FIELD OF THE INVENTION
The present invention relates to a crop shear in a hot strip mill located between
roughing stand, or stands, and finishing stands for cropping the ends of steel
plates prior to passage through the finishing stands. More particularly, the
present invention relates to a system for measuring accurate strip speed to be
provided as feedback to crop shear controller to cut the strip ends into optimum
lengths.
BACKGROUND OF THE INVENION
In a Hot Strip Mill (HSM), the slab traveling between the roughing mill and the
finishing mill is called a 'transfer bar'. The regularly shaped head and tail end of
each transfer bar must be trimmed in a 'crop shear' in order to avoid cobbles
when the transfer bar is rolled into a very long strip and subsequently coiled.
It too little of the head or tail is cropped the excess material may cause problems
when rolled, such as cobbling of the entire bar, marked rolls, or a torn off tail
that stays in the mill and cobbles the next bar. If too much is cropped the good
material that is wasted may be worth several million dollars per year.

Minimization of crop loss has been a great concern to the steel makers since the
inception of hot strip rolling process. This technology will not only help in
reducing the scrap loss due to crop cut but also have significant environmental
impact.
It has been observed that the loss of steel due to essential chopping of transfer
bar is significantly high in absence of a reliable strip feedback to the crop shear
controller. Erroneous speed calculation may either result in a larger cut (yield
loss) or a miss cut. Therefore, accurate speed feedback is critical for controlling
the yield loss thus produced.
In a conventional Hot Strip Mill, when a hot slab is rolled into thinner sections,
the edges of the stock deform and develop irregular shapes at the front and rear
ends. To square the ends and produce nearly rectangular shaped strip, the
irregular shaped ends thus formed are cut by shear before the stock is subjected
to further reductions. The cropped ends are disposed of as Scrap. Thus, the
length of the crop directly affects the net yield of the Hot Strip Mill. In a
conventional hot rolling mill, the irregular shaped front and rare ends are
chopped off by a crop shear. The crop ends are then disposed as scrap. Crop
loss alone contributes about 30% of the total yield loss of the mill. Considering
the environmental implication of the wastage thus produced, it becomes all the
more important to minimize crop losses. In the past, several attempts have been

made to tune the crop shear machine to minimize the cut length, but due to
non-availability of accurate transfer bar speed measuring device at high
temperature, the time to cut for the crop shear machine was not possible to
determine accurately. There are many other technologies that have been used or
being used for strip speed measurement. LASER Doppler Velocity Meter (LDVM)
is one such standard velocity measurement device and widely used in variety of
applications worldwide. In this particular problem, it failed to give accurate
measurements due to layer of flowing water on the strip surface and the hazy
. environment conditions, which is particularly not suited for LASER applications. A
device for measuring the transfer bar speed by adopting a pair of measuring rolls
on the pass line, had succumbed to very high ambient temperatures and
extremely adverse plant conditions. Another method which is being used is the
average uncoiling speed of the coil box, measured with the help of an encoder.
This measured velocity is found to be not accurate and fluctuating in nature. Yet
another innovative-attempt to measure the speed was envisaged using two hot-
meal-detectors kept at a fixed distance along the direction of rolling. The velocity
was computed by dividing the distance between them by difference of ON times
of respective HMDs. This arrangement is good but requires heavy maintenance
as the scales and water deposited on the tip of fiber optic probe of the HMD
needs to be cleaned frequently. Moreover, it is not possible to measure
acceleration in all the techniques described above, which is, indeed, a very
significant factor in calculating accurate time for the strip-head to reach crop
shear.

