JPS6363043B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a shape detection device for detecting deformation of a leading end and a trailing end of a steel material that occurs during rolling, for example, in a hot rolling process of steel.

FIG. 1 is a system diagram of this type of device according to the prior art, in which 1 is a red-hot steel plate, 2 is an imaging lens that forms an image of the steel plate, 3 is arranged in the width direction of the steel plate,
A group of photoelectric elements converts the radiant energy into an electrical signal; 4 is an amplifier that amplifies the electrical signal; 5 is a quantizer that digitizes this analog electrical signal into multiple bits; 6 is a quantizer that keeps the digital data constant. The imaging lens 2, the photoelectric element group 3, the amplifier 4, the quantizer 5, and the storage device 6 detect the radiation emitted from the red-hot steel plate 1, and detect the progress of the steel plate 1. It constitutes a first means for sequentially scanning according to the distance and storing it as a shape signal. 7
performs a histogram operation on the data group in the storage device on the data in the local region, that is, the shape signal of the intermediate region of the steel plate 1 (second means), and determines the outer circumference tracking starting point (fourth means). The arithmetic unit 8 is a third means for performing a spatial differential operation on all the data in the storage device, and the second arithmetic unit 9
is a fifth means for tracking and extracting the outer peripheral shape of the leading and trailing ends of the steel plate 1 from the values calculated by the first and second calculating devices 7 and 8, and is a third calculating device.

Next, the operation will be explained. The digital output obtained from the quantization device 5 indicates the amount of radiant energy from the steel plate 1, that is, a value corresponding to the temperature distribution of each part of the steel plate. Generally, the temperature distribution on the xx line at the tip of the steel plate 1 (FIG. 2a) is such that the edge portion a has a lower temperature than the internal region b, as shown in FIG. 2b. Further, as shown in FIG. 2c, the temperature distribution on the y-y line in the length direction is higher in the central region c than in the inner region b, and the distribution is stable. Therefore, ideally, the tip shape signals conforming to the above-mentioned characteristics would be sequentially stored in the storage device 6 at the next stage in a two-dimensional array.

However, in the actual rolling process, as shown in FIG.
occurs, and the temperature distribution on the
As shown in FIG. 2B, a part of the shape signal may be lost.

The first to third arithmetic units 7, 8, and 9 are for processing shape signals containing such disturbance information. This will be explained in detail below.

5A to 5C are diagrams for explaining the mode of signal processing of this first arithmetic device 7. Select any data in c,
A temperature histogram showing the radiation intensity distribution is obtained, and a temperature histogram as shown in FIG. 5b is obtained. In this figure, there is a peak 9c' with a high frequency of "blurring" 9c on the outer periphery of the steel plate 1 in the low temperature range, and a peak 9c' that shows the "blur" 9c on the outer periphery of the steel plate 1 in the low temperature range, and a peak 9c' that shows the scale 9a, the water rider 9b, and the steel plate itself in the middle temperature range b to the high temperature range c. Frequency peaks 1', 9
a' and 9b' appear. It has been experimentally confirmed that this histogram is very stable even if the size of the region c is slightly different.

Therefore, if we perform binarization using the temperature at the minimum frequency point, that is, the temperature at the "valley" position, as a threshold from this temperature histogram, we can obtain "blur" 9c in region c as shown in Figure 5c. Since the shape signal from which the portion has been removed is obtained, only the true steel plate can be extracted without being affected by the outer periphery "blur" 9c, scale 9a, sailing 9b, etc., and the outer periphery tracking starting point S in the tracking calculation described later can be extracted. can be determined (fourth means).

