WO2016048830A1

WO2016048830A1 - Object inspection system

Info

Publication number: WO2016048830A1
Application number: PCT/US2015/050965
Authority: WO
Inventors: Daniel Duane HAWK; Eric William GORHAM
Original assignee: The Steelastic Company, Llc
Priority date: 2014-09-23
Filing date: 2015-09-18
Publication date: 2016-03-31

Abstract

An inspection system is provided that includes at least one three-dimensional camera that is used to inspect an object to determine whether the object contains any defects. The defects that are capable of being detected by the inspection system include holes, tears, and improper thickness and overlap. The inspection system is configured to alert a user in the event that the object contains a defect.

Description

OBJECT INSPECTION SYSTEM

PRIORITY CLAIM

[0001] This invention claims the benefit of priority of U.S. Provisional Application Serial No. 62/054,112, entitled "Object Inspection System," filed September 23, 2014, the disclosure of which is hereby incorporated by reference in its entirety.

BACKGROUND

[0002] The present embodiments relate generally to a system for inspecting an object using a three dimensional camera.

[0003] Tire belt formation is a well-known practice that involves pulling multiple cords through an extrusion die. The extruder heats elastomeric material and coats the cords traveling through the die. Cooling drums adjacent to the extruder act both to pull the cords through the die and cool the fiber reinforced material before the cutting and splicing phase of production. After traveling through the cooling drums, the fiber reinforced material is allowed to hang with some slack in order to remove some residual forces. The fiber reinforced material is then drawn onto a cutting station. The cutting station includes a strip vacuum transfer, a cutter, and a belt conveyor. The strip vacuum transfer advances the fiber reinforced strip and positions it on the belt conveyor so that the cutter may cut the fiber reinforced material. The belt conveyor then indexes a predetermined distance. The strip vacuum transfer again advances the strip onto the conveyor so that the cutter again cuts it. This process results in a continuous belt of fiber reinforced material with the reinforcing cords lying at some angle typically not parallel to the central axis of the belt.

[0004] Defects can occur during the tire belt formation process that could potentially render the product unusable. For example, the tire belt formation process may result in a tire belt that contains holes or tears or has an improper thickness, width, or splice. To minimize or prevent these and other common defects from occurring, inspection systems are used to inspect the product. Traditional systems rely on a 2-dimensional camera to inspect the tire belt for defects. These traditional 2-dimensional camera systems require a strong backlight or front light to enable the 2-dimensional camera to detect certain defects, such as holes and tears. The light illuminates the inspection area and illuminates defects such as holes that pass entirely through the product.

[0005] The 2-dimensional camera, by its definition, is unable to detect defects that do not result in the complete penetration of the product because it is unable to detect differences in thickness. In other words, the 2-dimensional camera system is limited in its ability to detect variances in height and depth that are not detectable on the X and Y-axes. It is for at least this reason that a better, improved inspection system is needed to identify defects that are not detectable using a traditional 2-dimensional camera inspection system.

SUMMARY

[0006] One embodiment of the present invention includes an inspection system for inspecting an object including an inspection registration surface, a three-dimensional camera to measure the object on the inspection registration surface, a vision controller in

communication with the three-dimensional camera and applying algorithms to analyze the object based on user defined variables, a supervisory system which communicates to the vision controller, a host machine, and a user via HMI interface in order to properly report, categorize, and record all data captured from the analysis of the object.

[0007] Other embodiments of the present invention include three-dimensional camera being disposed above the inspection registration surface, or is configured to measure the width, gauge, splice height, splice height deviation, splice angle, edge linearity, surface coverage, or surface uniformity of an object.

[0008] Yet another embodiment of the present embodiment includes the vision controller using a series of algorithms to analyze each image and measure features and defects all being normalized to the inspection registration surface where applicable or wherein the supervisory system notifies the user and the host machine with an alarm when the measurements from the vision controller exceed a user defined parameter or subset of user defined parameters, wherein the supervisory system is fully customizable by the user to determine acceptable levels and machine alarm responses. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The invention can be better understood with reference to the following drawings and description. The components in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention. Moreover, in the figures, like referenced numerals designate corresponding parts throughout the different views.

