CALIBRATION ASPECTS OF MULTIPLE LINE-SCAN VISION

SYSTEM APPLICATION FOR PLANAR OBJECTS INSPECTION

Andrei Hossu and Daniela Hossu

University Politehnica of Bucharest, Faculty of Control and Computers

Dept. of Automatics and Industrial Informatics, 313 Spl. Independentei, sector 6, RO-77206, Bucharest, Romania

Keywords: Industrial Vision System, Line-Scan Camera, Dual-Camera Vision System, and Moving Scene in Robotic

Automation.

Abstract: The System Set-up Time it is one of the characteristics of an Industrial Vision System, besides the accuracy

performances and response time. Minimizing the set-up time while keeping the performances in accuracy

and in response time is one of the goals of any advanced Vision System. Starting from the purpose and the

required performances of the proposed Industrial Vision System, in the paper is presented a calibration

method developed for a multiple line-scan camera Vision System (in particular for a dual line-scan camera

system). The calibration method presented is based on analyzing the image of a calibration tool exposed to

the Vision System. There are presented the type of dimensional distortions identified from the experimental

results. The second part of the paper presents the calibration method. The Industrial Vision System

described in the paper is designed for silhouette inspection of planar objects located on a moving scene

(transport conveyor), in a robotic handling application (it is a pure 2D Vision System, the volumetric

characteristics of the analyzed objects being not relevant for the application). However the height of the

object is varying in time (from one set of objects to another). Due to the fact the distance between the

cameras and the objects is changing, the measuring results are affected. The proposed calibration method

allows the Vision System to self adjust the calibration parameters for a known change in height of the

objects, without affecting the accuracy system performances. In the final section of the paper are presented

some practical aspects of the proposed calibration method, and the balance between the off-line and the on-

line required computational efforts from the Vision System.

1 MULTI LINE-SCAN CAMERA

VISION SYSTEMS

CHARACTERISTICS

The class of the Artificial Vision Systems dedicated

for analyzing objects located on a moving scenes

(conveyor) presents some specific characteristics

relative to the Artificial Vision Systems dedicated

for static scenes. These characteristics are identified

also on the Image Calibration Process (Borangiu, et

al., 1994)., (

Haralick and Shapiro, 1992).

Figure 1 presents the model of the image

obtained from a dual line-scan camera Vision

System.

For this class of the Artificial Vision Systems we

could identify as relevant for the calibration process

the following characteristics:

- The system is using line-scan cameras for

the image acquisition.

- The system is a dual-camera.

- The obtained image has significant

distortions on (and only on) the image sensors

direction.

- There is an overlapped image area

between the two cameras. The end of the

acquisition line of the 1

camera is overlapping

the beginning of the acquisition line of the 2

camera. This overlapping area is significant in

dimension and is a constant parameter resulted

during the artificial vision system installation

process.

- There is a lengthwise conveyor distance

between the acquisition lines of the two

cameras. This distance is also a constant

parameter and its value is fixed during the

system installation process.

Hossu A. and Hossu D. (2008).

CALIBRATION ASPECTS OF MULTIPLE LINE-SCAN VISION SYSTEM APPLICATION FOR PLANAR OBJECTS INSPECTION.

In Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - RA, pages 36-40

DOI: 10.5220/0001480300360040

 SciTePress

2 THE PATTERN BASED

CALIBRATION TOOL

For the calibration process we adopted the method

of using a Pattern based Calibration Tool.

This Pattern based Calibration Tool represent a set

of blobs with a priori known dimensions and

locations for the real world (millimeters and not

image pixels) (Croicu, et al., 1998).

The outcome of using this type of calibration

technique was to obtain the following:

- Estimation with the highest accuracy of the

scene model parameters on the distortions

direction

- Estimation of the size of the overlapped image

area for both cameras

- The parallelism of the two acquisition lines if

obtain during the installation process, using the

support of the Calibration Tool

- The accuracy of installing the cameras in such a

way to obtain the perpendicularity of the

acquisition lines on the moving direction of the

scene (of the conveyor).

