WO2001088473A1

WO2001088473A1 - Inspection of a three-dimensional surface structure and the calibration of its resolution (solder paste deposit) using a camera, an optical sensor and an independent calibration mark

Info

Publication number: WO2001088473A1
Application number: PCT/DE2000/001563
Authority: WO
Inventors: Hubert Bellm; Ludwig Listl
Original assignee: Siemens Aktiengesellschaft
Priority date: 2000-05-17
Filing date: 2000-05-17
Publication date: 2001-11-22

Abstract

The invention relates to a device for inspecting a three-dimensional surface structure and to a calibration method using a camera (8) for recording a two-dimensional image and an optical sensor (7) for the three-dimensional measuring of a characteristic to be inspected. The camera (8) and optical sensor (7) are integrated in a sensor unit (6). A calibration mark (16), which is fixed to the inspection device independently of the test specimen (3) is used to calibrate the position and/or resolution of the camera (8) and of the optical sensor (7). This allows a programme-controlled calibration without having to reset the test specimen or inspection device.

Description

description

INSPECTION OF A THREE DIMENSIONAL SURFACE STRUCTURE AND CALIBRATION OF THE

RESOLUTION (SOLDER PASTE APPLICATION) WITH CAMERA AND OPTICAL SENSOR AND INDEPENDENT

calibration mark

The invention relates to a device for the inspection of a three-dimensional surface structure of an essentially flat test object according to the preamble of claim 1 and to a method for calibrating such an inspection device according to the preamble of claim 4.

From US Pat. No. 5,751,910 an inspection device for the application of solder paste on printed circuit boards is already known, in which the images recorded with a CCD camera are processed. Printed circuit boards for surface-mountable components, so-called printed circuit boards in surface mount technology (SMT), are produced in large quantities and in many variants. The components that can be surface connected are mechanically held on the circuit board by soldering their connections to metallized surfaces, so-called pads, and electrically connected to conductor tracks on the circuit board. For this purpose, a pattern of metallic pads is provided on the circuit board, which corresponds to the position of the connections of the component. Solder paste is applied to the pads using a screen printing process. The surface-mountable component is then fitted onto the circuit board. First, the component is held on the circuit board by the adhesive properties of the solder paste. After a subsequent heating of the printed circuit board, the connections of the component are soldered to the pads.

A trend towards ever increasing integration of components with a growing number of connections per component housing can be seen in assembly technology. In the so-called

Fine pitch range, the distance between two adjacent connections of a component is about 1 / 40th Inch. This also makes the pads on the printed circuit boards smaller and denser. About 80% of the soldering errors in the fine pitch area are caused by the solder paste printing. Examples of such errors are insufficient solder paste application or short circuits between adjacent pads due to inaccurate placement of the stencil during solder paste printing. In order to find these defects and to identify weaknesses in the manufacturing process, an optical inspection of the circuit board is carried out after the solder paste application. For this purpose, the above-mentioned American patent describes a device with a camera for taking an image of a solder deposit to be inspected. With the help of a neural network, three-dimensional features of the solder deposit are extracted from the two-dimensional image recorded during inclined lighting and cross-polarization of the light.

In order to be able to calculate the position of the solder deposits from the images, the position of the printed circuit board must first be recorded in the inspection device. For this purpose, the camera is moved with a positioning device in succession over two alignment marks. The alignment marks can be designed, for example, as round metallic pads or as printed cross marks. They are expediently arranged in the region of the corners of the printed circuit board. The images of the marks recorded with the camera are evaluated in a computer with a suitable image evaluation program and the position of the marks is calculated. Information about a possible shift or rotation of the circuit board is obtained from the position of the marks. On the basis of this information, the camera can be positioned over the desired group of solder paste depots, which corresponds, for example, to the connections of a component, and an image can be recorded for further assessment of the quality of the solder paste application.

However, the two-dimensional image processing method has the disadvantage that, in particular, deviations in the height of the solder deposits from a predetermined target value are only inaccurately can be averaged. This can result in a connection of a component not resting on the solder paste, for example if the solder paste application is too low, and in the subsequent soldering process no electrical connection between the component connection and the pad on the printed circuit board being established.

An optical sensor for three-dimensional detection of surfaces is known from DE 196 08 468 C2. The optical sensor described there, which works as a distance sensor based on the confocal optical imaging principle, enables automatic surface inspection with a high data rate at low system costs. With regard to the structure and further advantages of the known optical sensor, reference is made to the above-mentioned German patent.

The object of the invention is to create a device for inspecting a three-dimensional surface structure and to find a method for calibrating such an inspection device, by means of which the accuracy of the measurement of the three-dimensional surface structure is further improved.

To achieve this object, the new device for inspecting a three-dimensional surface structure of the type mentioned at the outset has the features specified in the characterizing part of claim 1. To this end, the new method for calibrating such an inspection device comprises the steps mentioned in the characterizing part of claim 4. Advantageous developments of the invention are described in the dependent claims.

The invention has the advantage that the volume and position of the solder paste application can be measured exactly. In the fine pitch area in particular, the solder volume is a very important process parameter, with the knowledge of which systematic process weaknesses can be eliminated. About a pure good / bad off In addition, measurement results are determined with the optical sensor on the geometric shape of the solder paste application. This data can be used directly for process improvement and is therefore of great benefit to the user. A calibration mark that is permanently connected to the inspection device has the advantage that it is independent of the test object. The calibration therefore does not impose any additional requirements on the quality of the marks on the test object, which are provided for example on printed circuit boards for determining the printed circuit board position with a camera. As before, two white crosses on the circuit board material can be approached with the CCD camera and the position of the circuit board on the inspection device can be measured using standard image analysis. With each conversion of the inspection device, for example a conversion to a new type of circuit board as a test object, a calibration must be carried out again. When calibrating with calibration marks on the circuit board, a special circuit board would have to be used as a calibration board. For each calibration process of the inspection device, the calibration board would have to be converted before a manual calibration process with manual approach to the marks could be carried out. This would be associated with the disadvantage of a considerable expenditure of time and money, which can be avoided with a calibration mark which is permanently attached to the inspection device and which is always in the same place and can in any case be approached automatically. The duration of a calibration process is only about 10 s. After completing the calibration process, the relative position of the optical sensor to the camera can be calculated exactly. Thus, not only can the camera be positioned exactly over the features to be inspected after measuring alignment marks on the test object, but also the optical sensor.

Since the calibration mark is firmly connected to the inspection device, its position changes when the DUT or a conversion to another DUT not. The mark can thus be approached automatically in any case and the individual steps of the calibration can be carried out automatically with the aid of a control computer which evaluates the image signals of the CCD camera and the optical sensor using image evaluation software. The same control computer can also be used to control the travel axes of a positioning device.

In the case of a roughly known resolution of the CCD camera, the alignment marks on a printed circuit board as a test object only have to be recorded in the image area. The position of the marks in the coordinate system of the inspection device can be calculated from the offset in the recorded image and the resolution of the image.

If the calibration mark, which is firmly connected to the inspection device, is such that it provides both a contrast in the gray-scale image of the camera and in the height image of the optical sensor, then advantageously only one mark is sufficient to calibrate both recording systems. In addition, due to the use of only one mark for both recording systems, the relative position of the optical sensor to the camera can be measured exactly. Compared to a separate measurement with different marks, falsification of the measurement result due to inaccuracies in the relative position of the marks to one another is advantageously avoided.

A calibration mark, which is designed as a circular cylinder standing on a plane, the base of which is aligned essentially parallel to the surface of the test specimen and the upper circular surface of which is treated in such a way that its gray value differs significantly from the gray value of the plane, has the advantage that it can be produced with little effort and represents a simple shape primitive, which can be measured with little computing effort using a stored model is correlated and, because of the contrasting edges, allows precise measurement both by the CCD camera and by the optical sensor.

An additional improvement in the accuracy in the measurement of the resolution of the CCD camera and the optical sensor can advantageously be achieved by taking at least two images each time, the respective recording positions being offset from one another in a plane parallel to the plane of the test object. The resolution of the CCD camera or the optical sensor is calculated exactly on the basis of the offset of the images of the calibration mark.

Based on the drawings, in which an embodiment of the invention is shown, the invention and embodiments and advantages are explained in more detail below.

Show it:

FIG. 1 shows a basic illustration of an inspection device,

FIG. 2 shows a plan view of the working area of an inspection device,

Figures 3A and 3B images of the CCD camera to determine the

Resolutions in the X and Y directions, FIGS. 4A and 4B images of the optical sensor for determining the resolution in the X direction and

Figures 5A and 5B images of the optical sensor for determining the resolution in the Y direction.