SUMMARY OF THE INVENTION
Accordingly, there is provided an improved system and a method of strip speed
measurement which controls the angular speed and momentum of Crop shear to
achieve desired cut. More specifically this is a machine vision system wherein a
CCD camera continuously monitors the strip movement. The images are in turn
fed into.an embedded hardware unit hosting the algorithm wherein the captured
frames are analyzed. It employs edge detection method on each captured image
frame to determine the optical flow between two successive frames obtained at
fixed time intervals.
These objectives are accomplished in accordance with the teachings of the
present invention by providing a new and improved method of strip speed
measurement which controls the angular speed and momentum of Crop shear to
achieve desired cut. More specifically this is a machine vision system wherein a
CCD camera continuously monitors the strip movement. The images are in turn
fed into an embedded hardware unit hosting the algorithm wherein the
captured frames are analyzed .It employs edge detection method on each
captured Image frame to determine the optical flow between two successive
frames obtained at fixed time intervals.
Since the Camera is installed at height with some inclination from moving strip,
the distance of imaging system from strip varies along the length of strip. This
may cause the problem in transformation of camera coordinate system to world
coordinate system i.e. obtaining accurate millimeter from pixel. So a camera
calibration algorithm is employed which transforms the detected edge positions

from camera coordinate system to world coordinate system more precisely. Strip
velocity is measured by using edge distances and fixed time intervals between
successive image frames.
The aforesaid Imaging system is placed nearby from the Moving Hot Strip. The
ambient temperature remains consistently around 60 °C. Thus the camera needs
to be properly cooled to get working in such a high temperature environment. A
pin hole camera enclosure is used to cool entire camera from high temperature
effects. Pin hole arrangement with air purging provides the facility to prevent the
passage of dust inside the enclosure and to clean up the dust accumulated onto
the glass in front of the lens as well as cooling of entire system.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS
Fig#l Roller table between Coil box and Crop shear
Fig#2 Schematic diagram of Strip velocity measurement System
Fig#3 Setup for Camera Calibration
Fig#4 Principle of World coordinates transformation from Camera coordinate
system
Fig#5 Installed machine vision system in Hot strip Mill.

Detailed Description of the Preferred Embodiments
Fig. 1 illustrates the manner in which irregular ends of a moving strip are
squared or cropped along head and tail end when moved through a rotary crop
shear.
In a hot strip mill, Metal slab 101 to be rolled are heated and then passed
through series of roughing stands (not shown in fig 1) to reduce slab (strip)
thickness. During this process the ends 110 become deformed because of lack of
restraint and form irregularities. These irregular ends must be squared off or
. cropped before feeding the strip into finishing stands. For this purpose a cutting
machine known as rotary crop shear 111 is used. Hot strip 101 after passing
through last roughing stand has deformed head 110 and tail (not shown in fig) is
rolled by coiler box and then decoils towards Crop shear. During decoiling it
travels on the rollers 102 towards crop shear 111.The rotary crop shear comprise
two massive rotary drums 105 and 106 having their longitudinal axes
perpendicular to the path of the strip through shear. The shear drums are
provided with cutting blades 107 and 112 and are driven in synchronism through
gears by DC drive motor (Not shown in fig). A reliable strip speed feedback is
required to the crop shear controller so that it crops head and tail end as short
as possible to maximize the length of prime material i.e. to reduce yield loss and
to avoid any delay due to miss cut. One innovative-attempt to measure the
speed was envisaged using two hot-metal-detectors 103 and 104 kept at a fixed
distance along the direction of rolling. The hot metal detector 103 and 104
actuates a contact closure in response to radiated energy from the hot strip, to
discriminate between the presence and absence of strip at this position .The