- Obtain a high accuracy on the distance

lengthwise the conveyor of the acquisition lines.

The shape and the dimensions of the pattern

adopted for the Calibration Tool force this

characteristic.

3 CALIBRATION TOOL

DESCRIPTION

In Figure 2 it is presented the pattern adopted for the

Calibration Tool used for the dual line-scan camera

Vision System (the dimensions are presented in

millimeters) (Croicu, et al., 1998)., (Hossu, et al.,

1998).

The characteristics of the adopted Pattern:

- The pattern contains dark blebs (marks)

placed on a bright background (with a high level

of light intensity for the image)

- The pattern is symmetrical on the vertical

direction (lengthwise the conveyor). The two

cameras have the acquisition lines parallel one

each other but located on different position on

the conveyor (due to the lighting system

adopted – built from two fluorescent tubes used

for obtaining the image from the reflection on

the object surface). 1st Camera will have the

acquisition line located on the top edge of the

lower section of the pattern, and the 2nd

Camera will locate its acquisition line on the

bottom edge of the upper section of the pattern.

- The pattern is partially homogenous on the

horizontal axis (the direction crosswise the

conveyor, the direction of the distortions)

- The pattern contains a characteristic of a small

difference (1 mm.) between the even and the

odd marks. This will force the installation

process to be very accurate in obtaining

the parallelism of the acquisition lines of the

cameras and also the perpendicularity on the

conveyor direction.

Figure 1: The image obtained from a dual line-scan camera Vision System.

CALIBRATION ASPECTS OF MULTIPLE LINE-SCAN VISION SYSTEM APPLICATION FOR PLANAR OBJECTS

INSPECTION

4 EXPERIMENTAL RESULTS OF

THE CALIBRATION PROCESS

In Figure 3 are presented the results obtained from

the Calibration process performed on the 1

Camera.

The Excel Cell used as support for representing

the results of the Calibration on the 1

Camera

contains the following:

- The 1

column (called Mark) contains the

number of the corresponding Mark existing in

the pattern

- The 2

column (called Calib. Cam)

contains the values of the coordinates of the

marks on the Calibration Tool.

These values are obtained from the “real

world” from direct measuring of the Pattern

applied on the Calibration Tool (represented in

millimeters).

- The 4

column (called Pixel(x)) represents

the coordinates of the existing Marks on the

image. These coordinates are represented in

pixel number.

- For both cameras we chose a Polynomial Trend

of order 3 for the approximation of the

conversion function of the image coordinates

values (pixels) into the conveyor coordinates

values (millimeters). Using this type of trend the

calibration method will provide results with a

maximum approximation error inside the

accuracy requirements of the particular Vision

System

- On the top and right side of the Excel Cell (the

first two rows and the last four columns) are

stored the parameters of the 3

Order

Polynomial estimated as trend.

- The last column (called Estimated (y)) contains

the estimated of the marks coordinates values

(on the conveyor), obtained from applying the

Polynomial trend of order 3.

- In the bottom right side of the Excel Cell is

presented the graphical chart of the trend of the

coordinates values of the marks on the conveyor

related to the pixel coordinates.

Figure 2: The pattern of the Calibration Tool used for the dual line-scan camera Vision System.

LL1 H H1 H2 alfa a1 b1 c1 d1

4636 1408 2592 17 31 3.01984E-03 1.448567E-09 -9.128585E-06 0.5088785 338.9459

Mark Calib Cam1 Real Cam1 Pixel (x) Estimated (y)