Subsequently, the second arithmetic unit 8 reads out the digital data group from the storage device 6, and for example, 3×3
A spatial differential calculation is performed using the window to emphasize the outer circumferential edge of the steel plate 1 tip shape signal. Figures 6a and 6b are diagrams for explaining the mode of signal processing, and Figure 6b is a signal waveform diagram showing the result of spatial differential calculation, where the maximum value is at the true edge of the steel plate 1. can get. Therefore, even if there is a disturbance as described above, only the outer peripheral edge of the steel plate 1 can be emphasized based on the magnitude relationship of the differential values, and the path in the subsequent tracking calculation can be determined. This is because the temperature at the edge of the steel plate 1 is sufficiently high compared to the background, and there is not much temperature difference compared to the internal surface scale, water riding, etc., which is typical of the temperature distribution of the steel plate 1. It can be said that this is graphical signal processing that utilizes features.

The processing results obtained by the first and second arithmetic units 7 and 8 as described above are stored again in the storage device 6, and based on this data, the third arithmetic unit 9
Perform circumferential tracking using . FIGS. 7a to 7c are diagrams for explaining the aspect of this signal processing. In the third calculation device 9, for example, from the tracking starting point S obtained from the first calculation device 7 as shown in FIG. 7a, Tracking is continued until the end point E is reached while sequentially searching for the maximum direction component of the spatial differential value calculated by the second arithmetic unit 8 in the clockwise direction. That is, as shown in FIG. 7b, a window M consisting of, for example, 3×3 picture elements with appropriate weighting coefficients is set, and the maximum directional component of the spatial differential value within this window M is calculated as the outer circumference of the tip shape. The trajectory of this maximum directional component is stored again in the storage device 6 as the "true" outer peripheral shape of the tip of the steel plate 1. The locus of this tip shape signal becomes a profile-only signal as shown in FIG. 7c (fifth means).

It goes without saying that the rear end of the steel plate 1 can also be treated in the same manner as the front end.

Since the shape detection device according to the prior art is configured as described above, its ability to track minute changes in the shape of a steel sheet is low. Since the processing speed is slow, it is difficult to extract the steel plate image efficiently. Furthermore, since there is no means for correcting data loss or omissions caused by scale, water riding, etc., there have been many problems, such as the tendency for strays or runaways to occur.

This invention was made with the aim of solving the above-mentioned problems, and after performing local histogram calculations and spatial differential calculations on the shape data in the storage device, three types of windows adapted to the shape of the steel plate are created. The present invention aims to provide a device that extracts steel plate images at high speed and with high accuracy by using a peripheral tracking process having the above function in combination with a function to correct data defects, omissions, etc.

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 8, 9 indicates two arithmetic units 7 and 8.
10 is a tracking calculation device that extracts the outer peripheral shape of the steel plate using three types of window settings that match the shape of the steel plate 1 based on the processing results. It is a correction calculation device that is a correction means that performs interpolation.

Next, the operation will be explained. In Figure 8,
In the tracking calculation device 9, first, as shown in FIG. 9a, for the typical shapes A, B, and C of the tip of the steel plate 1 obtained in the rolling process, each part is roughly divided into three shapes as shown in FIG. 9B. It is classified into three regions P, Q, and R. That is, by regarding regions Q and R as linear regions whose shapes change slowly, and region P as a curved region whose shape changes rapidly, it is possible to set an efficient window that matches the shape of each region. .

In FIG. 10, three types of windows M ₁ , M ₂ , M ₃ having weighting coefficients corresponding to the characteristics of each area are set, and the tracking start point S obtained by the second calculation device 8 is set.
This figure shows how the operation is sequentially switched in a clockwise direction until reaching the end point E. Directional weight coefficients W ₁ , W ₂ within three windows M ₁ , M ₂ , M ₃ , W ₃ , W ₄ are M ₁ ,
As shown in M ₂ and M ₃ (however, W ₁ > W ₂ > W ₃ > W ₄ )
It is set.

