WO2014117870A1

WO2014117870A1 - Method, measuring arrangement and system for inspecting a 3-dimensional object

Info

Publication number: WO2014117870A1
Application number: PCT/EP2013/054429
Authority: WO
Inventors: Ján ŽIŽKA; Tomáš KOVACOVSKÝ
Original assignee: Me-Inspection Sk
Priority date: 2013-02-04
Filing date: 2013-03-05
Publication date: 2014-08-07

Abstract

A method for inspecting a 3-dimensional object (5), particularly a vehicle tire or wheel, is described, where said method comprises the steps of: generating a beam of structured light using projecting means (2), said beam of structured light being adaptable regarding at least one characteristics, said beam of structured light being directed onto said 3-dimensional object (5) for illuminating a surface of said 3-dimensional object (5), capturing at least one image of at least parts of said surface using at least one camera (3, 3', 3"), transferring said image/s to computing means (4), calculating a 3-dimensional model of said surface based on said image/s using said computing means (4), and extracting structural information of said 3-dimensional object (5) based on said 3-dimensional model. Additionally an according measuring arrangement (1) and a system comprising the measuring arrangement are disclosed.

Description

METHOD, MEASURING ARRANGEMENT AND SYSTEM

FOR INSPECTING A 3-DIMENSIONAL OBJECT

The present invention relates to a method and a measuring arrangement for inspecting a 3-dimensional object. The present invention further relates to a measuring system comprising said measuring arrangement and a method for calibrating said measuring arrangement.

Optical testing of 3-dimensional object is well established in various technical fields. One important application comprises quality checks of products during or after production process. During optical testing, the 3-dimensional object is illuminated. The light which is reflected by the surface of the 3-dimensional object is detected by a detector. The 3-dimensional structure of the 3-dimensional object is extracted from the measuring signals of the detector and various additional piece of information, such as geometrical arrangement of light source and detector, structure of the illuminating light beam, etc. Problems arise, if the 3- dimensional object has a dark colour which results in poor quality of the measuring signals. Important representative of such 3-dimensional objects are vehicle tires and wheels. Standard solutions use cameras to get information in the form of 2D image. The image consists of 2D position and intensity information. This approach is highly influenced by illumination of the target, as well as target colour and surface. If the image contrast has poor quality, evaluation of it could be very difficult and non accurate, especially in tire industry, where object to inspect is black rubber and makes almost no contrast. Any evaluation of such input can lead into restricted results.

Some improvements can be found in devices with line displacement scanner. In these applications laser scanning technologies are used. A laser line is projected onto the surface of the tire. The image of the line is captured by a video camera under a given viewing angle between laser beam and camera. In EP 2 020 594 A1 the principle of operation is described. The principle is based on triangulation and results in depth information of the surface of the object. By relative movement of the tire or scanner a 3D image of the surface can be obtained.

The disadvantage of this solution is the need of tire rotation, or rotation/movement of line displacement scanner, which greatly increases the energy demands of the device operation. A further disadvantage is the cycle time to perform a complete scan of the surface. The tire has to be clamped in a rotating mechanism, then rotated, and released again after measurement. The time consumption is large and leads to a significant delay in the production cycle of tires. A further problem is that precise mechanic parts are required for the orientation of the laser projector and camera and for the rotation of the tire, whereby the device is composed of precise, cost-intensive parts. A further problem is that due to the poor contrast of the rubber strong lasers are used which require additional safety mechanisms to protect users from laser irradiation.

It is therefore an object of the present invention to improve and further develop a method and a measuring arrangement of the initially described type for measuring 3-dimensional objects with improved measurement results, reduced test time and reduced energy costs. Additionally, a measuring system capable of measuring 3- dimensional objects, such as tires or wheels, should be provided.