1 436 429.08037 175 427.7278

2 535.5 529.18545 380 531.0810

3 635 629.31390 576 629.3081

4 734.5 729.46574 777 729.5128

5 834 829.64098 979 829.7479

6 933.5 929.83962 1181 929.5853

7 1036 1033.08381 1390 1032.5400

8 1135.5 1133.32999 1595 1133.2616

9 1235 1233.59960 1799 1233.3085

10 1334.5 1333.89265 2004 1333.7361

11 1434.5 1434.71331 2210 1434.6181

12 1534 1535.05337 2415 1535.0504

13 1633.5 1635.41690 2621 1636.0883

14 1733 1735.80390 2825 1736.3342

15 1832.5 1836.21437 3028 1836.3487

16 1932 1936.64834 3231 1936.6953

17 2036.5 2042.15453 3443 2041.9242

18 2136 2142.63667 3644 2142.1760

19 2235.5 2243.14232 3844 2242.4668

20 2335 2343.67150 4046 2344.3760

Camera 1

y = 1.44857E-09x

- 9.12858E-06x

+ 5.08879E-01x +

3.38946E+02

500

1000

1500

2000

2500

0 500 1000 1500 2000 2500 3000 3500 4000

Pixel

Figure 3: Experimental results of the calibration process on the 1

camera.

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

Pixe l Width [m m ]

0.485

0.490

0.495

0.500

0.505

0.510

400 900 1400 1900 2400 2900 3400 3900 4400

mm / pixel

Figure 4: The distortions estimated on the acquisition lines direction.

Figure 5: Using the calibration results for a dual line-scan camera system.

In the figure we are using the following notations:

- x coordinate value in image – (pixels)

- y coordinate value on the objects scene (on

conveyor) – (mm)

dxcxbxay +++= for the 1

Camera and

dxcxbxay +++= for the

Camera

In Figure 4 is presented the graphical chart of the

evolution of the distortions estimated on the

acquisition lines direction. The two cameras are

covering around 4 meters wide view. The figure

represents the behavior of the two acquisition

cameras: in the are from 400 mm to 2250 mm it is

represented the behavior of the 1

camera and in the

area from 2250 mm to 4400 mm it is represented the

behavior of the 2

camera.

We can notice the distortions are effecting the

pixel width of the image form 0.49 mm/pixel to

0.507 mm/pixel.

Ignoring this variation of the pixel width along

the image acquisition line, would lead to

accumulation of very high errors in some areas of

the image, due to the fact the amount of pixels

contained in an acquisition line is high (8096).

5 CONCLUSIONS

The most important thing obtained from estimating

the scene parameters, is the estimation of the image

distortions.

Horizontal

matching

Sliding the

Camera

Vertical

matching

Sliding the

Camera

Image area

removed for

both 1

and

Camera

Cam 1

Cam 2

CALIBRATION ASPECTS OF MULTIPLE LINE-SCAN VISION SYSTEM APPLICATION FOR PLANAR OBJECTS

INSPECTION

This end of the process of distortion estimation

leads to obtaining a table for conversion the image

coordinates (pixels) in scene coordinates

(millimeters) for each of all the 8096 pixels of the

acquisition lines. This lookup table will contain

8096 values (double representation) representing the

real values of the scene (conveyor) coordinates of

each pixel.

The CPU effort of the image processing

algorithm will be minimal on converting the image

coordinates into conveyor coordinates, using the

pixel coordinate as the index of the offline built

lookup table.

In order to minimize the processing time of the

image, the polynomial estimation is not used on-line.

The polynomial estimation is used for building the

pixel to millimeters lookup table, in the offline stage

of the presented method. In fact using a polynomial

trend of 3

order or other type of trend is not

relevant for the response time performances of the

vision system but only for obtaining the required

approximation error. This stage, being offline and

using a relative small amount of data, another more

sophisticated approximation method could be used.

In Figure 5 are presented the ways the last two

steps of the calibration process are performed.

REFERENCES

Borangiu, Th., Hossu, A., Croicu, A., 1994. ROBOT

VISIONPro, Users Manual, ESHED ROBOTEC, Tel -

Aviv.

Croicu, A., Hossu, A., Dothan, E., Ellenbogen, D., Livne,

Y., 1998. ISCAN-Virtual Class based Architecture for

Float Glass Lines, IsoCE’98, Sinaia.

Hossu, A., Croicu, A., Dothan, E., Ellenbogen, D., Livne,

Y., 1998. ISCAN Cold-Side Glass Inspection System

for Continuous Float Lines, User Manual, Rosh-

Haayn.

Haralick, R., Shapiro, L., 1992. Computer and Robot

Vision, Addison-Wesley Publishing Company.

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics