign objects on beef pieces by comparing using a computer processor at least two of the captured images to determine differences between the compared images indicative of the presence of a foreign object on the beef pieces.