velocity was computed by dividing the distance between them by difference of
ON times of respective HMDs. This arrangement is good but requires heavy
maintenance as the scales and water deposited on the tip of fiber optic probe of
the HMD needs to be cleaned frequently. Moreover, it is not possible to measure
acceleration in all the techniques described above, which is, indeed, a very
significant factor in calculating accurate time for the strip-head to reach crop
shear.
As shown in Fig 5, the field unit of proposed machine vision based strip velocity
measurement system 502, at inclined positioned around 3 meters away from mill
center and at height of 4.5 meters, continuously monitors the advancement of
the hot strip 505 on the rolling table. Imaging System 502 is covered by a
canopy 503 to prevent system from excessive heat and water. The light rays 202
from the strip 201 goes inside the assembly 502 through a circular aperture (Pin
hole) 203 of diameter of 50mm. Pin hole 203 protects the camera optics and
electronics from moist iron dust and high ambient temperature .An air purging
outlet 205 is provided to prevent dust entering the housing 207.The light from
opening 203 is gathered and paralleled 208 to Camera lens assembly 210 by lens
206. The close up ring 209 is used in order to reduce minimum object distance.
Rays 208 paralleled by lens 206 is gathered and focused on CCD (Image Plane)
214 of Camera 215 by lens assembly 210 which has three individual rings for
?oom 211, focus 212 and aperture control 213. Camera 215 is driven by 12V DC
power provided by Adapter 219 which converts 220V AC to 12V DC. Port 216 of
Camera is used for both power and communication. The images captured by
camera 215 is sent to an embedded hardware 222 placed in control room, herein
known as the 'the host' through a physical Ethernet link 217 and 220, connected
to host 222 via PJ 45 connector 221.

As shown in Fig 3, a chessboard 302 of size 1800mm x 1600mm is placed on
roller table 301.With the help of chessboard 302 and perspective projection
geometry models, all the physical points of strip 506 in Fig 5 on the rolling table
are mapped very precisely to the image points in the camera plane 509 using
non linear least square optimizations. This way the exact transformation from
world coordinate frame (Xw, Yw, Zw) 508 i.e. rolling table, to camera coordinate
frame (Xd, Yd, Zd) 509 i.e. image, is obtained.
In fig 5, as the head of strip 501 comes in the field of view 507 of camera 502,
continuous images of strip are captured by camera 502 with time
stamps(tl,t2...tn) until it moves out from the field of view 507. The images
captured by the camera 502 are sent to host unit 222. The Control room unit 222
hosts the sophisticated image processing algorithms to compute optical
flow(dl,d2..dn) between two successive image frame, as shown in fig 6
obtained at a fixed time interval using edge detection.
Fig 6 shows the detected edges (d1, d2...dn) on successive images of strip
captured by camera at fixed time intervals (t1, t2...tn) in camera coordinate
frame(Xd,Yd,Zd)
dn = detected edge (Xn, Yn) on captured image in camera coordinate frame( in
pixel)
tn= time of captured image (in sec)

Detected edges in Image coordinate frame (Xd, Yd, Zd) are transformed in world
coordinates system (Xw, Yw, Zw) by using following transformation (As shown in
fig 4).

dw= detected edge (xw, yw) on captured image in world coordinate frame (in
mm).
Measurement is made more precise by employing a third order curve fitting on
cumulative displacements.
This process filters out the noise produced by erroneous measurements induced
because of flowing scales, fog and black patches due to localized cooling areas
near the edges of the strip. As a result, 'displacement' becomes a continuous 3rd
order polynomial function of 'time'.


'Strip velocity' and 'acceleration' are calculated using first and second order time
derivatives of 'distance', respectively

Velocity and acceleration together are used to calculate precise time for the strip
to reach crop shear to achieve optimum cut performance so that crop loss can be
reduced. These measurements are sent to the crop shear controller through an
Advantech USB module 223 (Fig2) connected with host 222. After receiving the
speed (and position) from proposed system (and the position of the transfer bar
by the HMD system) and cut reference from the OptiCROP system, crop shear
PLC controls the rotation of the blade to perform optimum cuts.