[0010] FIG. 1 is a perspective view of an inspection system of one embodiment of the present invention.

[0011] FIG. 2 is a front view of the inspection system shown in FIG. 1.

[0012] FIG. 3 is a perspective view of the laser used to inspect an object using the inspection system of FIG. 1.

[0013] FIG. 4 is a partial schematic of the inspection software, encoder, and cameras used for the inspection system shown in FIG. 1.

[0014] FIG. 5 is a perspective view of a tire belt making system that employs the inspection system shown in FIG. 1.

[0015] FIG. 6 is an operational flow chart depicting an exemplary inspection procedure.

[0016] FIG. 7 is a partial front view of the inspection and calibration system shown in

FIG. 1.

[0017] FIG. 8 is a side view of the inspection system as one embodiment of the presented invention.

[0018] FIG. 9 is a front view of the inspection system shown in FIG. 8.

[0019] FIG. 10 is an example of the inspection system mounted on a host machine.

[0020] FIG. 11 is a hardware flow chart depicting an implementation of the inspection system.

[0021] FIG. 12 is an operational flow chart depicting an exemplary inspection procedure.

[0022] FIG. 13 is an operational flow chart depicting a calibration and alignment procedure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] Referring to FIGS. 1-6, an inspection system is generally indicated by numeral 10. The inspection system 10 may be used in any tire belt making system or any other system where defects can exist. One example of such a tire belt making system can be found in U.S. Patent No. 7,497,241, the disclosure of which is incorporated herein by reference in its entirety. In the tire belt making system disclosed by the '241 patent, the inspection system 10 is located after the bias cutter. However, this is only exemplarily in nature, and it can be appreciated that the system 10 can be configured to be placed in any desirable location and can be used to inspect any other object that may have surface defects, such as sheets made out of polymers, metals, other composite materials, and the like.

[0024] As shown in Figures 1 and 2, the inspection system 10 includes a frame 12 having a bottom frame portion 14 connected to a top frame portion 16. The top frame portion 16 includes a top cross-member 18. Disposed within the frame is a drum 20. The drum 20 includes a rolling surface 22 and has a first end 24 and a second end 26 that are rotatably connected to the frame 12. This allows the drum 20 to rotate within the frame 12 during the inspection process.

[0025] A motor 28 is configured to impart a rotational force on the drum 20 to allow the drum 20 to rotate within the frame 12. The rotation of the drum 20 translates an object that is placed along the rolling surface 22 from one side of the frame 12 to the other and along to another device. It can be appreciated that the drum 20 may also consist of one or more miniature rollers to accomplish the same task of moving the object relative to the frame 12. The speed of the drum 20 can be controlled by the PLC and loop photo-eyes.

[0026] A feeding track 30, as shown in Figure 1, may also be used to orientate the object, which in this embodiment is a tire belt, prior to coming in contact with the drum 20. The feeding track 30 can consist of a series of rollers 32 and two alignment arms 34 to align the tire belt to ensure that it is properly orientated on the rolling surface 22 for the inspection process.

[0027] As shown in Figure 2, coupled to the cross member 18 are two three- dimensional cameras 36. The three-dimensional cameras 36 are positioned over the drum 20, and specifically the rolling surface 22, so as to inspect the object as it is being passed over the rolling surface 22 beneath the cameras 36. The drum 20 may be calibrated so that it is completely level and perpendicular to the inspection field of the three-dimensional cameras 36. The three-dimensional cameras 36 can measure various parameters of the object, such as its height, thickness (i.e. elevation), and depth in order to detect deformities within the object that is being manufactured or inspected. For example, the three- dimensional cameras 36 can measure parameters such as belt width, belt thickness, offset splice (i.e. dog ears), open splices, splice overlap, and splice thickness of a tire belt during the various stages of the manufacturing process. It also can detect holes and foreign objects that may be embedded within or disposed on the tire belt. One type of three-dimensional camera 36 that can be used with this system is the Sick ICD-3D 100 camera, manufactured by SICK Inc. of Minneapolis, MN. It can be appreciated that other three-dimensional cameras with similar functionality may be used with the present invention.