That is, windows M ₁ and M ₃ in regions Q and R
In addition to setting the maximum weighting coefficient W ₁ in the direction of progress of tracking, a relatively large weighting factor W ₂ is applied diagonally to the left toward the direction of progress in order to prevent "climbing" into the interior of the shape. In addition, like M ₁ and M ₃ , the window M ₂ in the region P has the maximum weighting coefficient W ₁ in the direction of travel, and is given a relatively large weight to facilitate tracking in the left and right diagonal directions.
It is distinctive in that it carries out W ₂ . Furthermore, the switching of the windows M ₁ , M ₂ , and M ₃ is performed by using the direction vector obtained in the tracking process by the correction calculation device 10 described later.
As shown in FIG. 12a, this is carried out when a predetermined number N ₁ , N ₂ or more of predetermined directional vectors exist in the step of sequentially storing them in a vector memory provided in the computer and reading them out. Therefore, depending on the shape of the steel plate, there may be a slight deviation in the timing of switching, but in practice this is an error that does not pose any problem.

Furthermore, as shown in Figure 12b, due to disturbances such as a scale riding on the outer periphery of the tip of the steel plate, the second
If a part of the output of the arithmetic unit 8, i.e., the spatial differential value, is missing (Ld) or a phenomenon equivalent to it occurs, a tracking error will occur, for example,
There is a risk of getting lost in the wrong direction. Therefore, the 12th
As shown in Figure c, among the direction vectors at each point in time stored in the vector memory in the correction calculation device 10 during the tracking process, the direction vector (Vx) immediately before falling into the above state is referred to, as shown in Figure 12D. After vector correction (Vc) is performed, tracking can be continued.

As described above, according to the device according to the present invention using the tracking calculation device 9 and the correction calculation device 10, it is easy to follow minute changes in the shape of the steel sheet, and moreover, compared to the device according to the prior art, the device according to the present invention can Since the search area is reduced, processing time can be shortened.

Furthermore, since a correction means is provided by referring to the immediately preceding vector for loss or omission of data caused by disturbances, extremely stable operation can be expected.

In addition, almost the same application is possible to the rear end portion of the steel plate.

In addition, in the above embodiment, the tracking calculation device 9 and the correction calculation device 10 were configured by hardware.
Similar effects can be achieved by software processing using a microprocessor or the like.

As described above, the present invention performs local histogram calculation and spatial differential calculation on digital shape data stored in a temporary storage device, and then performs circumferential tracking having three types of windows corresponding to the shape of each part of the steel plate. It also has a function to interpolate and correct missing data caused by disturbances.
This has the effect of providing a high-speed and highly accurate device.

[Brief explanation of drawings]

FIG. 1 is a system diagram showing the configuration of a shape detection device according to the prior art, FIGS. 2 to 7 are diagrams for explaining its operation, and FIG. 8 is a system diagram of an embodiment of the present invention. 9 to 12 are diagrams for explaining the operation of this embodiment. In the figure, 1 is a red-hot steel plate, 2 is an imaging lens,
3 is a photoelectric element group, 4 is an amplifier, 5 is a quantizer,
6 is a storage device, 7 is a first arithmetic device, 8 is a second arithmetic device, 9 is a tracking arithmetic device, and 10 is a correction arithmetic device. Note that the same reference numerals in the figures indicate the same or corresponding parts.

Claims (1)

[Claims] 1 A first means for detecting radiation emitted from a steel material, sequentially scanning the radiation according to the traveling distance of the steel material, and storing it as a shape signal; a second means for calculating a histogram showing the radiation intensity distribution for the shape signal; a third means for performing a spatial differential operation on all the shape signals stored in the first means; a signal serving as the valley of the histogram obtained by the second means; fourth means for binarizing the shape signal of the intermediate region using the value as a threshold to determine a starting point for tracking the outer periphery of the end of the steel material;
By tracing the outer periphery of the end of the steel material from the tracking start point while setting three types of windows suitable for the shape of the steel material based on the spatial differential value by the third means, the outer periphery of the end of the steel material A shape detection device comprising: a fifth means for extracting a shape; and a correction means for detecting data loss, omission, etc. in the tracking process of the fifth means and interpolating it.