In Figure 1 is an inspection device for optical

Testing of printed circuit boards shown, which can be integrated into a printed circuit board production line. The unchecked circuit boards are fed to the inspection device with a conveyor belt 1. The circuit board to be inspected is held in an inspection position with a tiltable stopper 2, the hidden parts of which are drawn with broken lines. After the inspection (-0 CO t to P ^{1 1}

Üi o Cπ o Cπ o Cπ o ω φ P ^J rt ^ 01 P- D tr ¹ Φ P- er ^ Oi P- oi rr α s: H oi 13 Φ rt P- Φ TJ rt

TJ φ P- P- P- P-: rt 01 P Φ P- P Φ Φ P- P- r + P P- O H P- O P- P H P P- P "P-

Φ P "P Φ P O P" Φ rt- Hi P- P oi 3 H o Φ Φ P Φ H P O ι-i P PJ P P o

W Φ li er ι PHP • rf Φ TS P- φ P ^J P ^J IQ P- PP ^J rt Φ P P- H rt P rf P P- P- P- P Φ N Φ H p- φ li PN ri≡ s Φ 01 Φ P- rt

P- o ι> φ P ι-i Pi N er> Φ 01 rt 3

≤ H P> φ Φ 01 o P Ό P- l-i V -i π Φ = ε

O tr ι-i iQ P P Ό ^ P- P- 03 P- P- H P- t O P- rt ≥i O - P-

P rt J tr Φ o P P- P ¹ p: φ> q 00 PO Φ P 01 ι-i O rt P TJ ι-1 tsi rt CO P

01 P- PO Φ P- D ιQ & CD P 'H Φ? φ P ^J O α Φ rt tr Φ rt ii rt φ P P-

Φ φ P 0) NP o Φ> rt -i Φ PN o PP ¹ P 3 NJ P- P- P ≤

P- ιq P- p: t- P ^J P ¹ ι-i O ri- PPPP Ul P ri ^" 3 P- Φ Φ X Φ O ιq P- α

P 01 rt yQ P- 01 (-0 P 'Φ P P- r_o 01 Hi φ PS, P- P- rt p- P- P P- H r ii Φ p. O P- P- H o φ tsi ω P yQ Φ rt CD o P yQ P- rt Φ rt rt P 01 rt P tr Q. H

P- P 01 O P- tr P φ φ J ω P Φ s? R 01 01 P H Φ N • P Φ

O P- rr P o rt- C0 P- PPDP o φ t ^| χ) P ¹ s; P s: PH 3 co

P ^J rt- P- - P W3 Φ o Hl OS Φ H er P ^J Hi s 01 Φ Φ 01 φ ιq P P- rt rt P φ PPP ^J P ¹ PP ^J φ Φ P Φ Φ Φ P- Φ DJ ιq li P ii P rt o

P P 3 P L iQ tr 01 P o CD 01 H P yQ. 01 P- P ι P- 01 3 φ s P 01 J

PP er φ IN o P- P Φ φ P Φ P ¹ O p- P ^ p Φ p- Φ TJ ι CD S! CD ι-is: HPP 3 3 Hi K tf P φ 'rt φ φ P Ω P P- φ er CD li 3 φ ^ q P φ to J Φ fu P- φ P ^J PP P- p- HP tr PPH

<! p- iQ X P- P- l-i E? P- P- Hi P φ oi rt ι: rt Φ

Φ Φ 01 * \ rt P> PP P- P- Φ rr ri ri- PX φ P- rt P- Φ er P to ι-i PP P- σ t p- P ¹ 01 P tr ¹ O φ P- SPP uq Hi C 3 er iQ P 1 α Φ o ¹ rr Ό Φ 01 01 Λ P- φ w Φ P- PP P- P P- P3 P

P P- P ¹ φ P P- lT "XO P- P-? P 3 P o 3 PP & o n H rt HP

P ιq tO O ri P ri- r - P- P 1 oi ri- P o P o U3 rt ¹ ιq PO

O φ P- PP! Λ p- Φ P- • ϋ P- TJ HP ¹ ^ 01 Φ Φ Φ P -

<! φ N p. P <- Hi P- ri- ι-i P cj Hi P-, Φ P- P i-i 01 P- 01

P o P- P P-. P ¹ ->CD> o P- O CU er Φ ι- <; P- P- yQ P- P Ξ rt PTP -J

0 P fJ P- w P ¹ H 1 φ ⁴ O ö P ≤ Φ P- g P rt PP Φ Φ Φ o P

P- φ P P- S ιq ϊö P- ri- P tu 01 φ li P ^J P Φ ι-i p- P- ^ 3 ι-i rt oi "XI ^ P ¹ er CD Φ P- PP P- rf 3 ιq J rt H φ ιq 01 P ¹ rt Φ rt π J! CD rr Hl P- <qoo Φ P φ P- Φ o P- P Φ OP ¹ Φ φ rt H er P