In accordance with the invention, the aforementioned object is accomplished by a method comprising the features of claim 1. According to this claim, such a method comprises the steps of:

generating a beam of structured light using projecting means, said beam of structured light being adaptable regarding at least one characteristics, being directed onto said 3-dimensional object for illuminating a surface of said 3-dimensional object,

capturing at least one image of at least parts of said surface using at least one camera,

transferring said image/s to computing means,

calculating a 3-dimensional model of said surface based on said image/s using said computing means, and extracting structural information of said 3-dimensional object based on said 3-dimensional model.

Further, in accordance with the invention, the aforementioned object is accom- plished by a measuring arrangement comprising the features of claim 6. According to this claim, such a measuring arrangement comprises:

projecting means, said projecting means generating a beam of structured light, which is adaptable regarding at least one characteristics, and being configured such that said beam of structured light illuminate a surface of said 3- dimensional object,

at least one camera, said camera/s being configured for capturing an image of at least parts of said surface of said 3-dimensional object, and

computing means, said computing means receiving said image/s and calculating a 3-dimensional model of said surface based on said image/s, said computing means further extracting structural information of said 3-dimensional object based on said 3-dimensional model.

A system for inspecting a 3-dimensional object is provided by claim 13. A method for calibrating a measuring arrangement is provided by claim 15.

According to the invention it has first been recognized that inspection of a 3-dimensional object can be performed substantially without movement of the measuring device. At the method and arrangement according to the invention it should even be avoided to move the 3-dimensional object during measurement. Nevertheless, an accurate and fast inspection of this measured object is possible. This can be achieved by a combination of the usage of structured light, which is projected onto the surface of the 3-dimensional object (i.e. the measured object), and at least one camera. In a first step, a beam of structured light is generated by projecting means. This beam of structured light is used for illuminating a surface of the 3-dimensional object, where this surface is structured, i.e. there are areas or spots of different height on the surface. Due to the structured surface of the 3-dimensional 3- dimensional object, the pattern, which would be generated by the structured light on a plane surface, will be distorted and the light will cause shades on the surface.

According to the invention, the structured light is further adaptable regarding at least one characteristic. By choosing appropriate characteristics, a region of interests on a surface of the 3-dimensional object can be inspected without movement of the measuring arrangement. The changed characteristic may comprise a change of the light pattern created by the beam of structured light and/or a change of colour of the structured light and/or a change of colour of parts of the pattern created by the beam of structured light, etc.

In a second step, at least one image of at least parts of the surface of the 3- dimensional object is captured. In a third step, the captured image is transferred to computing means.

In a fourth step, a 3-dimensional model of the surface is calculated based on the captured image/s using the computing means. The 3-dimensional model might be "low-level", i.e. the 3-dimensional model might be reduced to particular structures of interest or might be low in detail. The 3-dimensional model is preferably adapted to the specific surface to be inspected.

In a fifth step, structural information of said 3-dimensional object is extracted from the 3-dimensional model. The structural information refers to structures which are preferably small compared to the 3-dimensional object. One example of such structural information includes the DOT (U.S Department of Transportation) code, the producer's name, bar codes, colour codes, etc. However, the structural information may also include labelling errors (like missing characters or letters, redundant characters or letters, incorrect tire labels) or production errors (like bulges or dents).

There are several ways how to design and further develop the teaching of the present invention in an advantageous way. To this end it is to be referred to the patent claims subordinate to patent claims 1 , 6 and 13 on the one hand and to the following explanation of preferred embodiments of the invention by way of example, illustrated by the figure on the other hand. In connection with the explanation of the preferred embodiments of the invention by the aid of the figure, generally preferred embodiments and further developments of the teaching will we explained. In the drawing

Fig. 1 shows a lateral view of measuring arrangement according to the invention comprising three cameras,

Fig. 2 shows a front view of the measuring arrangement of Fig. 1 ,

Fig. 3 shows an example of the scanning area of the three cameras of the measuring arrangement of Fig. 1 and

Fig. 4 shows an example of a structure to be inspected by the measuring arrangement.