Other References
1. WORLD STEEL UNIVERSITY WEBSITE, Hot Rolling Link:
http://www.steeluniversity.org/content/html/eng/default.asp?catid=30&pageid=
2081272061
2. BERTHOLD K.P. HORN and BRIAN G. SCHUNCK, Determining Optical
Flow, Artificial Intelligence Laboratory, Massachusetts Institute of
Technology, Cambridge, MA 02139, U.S.A.
3. ADELSON, E H. AND BERGEN, J R. 1986. The early detection of motion
boundaries. In IEEE Proceedings of Workshop on Vtsual Motzon (Charleston,
S.C., May). 151 156.
4. ADR5 G. 1985. Determining three-dimensional motion and structure from
optical flow generated by several moving objects. IEEE PAMI 7, 4, 384 401.
5. ANANDAN P. 1989. A computational framework and an algorithm for the
measurement of visual motion. Int. J. Comput. -Vzszon 2,283-310.
6. BLACK, M. J. AND ANANDAN, P. 1993 A framework for robust estimation of
optical flow. In Proceedings of ICCV (Berlin, May), 231 236
2
7. FERMIN, I. IMIYA, A. 1994. Two-dimensional motion computation by
randomized method. Tech. Rep. TR ICS-4-6-1994, Dept of Information and
Computer Sciences, Chiba University, Japan.
8. E. De Micheli, V. Torre , S. Uras, The Accuracy of the Computation of Optical
Flow and of the Recovery of Motion Parameters, IEEE Transactions on
Pattern Analysis and Machine Intelligence, v. 15 n.5, p.434-447, May 1993
[10.1109/34.211464 ]
. 9. REICHARDT, W., SCHLOGL, R. W., AND EGELHOAF, M. 1988. Movement
detectors of the correlation type provide sufficient information for local
computation of 2d velocity fields. Naturwissenschaften 75, 313-315.

10. SOBEY, P AND SmNIVASAN, M.V. 1991. Measurement of optical flow by a
generalized gradient scheme. J Opt. Soc. Am. A 8, 9,1488-1498
11. SPETSAKIS, M. E. 1994. An optical flow estimation algorithm that uses
Gabor filters and affine model for flow. Tech. Rep., Dept of Computer Science,
York University.
12. Tsai, R. Y. HUANG, T. S. 1984. Uniqueness and estimation of three-
dimensional motion parameters of rigid objects with curved surfaces. IEEE PAMI
6, 1,13 27.
13. ZOLTOWSKt, M.D. 1987 Signal processing applications of the method of
total least squares. In IEEE 21st Annual Aszlomar Conference on Signals,
Systems, and Computers (Pacific Grove, CA, Nov.), 290 296.

WE CLAIM :
1. A Strip Velocity Measurement system to measure the hot strip speed
in hot strip mill, comprising a high resolution camera unit, an
embedded hardware unit which hosts an analysis software, a telephoto
lens to focus the strip image onto the CCD surface, a pin hole camera
enclosure arrangement for accommodating an imaging system which
makes it to work in very harsh environment, a clean air purging unit
which delivers clean air, and a flow of the clean air which is
delivered from said clean air unit, a calibration unit which consist of a
plurality of checker boxes distributed over a steel sheet for
determining extrinsic and intrinsic parameters of the camera, wherein
the system employs edge detection to determine the optical low
between two successive frames obtained at a fixed time interval,
wherein the edge detection based on adaptive threshold in each frame
is computed to determine the optical flow between the successive
image frames and wherein the hot strip speed is measured with its
acceleration to allow crop shear to cut the deformed ends of strip very
precisely.

2. The system as claimed in claim 1, wherein the hot strip speed is
measured with an accuracy of + 5 mm/s.
3. The system as claimed in claim 1, wherein hot strip speed is
measured even in the presence of flowing scales and patches on strip

ABSTRACT

The present invention relaters to a Strip Velocity Measurement system to
measure the hot strip speed in hot strip mill, comprising a high resolution
camera unit, an embedded hardware unit which hosts an analysis software, a
telephoto lens to focus the strip image onto the CCD surface, a pin hole camera
enclosure arrangement for accommodating an imaging system which makes it to
work in very harsh environment, a clean air purging unit which delivers clean
air, and a flow of the clean air which is delivered from said clean air unit, a
calibration unit which consist of a plurality of checker boxes distributed over a
steel sheet for determining extrinsic and intrinsic parameters of the camera,
wherein the system employs edge detection to determine the optical low
between two successive frames obtained at a fixed time interval, wherein the
edge detection based on adaptive threshold in each frame is computed to
determine the optical flow between the successive image frames and wherein
the hot strip speed is measured with its acceleration to allow crop shear to cut
the deformed ends of strip very precisely.