• p: CD Φ Φ Φ r + P '3 ιq H Φ rt Hi P' φ s q P er 3 Φ H P

P P- P & P Φ rr er p: φ H Φ P PJ Φ Φ 3 PPP <! α P- er C Φ P 21 D ω P_ P ¹ li P - rt P s: p- φ P Φ

P- P- er li Φ P P> P CD φ H p- li tr ¹ P Φ φ rr Hl P ^* CO oi ii φ P li P- Φ P t r + ^ q ιq ι-i N φ rr φ φ P- PH Φ t 01 H Φ TJ JS- 01 i P- Φ 01 01

Φ φ Hi P- P P- P- ri- P CL rt PJ φ P o Φ ϊ * Φ ii CD OOP <l P CD 01 li c P- P r + N Φ PH P P- • 3 ι-i H li PP

P cυ P P rt H 3 P CD <Φ p- rr φ 01 Φ φ rt P P Φ Φ er rt P t

3 P P- Φ P- & P o P & H H H P rt- o i-i P- • 01 rt α 3 φ er 01 rr

P- Hl C ιQ <P- rr Hl PH r + Hi t Φ P 01 ^• σ O ιq * Ü N σ Φ P- tr ¹ P- P er Φ li P CD o -j CD P? oi o ri- rf- α φ Φ JH rt φ rt PP n 01 WN - Φ P- V Φ P ^J P- 'UJOHP P- Φ PH p- P Φ P

P- rt Φ Φ P φ P- Φ li P- Φ HP s: P Φ Φ P 01 r 01 li rt 01 er Φ rt PH φ P Φ P- P Φ P Φ er er ιq Φ PW P- n ri- Φ P- <! rt Φ P- Φ tq Cπ

P> 3 P ¹ P- ri- Φ l_l. P φ P- 01 D φ H r + P - Φ PT) o Φ Φ 01 P ii 01 l &

3 P-CT> IP φ. 3 P 3 ii O & oi P- P ^J Φ Φ ϊ H, P ι-i P "oi Φ • ö oi rsi P P-

CD rf PC Tl _J z Φ o ii Φ P- O Φ P- H> H Φ oi NP "Φ 3 P ¹ rt P oi φ li 01 PO φ tr ¹ CΛ rf ι P Φ P P- P φ P- Φ P- P li PJ P TJ

? O TJ P 01 P- CD r + tr ¹ O φ P • rt- Φ m er o P ¹ P- o P Φ 1-3 rt rt ≤ Φ l →

Φ Φ -i φ Φ P P- P "^ q Φ 0 ⁾ HP ^J Φ ςu Q. Φ Φ P" P tr ιq P ι-i rt P- Φ Λ 'φ li ^ q? V p. α 1 P- Φ P ^ q Φ CD rr α P Φ σi o P Hi rf ty li rt P Φ O P- rt P-

P ¹ Φ rr iQ 1 Φ Φ PQ P- Hi ι-io 01 P- g p- Φ Φ cf PPP rt P- rt Tl I P- tr T) Φ 1-1 1 Φ φ p: Ö o P P- n rt P- er Oi Φ Φ o Φ

P O Φ H P 1 ι-i N er 1 H P t P Φ 1 P- P i-i P ι-i

P- I P 1 p: Φ rt- Φ P φ Φ 1 Φ ι-i P Φ o 1

01 01 Hl li H li * - ~ J rt 01 1 P rt 1 I _* 1

J co to t P ¹ P>

Cπ o Cπ o Cπ o Cπ

Working segment 24 is a panel 25 on which circuit boards, not shown in FIG. 2, are arranged as test specimens. A panel coordinate system with axes XN and YN, which are also marked as arrows, is defined by the position of the layout of the circuit board on the panel. The user coordinate system is shifted in position with respect to the machine coordinate system. The offset of the panel has the coordinates XMO and YMO. With the width dimension X and the length dimension Y of the panel, the panel coordinates XN and YN of a point P can be converted into machine coordinates using the following formulas:

XM = DimensionX - XN + XMO and YM = YN + YMO.

With a positioning device, not shown in FIG. 2, the sensor unit 21 in the inspection device is moved in the direction of the axis XM, the panel 25 in the direction of the axis YM. Overall, the sensor unit 21 can thus be positioned relative to the panel 25 in the X and Y directions. In the snapshot shown in FIG. 2, the optical sensor 22 is just above a calibration mark 27. If the calibration mark 27 is exactly centered in the image of the optical sensor 22, the coordinates SensPosX and SensPosY are determined. The machine coordinates XM and YM are converted into the axis coordinates X-axis and Y-axis of the positioning device according to the formulas:

X axis = XM + SensposX = DimensionX - XN + XMO + SensposX and

Y axis = SensposY - YM = SensposY - YN - YMO.

The camera 23 is positioned over the calibration mark 27 in a corresponding manner. Coordinates CamPosX and CamPosY of the camera 23 with a centered image of the calibration mark 27 and the respective conversion formulas for the conversion into the co CO to to P ¹ P ¹

Cπ O Cπ o Cπ o cπ w P ^ P>

P P- P Ω

3 P P tr

Φ PL P ^J o ü φ o Φ

P ι-i ιq P

03 V

00 s: o

P- P- o oi 03 φ l-i

P- rt PL

P er P-

PL yQ Φ P φ P- P

01 3 rt o Ω Φ l-i tr o P

N Z TJ

P pj: rt PL

<q H P ro

01 N oι H

Z rt Ω

Φ • tr TJ p- Φ o oi α P 01 φ Φ P-

H co rt ω Φ P-

O o P O

TJ oi PP ¹

Z o rt o P- o t P- i-i φ

01 H

Ω t Φ

P- tr to P-

P Φ • P

H

X o P-

1 φ Ω

P tr

P 03 rt

P ^J o P oi HP ιq

P -J

P φ

Ω P ii tr P ^y Q

PL φ

P- he

P ex Φ

P- P

<- <φ

1 01 u o P-

P- o Ω

Ω α tr t /

1

co co t ro P ¹ »

Cπ o Cπ o Cπ o Cπ

ιq 03 Cd Z P- SV co? 1 V o s: 3 t ^ X P ≤ Z PL PL t ö H 11 P oi 3 03 3 P- O P- er z rt

Φ P- er Φ PO: φ PO: TJ Φ Φ PP 1 P ¹ P Φ φ P φ P P- φ φ P rt φ Ω P 03 TJ 3 ii P- the P Φ ι-i PL PPH "P rt li oi P ¹ 3 03 ιq P ii 3 1-1 P ^ Φ PP Ω p- oi tr H rt rt P ι-i P

P- PL P PL Φ P 03 P- P P- PL 01 P- Φ d Φ P Φ P- PL tr Φ oi Φ V •> _* P- to Ω P υ3 p- Φ Φ 3 φ o er φ 01 φ φ er li P SV P φ li Cd er <d φ H Φ P φ 03 ö tr

P P ι-i l-i P Ω P P H P PL o rt P P P- ι-i Φ P P PL 3 P Φ Ω 1 rt <er

Φ α P-. LJ. 03 p- tr P- PP ¹ 01 P- 1-1 PL φ P co P ¹ P- tr Q • o Φ rt P- φ φ PL φ P Φ PL Φ co P- Hi to Z rt 00 o. Φ 3 s P ii Z Φ σ> P Φ H ii N

Φ φ 03 s: §; P- -J H P P P- l-i P o Φ H-. 3 H Φ PL φ SV Φ P φ P P α rt p:

P o Φ φ 3 tr Φ d 3 Φ V P H Φ P- d 03 P- ι-i o φ H oi SV P d P- Φ ι

> TJ er P- d P P o oi P P- κ | P P P. i-i PL rt P 03 φ V TJ P PL o P o er Φ P- P-

O d H Φ P ^J P ι-i P φ s HP 1 P ^J yQ Φ 3 P rt yQ P 01 3 rt Φ li P tr φ P- P ¹ P- J Hi d: P- 03 φ PL SV PL PP PL SV * f φ φ PP 3 Φ P φ P P- PP 01 P o P 03 tr Ω rt H "Hi P Φ 01 03 φ Φ Cd P- oi er li P- P" X ιq 03 H "01 d tr 03 PL o P tr

P- O: L N P PL p. o Ω l-i P- Ω SV SV rt J \ φ Ω Hl cn Φ o Hl

03 03 P- P- Ξ yQ φ P> Φ H tr P> tr S P O: φ d; PL <P tr P O P H PL er rt PL

Ω d P Φ Φ ii li CT) ri oi P- tr ¹ σ> PL rt P- Ω PP PL P Φ φ _"* Φ Φ 3 N 3 oi Φ Φ φ tr P yQ P- O: PP dv tr P P - V er H li P P- d P- Jjd ii 03 P 1-1

Φ yQ 03> & α t er Φ yQ s: P P l-i rt ro P φ rt Φ 3 er P P H rt er o P-

P 03 d Cd φ o Φ Φ P Φ Φ d yQ o Φ PP 03 Φ N «φ φ co φ P er o ¹ Φ

Z TJ Hi P- P ö PV d SV li Hi oi • P Φ Φ P 03 P- φ 3 Ω n <i P- o PP to Φ PPP ^J 1 PPP o PL PL yQ yQ V Ά PL P- 03 03 P tr no PL ö Hi WP "φ HHP P. PL vq P PL o Φ φ φ φ o P Cd PP Φ Ω P- φ TJ 01 TJ ö li PL Φ 1 Hi P P-

P rt P tr φ φ P H P l-i ι-i P P 03 TJ φ φ 01 tr er P P- o l-i ro 1 rt P l-i W φ er

03 Φ 3 H i-i 3 O: rt PL L O Ω P- P- t-i 01 P t-i Φ l-i p: P- ^ Φ P P P H- H

O Φ φ P φ φ P- *, PL 3 P P er O Φ P- N Hi P P P- Lq P 3 φ er P-

HP, Φ J P.> ii Φ P li PP dp PP li PL N Α J rt Φ d 03 -J ro 3 P ¹ 01 P ¹ Φ li Φ

01 Φ o Φ d P P P P "l-i Φ P 01 P- Φ P O rt i-i z P- H Φ tr SV 01 H i P-

03 Φ 03 H Hi <q o rt P- Ω P rt yQ Φ H H P- Z 3 SV φ 01 P l-i P o P P Φ 3

- 1 P P- P ¹ oo Hi φ π Φ er tr • φ H P- li 01 p- PO: P- φ yQ MP Hl PVP ¹ HP

P rt «O: p: O ö PHP er 3 Ά P Φ Ω ι-i ii P 03 tr li rt rt o co P- 3 ti z H P- P 01 er 3 1 P- Φ ö P P- PP 01 SV - SV P Φ li N 3 00 ι-i P er PV φ 01 er O d φ φ w PL Φ P- P- PL HP ¹ TJ rt Φ φ φ d P- PL P Hi P li H φ ι-i Φ P P- P 03 PL rt P Φ HP φ PL SV P- P- d PP PL yQ 01 rt 3 P- oi o P- SV

PL SV P Φ er ιq rt P- ii 3 H 3 N φ Φ er N ι-i PL P ¹ • Φ Φ CD: rt P- φ rt SV 03 Φ φ P ¹ φ O Φ P li P- Φ p- Φ P Cd 03 P rt ι-i P- s: co P- σi 3 P rt rt • P ii CT

PP P- φ H oi H Cd li P- o 3 Φ P ¹ P- Φ Φ φ φ _{^ *} PN d ii P 3 P ¹

Hi N φ p> Ω P P- SV P ^> P i φ li li PP tr ¹ d P tr φ φ d P σι TJ

P- O d Φ i-i P rt d tr φ PL φ P H Φ rt 01 rt PL O P- yQ o $ 01 Ω H O

P V Φ z 3 S Hi Φ co P l-i P o N d P Φ O P- rt d: Ω tr P tr tr V P 03

P P- φ P rt Z {- • P 3 ^ P ¹ PN PL TJ d P PL li N φ α er tr PL Φ rt P g φ 3 P- er P ¹ P P- 1-1 φ Φ O: er P- cn 01 Φ φ rt 3 "3 Φ Hi 01 d Φ Φ Φ H P- P P- P- rt

Φ Φ PP "SV ι-i li 03 Φ rt s: d li P- P d: 3 P- J li NH <PL SV 3 P ¹ S P-

SV 01 P 01 φ PL d ü ^" ι-i rt P- Φ ιq φ 03 φ X ι-i 3 O d Q φ li σi PO

P PL er φ P 3 Φ Φ PH rt P Ω P- \ g ιq rt PL li P ¹ φ P o co yQ P

P s Φ P- P> Φ P ιq φ Ω rt Φ tr PK! φ PL d Φ Hi Φ PP tr ¹ Φ 01 α P ¹ φ P-

P P- HP σ> ii ω tr TJ p. Φ P 3 Φ φ li P- PPOP ¹ P n yQ PP ι-i V 1 P- P Φ rt SV φ PL 01 P d φ TJ p- PPZV φ LP vo o φ P ιq o K Φ ii

Φ H <i φ P Ω <! φ P φ PP li Cd rt P- 3 d yQ N ö P p- φ t TJ O: Hi P ¹ tr

H o Φ P- d tr o 03 P rt V O P- o P ι-i P <L Φ φ d 1 t P tr Φ t P

03 HP Hi P ii ιq • rt Cd yQ PL φ d P. φ Φ PP TJ w PL V PL P ¹ P Φ ii li

= 3 V 03 Φ "3 Φ rt o ro Φ P- H PL ro P i-i ro 1-1 i-i PL O P- P φ φ ro P- d P rt P •