A measuring arrangement which may scan the complete surface of a 3-dimen- sional object is composed of a data projector (projecting means) and one or more cameras. The number of cameras depends on the size of the 3-dimensional object, the size of the area of interest and the required resolution of the measuring device. The relative position of the data projector and the camera is defined by the angle between the centre line of projection and the centre line of view, which cannot be changed after the scanner calibration. The size of the angle depends on the type of the scanned object surface, the size of the scanned object and the space which is available for the scanner placement.

The principle of measurement is based on triangulation with structured light: The data projector which may be a commercially available video projector projects a pattern of light onto the surface of the object.

During a standard measurement the data projector projects white and black stripes with specific widths and distance between the stripes onto the scanned object. The pattern of stripes is shifted by a specific distance. The number of shifted pattern projected on the scanned object depends on the size of the object, the required resolution and the cycle time which is available.

For example, the number of pattern is at least 4. Then every scanned point on the object is illuminated at least 2 times by a white and 2 times by a black stripe. The width of stripes projected on the scanned object may vary for particular patterns. Instead of stripes it is possible to use other patterns - for instance chessboard, longitudinal stripes, grey code, sinusoidal stripes, or their combination. The pattern projected onto the surface of the object is then observed by at least one camera which is mounted under a certain angle relative to the projector. Using triangulation principle the contour of the surface can be detected. The images of the camera are calculated using a computer (computing means) to obtain depth information of the surface. Standard black-and-white video cameras can be used.

In a further embodiment of identification and measurement requirement of coloured structures on the surface, the projector can be operated in a colour mode to project different colour onto the surface, but still using black-and-white cameras (or monochrome cameras). This results in an enhancement of contrast for coloured structures on the surface. For example, a red dot on a black rubber surface has different contrast when it is illuminated by red or blue light. In the simplest embodiment fully coloured projections are used: For example, first the complete scene is illuminated by red light, then green light and finally blue light. Coloured structures on the black surface then produce differing contrasts within the images of the camera. The images then are combined to a final contrast (black-and-white) image. Taking into account, which image has the highest contrast at which colour, the colour of the structure can be detected even with black-and-white cameras. There are many different marks on a tire for instance full white, red or yellow circle, colour ring, square, triangle, star, semi-arch, or identification and measurement requirements of colour letters on sidewall for instance N, P, 0... 10 and other alphabet letters and measurement of their mutual position, or position towards a specific point on the tire for instance DOT code or barcode. ln both cases - black-white pattern projection as well as colour pattern projection - black-and-white camera is used for image capturing.

In the following, two preferred scanner versions will be described: Version A:

A typical assembly of a surface scanner designed for tire sidewall scanning for the purpose of recognition of small structures on the surface, for instance printing, labels in the size of a DOT code or bigger than size of a DOT code, bulges and dents, is one data projector and 3 cameras.

A first camera with a fast capture rate - for instance 30 frames per second - with a low resolution - sufficient for tire position and dimensions (outside diameter, inside diameter, tire height) - is used in a first step to determine the position of the tire. The first camera recognizes the tire position on the conveyor within the measurement range of the sensor. The field of view must be big enough to ensure that the tire is within this field during measurement. The image from this first camera defines the area to be incorporated in the algorithm for creation of a 3D model when scanning with the other two cameras.

The second and third cameras can be slower, but with a high resolution - for instance 20 megapixel - sufficient for identification and subsequent recognition of the desired surface structure. These cameras use a different field of view on the tire to obtain a complete image of the surface.

The acquired image is used for creation of 3D model of a part of the image (based on first camera scan), necessary for example for identification of printing. The letters of printing in tires have enhanced surface structure and can be detected by the described method of surface scan. Letters on the surface may be: the name of the tire manufacturer, the information of type and size of tire, or the code of production (DOT code), other information imprinted on the sidewall of tires or defects on sidewall of tires, for instance bulges and dents. ln a further aspect, this scanner enables identification and measurement of colour marking on the tire by using coloured patterns to be projected onto the surface still using black-and-white cameras.

Version B:

An assembly of a scanner with lower resolution, but higher speed which is designed for scanning and recognition of tire label on the sidewall is a data projector and 2 fast cameras with a resolution sufficient for identification and subsequent recognition of tire name and type. The scanner speed is at least 30 fps and resolution is 4Mpix. The image acquisition takes place in one step; subsequently the 3D model intended for further processing is being formed from acquired image.

This version enables identification and measurement of colour marking on the tire.

The general process of tire scanning on conveyor is described below:

1. The tire is positioned within the measurement range of the scanner. Tire transport is independent from the measurement. The tire transport can be ensured manually, on a conveyor belt or roller conveyor as well as with a robot. Tire can be with or without a rim, inflated or not.

2. The tire is not allowed to move before and during the scanning.

3. Based on the signal indicating that the tire is standing still in the measurement range of the scanner, the scanning sequence according to the scanner type - Version A or Version B is started. 4. After finalization of the scanning process, the scanner sends information to the conveyor operating system that the conveyor can transport the tire out of the measuring range. 5. During conveyor movement - departure of the scanned tire and arrival of a new one into the measurement range of the scanner - the result of the previous measurement is calculated.

6. Scanning of current tire image and measurement results calculation of the previous tire can be realized simultaneously.

The surface scanner of the present invention is a very compact system which can be placed on various machines and devices such as uniformity testers, lines for geometry measurement and other devices for handling and testing of finished tires. A further application can be the inspection of tires in a car production line: At the end of the production line the tires can be controlled for correct size, type, manufacturer and rotation of tire. The result of calculating the captured images in step 3-5 is a 3D-model of the surface of the tire. From the 3D-model various information of the tire can be obtained:

DO T code reading

A typical application for the A version scanner is identification, recognition and reading of the DOT code on the tire. When applying a standard scanning process for the A version sensor and standard scanning process described above a section of the surface, which contains the DOT code, is selected. The 3D-infor- mation of the highlighted letters which represent the DOT-code is then transferred into 2D black-and-white information. Using standard OCR (optical character recognition) the letters of the DOT-code can be transferred into computer-readable code, for instance ASCII. Based on the general rules for character coding to DOT code with an output from OCR, the system reads the DOT code. The read DOT code may be used for the conveyor system to transport the scanned tire.

Tire Identification

Typical application for the B version scanner is identification and recognition of scanned tire. This system is used in establishments, where bar codes are not used for tire identification, or in establishments which hold tires from various producers. The system recognizes the producer of the tire, for instance Continental, Pirelli, and the tire type for instance 155/80 R13 79 T. For recognition of tire producer and type a comparison of the scanned 3D model according to the description above with the defined model is used. OCR utilization is in this case complicated, as the tire name and type do not have a homogenous character form and vary even at different tire types from one producer.

Error detection on tires

Another typical application for the A version scanner is identification, recognition and error classification on tire sidewall for instance missing character, letter or redundant character, letter or incorrect tire labelling, bulges and dents. Using the scanner, it is possible to check outside tire dimensions and tire labelling correctness, i.e. whether the tire labelling corresponds with the outside tire dimensions. This application involves a combination of application "DOT code reading" and "Tire identification" and is extended to scanning of tire sidewall printing and error classification. Detection of marking quality

Depending on the desired accuracy of marking error, detection A version or B version of the scanner is used. Compared to the application described above scanning can be extended to colour patterns, which extends the scanning time slightly. The output of such a system is label position in relation to the reference point (BarCode, DotCode), the label colour, the definition of compliance or noncompliance with the specified label (colour, size, label type), mutual label position, mutual position of labels and tire bead. Application of marking quality detection can be combined with applications described above.

Advantages of the measurement according to the invention

• Reduction in energy costs compared to systems with line displacement sensors, which require a complex mechanics and tire rotation

· No rotation of tires required

• Significant reduction of machine tact for DOT code reading (ca. 60%)

• Significant reduction of initial investment (ca. 60%)

• Data projector not using 3B class lasers- increase of safety

• Significant reduction of space required for device installation

Calibration process:

Initial scanner calibration is carried out in several steps - order can be changed. For a routine calibration the steps can be merged.

1. Calibration of camera objectives and/or lenses (Scanner version A,B)

2. Calibration of data projector optics (Scanner version A,B)

3. Calibration of first camera position and data projector (Scanner version A,B)

4. Calibration of second camera position and data projector (Scanner version A,B)

5. Calibration of third camera position and data projector (Scanner version A) 6. Calibration of first and second camera position (Scanner version A,B)

7. Calibration of first and third camera position (Scanner version A)

For a standard calibration steps 1 - 3 are not applied. Steps 4 -7 can be merged into one step.

Fig. 1 shows an embodiment of measurement arrangement according to the invention (A version scanner). The scanner 1 is a compact unit which contains projecting means 2 (data projector), a first, fast low-resolution camera 3, two high- resolution cameras 3', 3", and computing means 4 (computer). The scanner is placed above a production line of tires. A 3-dimensional object 5 (tire) is transported on actuating means 6 (conveyor belt). During a scan the tire has to be stopped. The first camera 3 detects the position of the tire 5 on the conveyor 6 which can vary within the scanning area 7. The image of the projector 2 is not focused onto the tire; a defocused configuration is used instead. The reason is that the focal plane of the projector 2, which is perpendicular to the direction of projection, is not in plane with the surface of the tire due to the tilted mounting position of the projector. Therefore focusing onto the tire is not possible over the whole surface. In this case a defocused configuration gives better results of the images. It may be mentioned that in a different geometrical configuration focusing of the pattern onto the surface may be more convenient. The measurement range of the scanner 1 is 900 mm x 900 mm x 320 mm (width x length x height) in this configuration. Fig. 2 shows the scanner 1 of Fig. 1 in front view. The measurement arrangement is located in a housing. The housing comprises a window 8 through which the structured light may leave the housing. Behind the window 8, the video projector 2 is located. Fig. 3 shows a typical scene of the scanning process of a tire. The tire 2 is transported by a conveyor belt (not shown in figure) into the scanning area 7. The video projector 2 illuminates the whole scene using structured light. The first camera 3 takes an image of the scene with a given size 9 which depends on the configuration of the scanner, e.g. the x- and y- number of the camera pixels, the focusing properties of the camera objective and the distance of the camera to the object. The area of the image has to be larger than the position which the tire can have during the scanning process. With this first image of the tire the position of the tire (in a coordinate system of the scanner) is detected. The second 3' and third 3" camera has the same imaging area like the first camera. With the position information from the first camera 3, however, the area of interest 10, 1 1 of the second 3' and third camera 3" can be determined. The area of interest 10, 1 1 is smaller than the whole imaging area, but large enough to cover the area of the tire. The reduction of the area of interest results in a much faster calculation of the 3D-model due to the fact, that only a small area (and number of pixels) must be calculated.

Fig. 4 shows typical image resulting of a scanning process. The depth information which comes out of the 3D-model of the tire surface is transformed into a black- and-white information. The letters on the sidewall of the tire can then be detected using standard OCR-technique

Many modifications and other embodiments of the invention set forth herein will come to mind the one skilled in the art to which the invention pertains having the benefit of the teachings presented in the foregoing description and the associated drawings. Therefore, it is to be understood that the invention is not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

L i st of ref e re n ce n u m e ra l s

1 measurement arrangement

2 projecting means

3 camera

4 computing means

5 3-dimensional object (tire)

6 actuating means (conveyor belt)

7 scanning area

8 window

9 captured area of the first camera

10 area of interest (second camera)

11 area of interest (third camera)

Claims

C l a i m s

1. Method for inspecting a 3-dimensional object, particularly a vehicle tire or wheel, comprising the steps of:

generating a beam of structured light using projecting means (2), said beam of structured light being adaptable regarding at least one characteristics, said beam of structured light being directed onto said 3-dimensional object (5) for illuminating a surface of said 3-dimensional object (5),

capturing at least one image of at least parts of said surface using at least one camera (3, 3', 3"),

transferring said image/s to computing means (4),

calculating a 3-dimensional model of said surface based on said image/s using said computing means (4), and

extracting structural information of said 3-dimensional object (5) based on said 3-dimensional model.