[0028] Furthermore, it can be appreciated that the location and the number of three- dimensional cameras 36 are application dependent and may vary from application to application. For example, and without limitation, there may only be one three-dimensional camera 36 or more than two, depending on the size of the object to be inspected. For example, in one embodiment used for the tire belt, if the tire belt width is less than 230 mm, only one camera may be required. Two cameras may be used for widths up to 471 mm. Of course, these width dimensions are for one particular application that is using one particular 3-D camera, and the inspection field width of the three-dimensional camera used in the system 10 may vary depending on the type of camera used.

[0029] In addition, the location of the three-dimensional cameras 36 may change depending on the orientation of the surface to be inspected. If the side or bottom surface of the object is to be inspected, then the three-dimensional cameras 36 may be disposed to the side or underneath the object, respectively.

[0030] The functionality of the three-dimensional camera 36 enables inspection of an object in a manner that is not possible by a traditional two-dimensional camera. In addition to the parameters discussed above, the three-dimensional camera 36 may also be used to detect holes, open splices, or tears within the surface of the object that do not penetrate all the way through the object. Such deformities would not be detectable by a two-dimensional camera because they are only detectable by measuring the thickness of the material about an axis that is perpendicular to the surface of the object. [0031] A laser 38 is disposed between the two three-dimensional cameras 36 as shown in Figure 2 and mounted on the cross member 18. The laser 38 is used to illuminate the region of the object that is being scanned or monitored by the three-dimensional cameras 36. In this embodiment, the laser 38 is aligned with the two cameras 36. However, it can be appreciated that the laser 38 may also be positioned at a different location, such as to one side of the three-dimensional cameras 36. The width of the laser projection on the object may be at least as wide as the width of the inspection field generated by the three-dimensional cameras 36. The laser 38 in this embodiment is separate and apart from the two three- dimensional cameras 36. This is because the lasers disposed within the three-dimensional cameras 36 project beams that partially overlap with one another, which results in

measurement errors by the three-dimensional cameras 36. This is because unless the mechanical alignment of the two lasers is precise, the adjacent three-dimensional camera 36 may show a discontinuity in the overlap region.

[0032] By using a third independent laser 38, neither three-dimensional camera 36 sees an overlapped laser line. The data captured by the two three-dimensional cameras 36 from the overlap region captured by the three-dimensional cameras 36 may be manipulated so that the discontinuity is removed. Moreover, by using an independent laser 38, a "Class 2" laser may be used to accomplish the measurement of wider belts at an acceptable resolution and speed. This may not be the case with the laser within the three-dimensional camera 36 because the three-dimensional camera 36 must be farther away to "see" the entire belt width and a stronger (brighter) laser may be needed, which may require eye protection.

[0033] As shown in Figure 3, the laser 38 has a laser beam angle a that is projected onto the object. As mentioned above, the angle a must be wide enough to cover the portion of the object that is being measured or inspected by the three-dimensional cameras 36. It can also be appreciated that more than one laser 38 may be used if there are discrete sections of the object that need to be inspected or measured such that the two laser beams do not overlap with one another.

[0034] An encoder 40, as shown in Figure 2, may be in communication with the three-dimensional cameras 36 so as to retrieve and process the data collected from the three- dimensional cameras 36 for processing by an inspection software 42. The encoder 40 is used to clock image profiles to the camera system 36 in order to build a three-dimensional image of the belt material. This information and other related data can be recorded to log files or to a data acquisition computer.

[0035] The inspection software 42, as shown in Figure 4, is in communication with the encoder 40, which in turn is in communication with the cameras 36. It can be appreciated that the software 42, encoder 40, and cameras 36 may be entirely or partially in wireless communication with one another. It is also contemplated that the three-dimensional cameras 36 may be in direct communication with the inspection software 42.

[0036] One type of inspection software 42 that can be used with the system 10 is IVC Studio 3.2, manufactured by SICK Inc. of Minneapolis, MN. It can be appreciated that other types of software may also be used with the inspection system 10. The inspection software 42 is designed to configure and calibrate the camera(s) to inspect or monitor the

object/product and compare the characteristics of the object/product to parameters that are inputted by a user.

[0037] The three-dimensional cameras 36 rely on precise calibration and alignment in order to function and operate in the intended manner. The software system 42 also includes a calibration feature that enables the three-dimensional cameras 36 to be calibrated prior to use. As shown in Figure 7, the inspection system 10 may include a calibration fixture 48 to aid in the calibration process. The calibration fixture 48 allows the inspection software 42 to calibrate the three-dimensional cameras 36 by leveling the cameras 36 using the laser beams of the three-dimensional cameras 36 with respect to the calibration fixture 48 and thus the drum 20 and rolling surface 22. Once the three-dimensional cameras 36 are leveled, the calibration fixture 48 can be flipped over such that a thin groove is showing. The inspection software 42 can then be used to align the laser beams of the cameras 36 such that they are centered inside the small grove of the calibration fixture 48. The three-dimensional cameras 36 can be adjusted using set screws 50 within the camera brackets 52 that are attached to the cross member 18. Alternatively, the cross member 18 includes slotted holes 54 that enable the entire cross member 18 to be adjusted to calibrate the three- dimensional cameras 36. The three-dimensional cameras 36 can also be calibrated to measure the overall belt width by using a calibration bar. The fixture surface is calibrated by initializing the software 42 to capture the image of the surface.

[0038] Another aspect of calibrating the three-dimensional cameras 36 includes using the lasers built into the three-dimensional cameras 36 to align them to the drum 20. To do so, the laser beams of the three-dimensional cameras 36 are aligned with the laser beam that is generated by the laser 38 in a manner such that the leaser beams of the three-dimensional cameras 36 do not overlap but are collinear with one another. Once all the beams are aligned, the laser beams of the cameras 36 are turned off while the laser 38 remains on and is used during the inspection process. In addition, and to the extent necessary, the drum 20 may also be leveled using jack screws 56. Preferably, the drum 20 is level to the cameras 36 as well such that a portion of the rotating surface 22 is perpendicular to the inspection field generated by the three-dimensional cameras 36.

[0039] Typically, as mentioned above, the inspection system 10 can be used with a tire belt making system. A discussion of one process of cutting and splicing the tire strips to manufacture a tire belt can be found in above -referenced U.S. Patent No. 7,497,241. For example, the system 10 may be placed after the bias cutter of a tire manufacturing system. The inspection system 10 will be positioned at a location to allow it to inspect the tire strips once they have been cut and spliced together.

[0040] A discussion of the operation of the inspection system 10 in the context of inspecting a tire belt follows. However, it can be appreciated that the inspection system 10 can be used for other types of materials and the discussion below is not intended to limit the scope of the present invention.

[0041] As shown in Figure 5, the inspection system 10 is disposed after the cutter 44 in this configuration. The tire belt 46 is positioned onto the feeding track 30 and on top of the rolling surface 22 of the drum 20. The laser 38 projects a laser beam across the width of the tire belt 46 and the two three-dimensional cameras 36 inspect the illuminated portion of the tire belt 46 for any defects. Specifically, the two three-dimensional cameras 36 gather the thickness data of the tire belt 46 and combine it with the encoder feedback 40 to generate a three-dimensional image of the tire belt 46. Once the three dimensional image is generated, dimensional data of the image are compared to the user inputted parameters by the inspection software 42. As discussed above, these parameters include, but are not limited to, the belt width, splice dog-ear (i.e. offset splice), open splices, and splice thickness. In addition, the parameters may also include belt thickness to determine if there are any tears or holes in the tire belt.

[0042] Referring now to Figure 6, a flow chart, designated generally by the numeral 100, is representative of one embodiment of computer readable media tangibly embodying a program of instructions that could be contained in the inspection software 42 or central control unit for inspecting the tire belt 46. The method steps of the software may be programmed to any computer or machine-readable media, and performed by a suitable computer such as a control unit. The process begins when the inspection system 10 is initialized 102. The central control unit may inquire if the inspection software 42 is enabled 104. If not, the central control unit will take no further action. If the software 42 is initialized 102, it will inspect the object, which in this embodiment is the tire belt, to determine whether the inspected section falls within the user specified parameters 106. If the tire belt section 46 falls within the user specified parameters, no action is taken. If the tire belt section 46 falls outside of the specified parameter, the software 42 sends a notification to a user and a command to stop the manufacturing line 108. It can be appreciated that the parameter contemplated may be a single parameter or a host of parameters and that the command to stop the manufacturing line may occur if any one of parameters are violated or if only certain parameters are violated.

[0043] The Inspection System may provide distinct advantages such as ease of use, simplicity of mechanical and alignment and calibration. It may also allow for low cycle times and its ability to be retrofitted with existing systems such that the foot print of the overall equipment is substantially the same (a key advantage for potential customers). The inspection system offers a single solution for all measurement, feature, and defect detection for a variety of types of rubber produced belts and strips.

[0044] In another embodiment, an alternative system 10' is capable of mapping the entire surface of rubber strip and measure multiple features and defects simultaneously using a single measurement system. Features can be defined by, but not limited to: width, gauge, splice height, splice angle, edge linearity, and surface uniformity. Defects can be defined by, but not limited to: "dog-ears" (non-linear edge), poor coverage of rubber/holes in strip material, splice heavy/light, open splice, splice height variance, missing/broken cords, and deviation of surface profile.

[0045] The cameras can be mounted on existing line to eliminate the need for additional floor space while also maintaining existing cycle times. The system is capable of being configured for the defects defined by system and allows the machine to be stopped when any feature or defect goes outside of the user defined tolerance. The operator is alerted using graphical and visual information to make educated decisions about repairs. Data is also logged for records of materials features and defects for further review and evaluation.

[0046] This embodiment does not rely on using a high precision surface to accurately measure the product 2 dimensional or 3 dimensional profile. Other attempted systems require extensive integration, larger machine footprints, and slower speeds due to product contrast. This embodiment relies on multiple high resolution line scan cameras coupled in series to measure a full strip at high speeds in 3D. The system also eliminates difficult mechanical alignments and calibrations from being required as well as high precision surfaces to measure the belt against by using both vision applications algorithms and a precision mechanical mounting fixture.

[0047] Additional detail of this embodiment will be provided below. While the system relied on 3-D cameras, the use of multiple 2D camera systems in addition to width and height sensors to measure features and defects could be used, in addition to having a separate station to measure all features and defects at single or varying stages with 2D and 3D capable measurement systems. A separate station having a precision surface, such as a rotating drum, flat table, or belt to comprise of the background surface. Also, overlapping of the lasers can cause image distortion. This can be overcome by laser frequency shifting or camera offset or polarize filters.

[0048] As shown in Figure 8, the alternative system 10' is placed above an inspection registration surface 51. As the product 41 (or object) traverses beneath the 3D cameras 21, the cameras 21 capture 2D and 3D profile measurements of the product 41. The laser from the camera 21 illuminates the product 41 to allow the cameras 21 to accurately and repeatedly measure the product 41. The beam angle 31 is adjusted such that they do not overlap with one another in this embodiment. Figure 10 shows an example of the alternative system 10' being mounted on a system. In this example, it is a bias cutter.

[0049] It is important to note that the Vision Controller 123 and the Supervisory System 124 (Figure 11) are located in the same hardware platform in this case, though it is not a requirement for the Inspection System, the initial system was developed as such. The Vision Controller 123 refers to the software and communications required for controlling the 3D cameras 21 and analyzing the images for features and defects. The Vision Controller 123 also passes this information on to the Supervisory System 124. The Supervisory System 123 then takes the raw data provided by the Vision Controller 123 and compares it to a set of user defined parameters that an operator entered into the Supervisory System's 124 operator interface. The Supervisory System 123 will then record the results and alert the operator through the HMI interface in addition to alerting the Host Machine 125.

[0050] An encoder 122 on the inspection system is used to provide accurate incremental distance measurements to the Vision Controller 123. As the product 41 traverses beneath the cameras 21, the encoder 122 increments at a specific count related to the linear distance traversed by the product 41. The Vision Controller 123 then uses this information to control the 3D vision cameras 21 and accurately construct an image of known width, length, and elevation. The Vision Controller 123 then takes this information and using a series of algorithms and user defined parameters, analyzes the image for various defects and features such as improper width, gauge, splice height, splice angle, edge linearity, and surface uniformity. The Vision Controller then uses the Supervisory System 124 to speak with the Host Machine 125 and alert the operator and the Host Machine 125 of any relevant information found.

[0051] The flow chart 100 for the Vision Controller 123 and the supervisory system 124 analysis is detailed in Figure 12. As shown in this figure, when the alternative system 10' is in an automatic mode 101, the supervisory system determines if the inspection system is enabled 102. If the system is enabled, the Vision Controller determines if the product is moving 103 and the actual rate it is moving. It uses this information to trigger the 3D cameras 21 and captures and image of the product 104. Once and image has been captured in the buffer 105 the Vision Controller uses various algorithms to analyze features and defects of the product. All raw data found by the Vision Controller 123 is then reported to the Supervisory System 124. The Supervisory System 123 then evaluates the raw data based on a set of parameters defined by the user 107. If the raw data (and features and defects) found by the Vision Controller is outside of the tolerance, the Supervisory System 123 will alert the operator and display the results of the feature(s) or defect(s) that are outside of the tolerance 108. Based on the user defined setting, if the feature(s) and/or defect(s) found require operator involvement for repairs or further review 109, the Supervisory System 123 will alert the Host machine to stop the product 110. The Supervisory System 123 will then record all results and post the information to the host as needed 111.

[0052] If no defects were found from the Vision Controller analysis and all features are within the user defined tolerances, the supervisory system will display the results, record all values of feature(s) and defect(s) 111 and loop back to capture and analyze another image in an infinite loop as long as the host system is in automatic 101, the system is enabled 102, and the product is moving 103.

[0053] A key benefit of the alternative system 10' is the simplicity of calibration for accuracy of feature and defect measurement. The flowchart 130 shown in Figure 13 details the calibration of the Inspection system. When the operator enters calibration mode 131, they are required to install the precision calibration fixture 132 beneath the cameras 21. After it is in place, the calibration is then enabled 133. With the calibration enabled, it will measure the calibration 134 and compensate it based on the known values. These results will then be saved 135 in the Vision Controller 123. The Vision Controller 123 will then use these values to compensate all measurements to provide accuracy of measurement. The calibration mode will then be exited once the procedure 130 is completed.

[0054] While various embodiments of the invention have been described, the invention is not to be restricted except in light of the attached claims and their equivalents. Moreover, the advantages described herein are not necessarily the only advantages of the invention and it is not necessarily expected that every embodiment of the invention will achieve all of the advantages described.

Claims

Claims We claim:

1. An inspection system for inspecting an object comprising:

an inspection registration surface;

a three-dimensional camera to measure the object on the inspection

registration surface;

a vision controller in communication with the three-dimensional camera and applying algorithms to analyze the object based on user defined variables; a supervisory system which communicates to the vision controller, a host machine, and a user via HMI interface in order to properly report, categorize, and record all data captured from the analysis of the object.

2. The inspection system of claim 1, wherein the three-dimensional camera is disposed above the inspection registration surface.

3. The inspection system of claim 1, wherein the three-dimensional camera is configured to measure the width, gauge, splice height, splice height deviation, splice angle, edge linearity, surface coverage, or surface uniformity of an object.

4. The inspection system of claim 1, wherein the vision controller uses a series of

algorithms to analyze each image and measure features and defects all being normalized to the inspection registration surface where applicable.

5. The inspection system of claim 1, wherein the supervisory system notifies the user and the host machine with an alarm when the measurements from the vision controller exceed a user defined parameter or subset of user defined parameters.

6. The inspection system of claim 1, wherein the supervisory system is fully

customizable by the user to determine acceptable levels and machine alarm responses.