Φ o Φ l-i Φ rt Φ J P CS3 P Φ i-i 01 Ω PL 3 <φ 03 P 3 ι-i p- 03 er 01 er P

P- TJ rt P P- rr P PL PL ä H o tr P- <! φ O φ P P- 01 Φ rt H ιq P- 03 ι ü

03 N N O Z * P- Φ Φ p O i-i φ ro 01 O 1-1 rt J H -i O P- φ O φ P

Φ P rt d 13 φ tr 03 ro Φ 03 H PL OP li 03 α 3 P P- P- P d <J Φ er PL Z l N d H g li P Ω ι-i 3 P <φ ö ~ j φ P 1 d 1 φ ON 1 Φ rt H P- or

03 Φ PL Hi tr Φ H 1 P 3 P 01 01 P P- 00 ii P- 1 P ¹ PL tr 1 P φ rt Φ ι-i P- ιq oi P- Φ 1 1 PL Φ PP 1 P φ φ 1 φ 01 P 1 rt

determined by the optical parameters of the sensor and the speed of the axes of the positioning device.

The images recorded with the optical sensor 7 and the CCD camera 8 can be evaluated by correlation with a previously defined pattern of the calibration mark 16. This pattern can be created, for example, in terms of program technology or can be generated from a previously recorded image using image processing software. The pattern is stored in a file in the memory of the control computer 14 so that it is available to the image evaluation software during the X / Y calibration. An operating unit 29 is provided for making the operating inputs required for calibration X / Y and for displaying the results or the status of the inspection device. It is also used for operation during the actual inspection process.

When the calibration X / Y program has been started, the CCD camera 8 first records two images of the calibration mark 16 at different positions. FIGS. 3A and 3B show basic representations of the images taken. A circle 31 and 32 respectively marks the position of the calibration mark 16 in FIGS. 3A and 3B. The axes XM and YM of the machine coordinate system are additionally shown in FIGS. 3A and 3B. The position of the center of the image is in Figure 3A on the machine coordinates (XMP1, YMP1), in Figure 3B on the machine coordinates (XMP2, YMP2). The center of the image of the calibration mark 16 is located in FIG. 3A at the coordinates XI ', Yl' of an image coordinate system with the axes X 'and Y', in FIG. 3B at the coordinates X2 ', Y2'. The position of the center point can be determined in a simple manner using suitable image evaluation software. The scale of the axes X 'and Y' corresponds to the number of pixels in the image. From the so determined

Values can be easily and with high accuracy a pixel resolution CamPX in the direction of the X axis and a pixel resolution Ca PY in the direction of the Y axis can be determined using the following formulas:

_nΛ XMP1 - XMP2

CamPX ≡

X2'-XV

and

YMP1 - YMP2

CamPY =

72'- YV

The coordinates of the center of the image of the CCD camera 8 can also be calculated with this:

CamPosX = XMP1 + CamPX (XMAX '/ 2 - XI') and CamPosY = YMP1 + CamPY (Yl ^f - YMAX '/ 2).

The values XMAX 'and YMAX' represent the total number of pixels in the direction of the X 'and Y' axes.

The optical sensor 7 is then positioned over the calibration mark 16 and scanned with a first scan area in the X direction, so that a height image is recorded, as shown in FIG. 4A. The image according to FIG. 4B is recorded with the optical sensor 7 via a second scan area in the X direction. Axes XM and YM of the machine coordinate system and axes X ' ¹ and Y' ^{τ of} the coordinate system of the optical sensor 7 are shown in FIGS. 4A and 4B. The center of the recorded image of the calibration mark 16 is located in FIG. 4A at the machine coordinate XM1, in FIG. 4B at the machine coordinate XM2. The two values XM1 and XM2 are, of course, the same, since the position of the calibration mark 16 within the inspection device has not changed in the period between the taking of the two images. In FIG. 4A the origin of the image coordinate system lies on the machine coordinate XMS1, in FIG. 4B on the machine coordinate coordinate XMS2. The center of the image of the calibration mark 16 is located in FIG. 4A at the coordinate Y1 ″ of the image coordinate system, in FIG. 4B at the coordinate Y2 ″. A pixel resolution SensPX of the optical sensor 7 in the X direction is calculated with these values:

_ _Nv XMS2 - XMS1

SensPX =.

72 "-71"

FIGS. 5A and 5B show images recorded with the optical sensor 7 for a third and a fourth scan area in the Y direction. Axes XM and YM of the machine coordinate system and axes X '' 'and Y' "of the image coordinate system of the optical sensor 7 are again shown. The origin of the image coordinate system for the third scan area lies on the Y coordinate YMS3 of the machine coordinate system, for the fourth scan area on the Y coordinate YMS4. The center of the image of the calibration mark 16 lies in the image coordinate system on the coordinate Y3 "" for the third scan area in FIG. 5A and on the coordinate Y4 "" for the fourth scan 5B, from which a pixel resolution SensPY is determined to:

_ _Nv YMS3 - YMS4

SensPY =.

74 '' - 73 ''

The coordinates SensPosX and SensPosY of the theoretical center of the calibration mark 16 can be calculated, for example:

SensPosX = XMSl - SensPX ■ YMAX '' / 2

and

SensPosY = YMS3 + SensPY ■ YMAX "'12. In this, YMAX '' and YMAX '''are the maximum number of pixels in the direction of the Y''axis and in the direction of the Y''' axis of the image coordinate systems.

Finally, the calculated values of the variables CamPosX, CamPosY, SensPosX, SensPosY, Ca PX, CamPY, SensPX, SensPY and a value of the height resolution and the height measuring range of the optical sensor 7 are numerically output on the display of the control unit 29.

As an alternative to the mean values CamPosX, CamPosY, SensPosX and SensPosY of the recorded images of the calibration mark 16, the position of the origin of the image coordinate system can of course be calculated and displayed in a simple manner.

Claims

claims

1. Device for inspecting a three-dimensional surface structure of a substantially flat test object (3) with a camera (8) for recording a two-dimensional one

Image of at least a partial area of the surface, with a positioning device (9 ... 13) with which the camera (8) and test specimen (3) can be positioned relative to one another, characterized in that at least one additional optical sensor (7) for three-dimensional detection a portion of the surface of the test object (3) is provided that the optical sensor (7) and the camera (8) are combined in one sensor unit (6) and that a calibration mark (16) attached to the inspection device independently of the test object (3) is present, by means of which the position of the camera (8) and the optical sensor (7) can be determined after taking an image or after measuring the calibration mark (16).

2. Device according to claim 1, characterized in that the calibration mark (16) is such that it provides both a contrast in the gray scale image of the camera (8) and in the height image of the optical sensor (7).

3. Device according to claim 2, characterized in that the calibration mark (16) as one on one level

(17) standing circular cylinder is formed, the base surface of which is aligned substantially parallel to the surface of the test specimen (3) and the upper circular surface is treated in such a way that its gray value differs significantly from the gray value of the plane.

4. A method for calibrating an inspection device according to one of the preceding claims, characterized ge enn- zei chne t, that the camera (8) is positioned above the calibration mark (16) by the positioning unit (9 ... 13), that at least one two-dimensional image of the calibration mark (16) is recorded with the camera (8) that the Calibration mark (16) position (CamPosX, CamPosY) and resolution (CamPX, CamPY) of the camera (8) are calculated so that the positioning device (9 ... 13) positions the optical sensor (7) over the calibration mark (16) that at least one three-dimensional image of the calibration mark (16) is recorded with the optical sensor (7) and that the position (SensPosX, SensPosY) and resolution (SensPX, SensPY) of the optical sensor (7 ) be calculated.

5. The method according to claim 4, characterized in that with the camera (8) and / or the optical sensor (7) in each case at least two images of the calibration mark (16) are recorded, the recording positions each in one to the plane of the test object (3rd ) parallel plane are offset from each other, and that the resolution ^' (CamPX, CamPY; SensPX, SensPY) of the camera (8) or the optical sensor (7) is calculated on the basis of the offset of the images of the calibration mark (16).

6. The method according to claim 5, characterized in that two images of the calibration mark (16) are recorded with the camera (8), the recording positions being offset from one another in a direction substantially parallel to the plane of the test specimen (3), based on the Images of the calibration mark (16) the resolution of the camera (8) is calculated so that two images of the calibration mark (16) are recorded with the optical sensor (7), the recording positions in a first being essentially parallel to the plane of the test specimen (3) Are offset towards each other, that the resolution of the optical sensor in the first direction is calculated on the basis of the images of the calibration mark (16), that two images of the calibration mark (16) are recorded with the optical sensor (7), the recording positions being in a second to the plane of the test object ( 3) are offset substantially parallel to each other, perpendicular to the first direction, and that the resolution of the optical sensor in the second direction is calculated from the images of the optical sensor.