2. Method according to claim 1 , said structured light projecting a pattern of light onto a surface, wherein said pattern of light comprises a periodical structure, preferably black-and-white strips, chessboard, grey code and/or sinusoidal stripes.

3. Method according to claims 1 or 2, said step of generating a beam of structured light and capturing at least one image are iterated at least twice, wherein at least one characteristics of said structured is changed between two succeeding iterations.

4. Method according to one of claims 1 to 3, said characteristics comprises the colour of said structured light and/or said pattern of light.

5. Method according to one of claims 1 to 4, said step of extracting structural information comprises the steps of:

transforming said 3-dimensional model into a 2-dimensional black-and- white image (monochrome image) and

performing OCR - Optical Character Recognition - upon said black-and- white image (monochrome image).

6. Measuring arrangement for inspecting a 3-dimensional object, particularly a vehicle tire or wheel, comprising:

projecting means (2), said projecting means (2) generating a beam of structured light, which is adaptable regarding at least one characteristics, and being configured such that said beam of structured light illuminate a surface of said 3-dimensional object (5),

at least one camera (3, 3', 3"), said camera/s being configured for capturing an image of at least parts of said surface of said 3-dimensional object (5), and computing means (4), said computing (4) means receiving said image/s and calculating a 3-dimensional model of said surface based on said image/s, said computing means (4) further extracting structural information of said 3-dimensional object based on said 3-dimensional model.

7. Measuring arrangement according to claim 6, said projecting means (2) being capable of generating structured light of different colours and/or different pattern of light.

8. Measuring arrangement according to claim 6 or 7, said projecting means (2) comprising a data projector, preferably a video projector.

9. Measuring arrangement according to one of claims 6 to 8, said measuring arrangement (1 ) being configured to extract structural information which are small relative to the 3-dimensional object (5).

10. Measuring arrangement according to one of claims 6 to 9, wherein said structural information comprises letters, digits, bulges and/or dents.

1 1. Measuring arrangement according to one of claims 6 to 10, said camera/s (3, 3', 3") being a black-and-white camera (monochrome camera).

12. Measuring arrangement according to one of claims 6 to 1 1 , said measuring arrangement (1 ) being configured such that during capturing of a current 3-dimensional object (5) said 3-dimensional model of a previous 3-dimensional object is calculated and/or said structural information of a previous 3-dimensional object is extracted.

13. System for inspecting a 3-dimensional object, particularly a vehicle tire or wheel, comprising:

a measuring arrangement (1 ) according to one of claims 6 to 12, said measuring arrangement (1 ) being capable of measuring said 3-dimensional object within a scanning area (7),

actuating means (6), said actuating means (6) moving said 3-dimensional object into the scanning area,

controlling means, said controlling means being configured for controlling said actuating means (6),

wherein said computing means (4) being capable to send control instructions to said controlling means for triggering movement of said 3-dimensional object (5).

14. System according to claim 13, said actuating means (6) comprising a conveyor or a robot.

15. Method for calibrating a measuring arrangement according to one of claims 6 to 12, said measuring arrangement (1) comprising at least two camera (3, 3', 3"), said method comprising the steps of:

Calibration of objectives and/or lenses of said cameras (3, 3', 3"),

Calibration of optics of said projecting means (2),

Calibration of position of cameras (3, 3', 3") relative to projecting means (2), and

Calibration of position of said cameras (3, 3', 3") relative to each other.