JPH04238592A - Automatic bundled bar steel tally device - Google Patents

Abstract

PURPOSE:To improve the quality of a product by enabling tally for all the bar steel to be performed by accurately and automatically detecting the number of bundled bar steels by image processing. CONSTITUTION:The above device is equipped with an image pickup device 4 which image-picks up the end face of the bundled bar steel 1, and an image processing device 5 which applies the image processing to image-picked up data and detects the number of bundled bar steel 1. The image pickup data of the end face of image-picked up bundled bar steel is divided into the appropriate number of screen pats corresponding to the luminance of the data by the screen division means 5b of the image processing device 5, and binarization in accordance with the luminance is applied at every divided screen part by a binarization means 5c. Those image pickup data re divided into images at ever bar steel by a circular separation method by an image division means 5d and the image of the image pickup data at every bar steel is tallied by labeling processing by a tally means 5e. A tally result is outputted as the number of bundled bar steels.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention relates to a device used in a system for managing the number of bundled round steel bars on a conveyance line, and in particular, an automatic counting device for bundled steel bars that detects the number of products by image processing. Regarding equipment.

0002

2. Description of the Related Art Conventionally, steel bars have been counted before they are bundled, and after the final bundling, a visual inspection is performed by a person.

0003

[Problems to be Solved by the Invention] However, as mentioned above, conventional steel bar counting relies on human power, so as the number of bars increases, there is a possibility of false detection, and the number of throughputs is reduced. However, there are limitations, and sampling inspections are the main method.

[0004] Therefore, it is possible to automate the bar counting using an image processing device, but even if an image processing device is simply introduced, the image processing of objects with circular cross sections such as steel bars that are in close contact with each other will inevitably require an operator, etc. The problem is that manual correction is required, which cannot be handled by on-site operators.

[0005] The present invention aims to solve such problems by making it possible to accurately and automatically detect the number of bundled steel bars by image processing, and to enable the counting of all bundled steel bars. The purpose of the present invention is to provide an automatic counting device for bundled steel bars that improves quality and realizes unmanned counting work.

[0006]

[Means for Solving the Problems] In order to achieve the above object, an automatic bundled steel bar counting device (claim 1) of the present invention includes an imaging device that images the end face of a bundled steel bar, and an image capturing device that captures the image data. an image processing device that performs image processing to detect the number of bundled steel bars; Binarization means that performs binarization according to the brightness of each screen portion of imaged data; Image dividing means that divides the binarized imaged data into images for each steel bar by a circular separation method; and the image dividing means. The apparatus is characterized in that it includes a counting means for counting, by labeling, images for each of the steel bars of the imaging data divided by the means.

[0007] The automatic bundled steel bar counting device (claim 2) of the present invention also includes an imaging device that images the end face of the bundled steel bar, and an image processing device that performs image processing on the imaged data from the imaging device. an image processing device that detects the number of lines, and the image processing device performs differential processing on the imaging data to obtain a differential image, and performs ring-shaped pattern matching processing on the differential image to perform counting. It is characterized in that it is constructed with a counting means for performing the following.

[0008]

[Operation] In the automatic counting device for bundled steel bars according to claim 1, the image data of the end face of the bundled steel bars captured by the imaging device is divided into an appropriate number of screen parts according to the brightness by the screen dividing means of the image processing device. After the captured image data is divided into 2 parts, each divided screen part of the captured image data is binarized by a binarizing means according to its brightness. Then, the binarized imaging data is divided into images for each steel bar by the image dividing means using the circular separation method, and then the images for each steel bar in the imaging data are counted by the counting means in a labeling process. , the counting result is output as the number of steel bars in the bundled steel bars.

[0009] Furthermore, in the above-mentioned automatic steel bundling bar counting device of claim 2, the imaging data of the end face of the bundling steel bar taken by the imaging device is subjected to differential processing by the differential processing means of the image processing device, so that This is a differential image in which the features of each bar image are extracted not only by the brightness difference with the background but also by the gradient of brightness and darkness at the edges of the steel bar image. And for that differential image,
By performing ring pattern matching processing using the counting means, the number of steel bars of the bundled steel bars can be obtained.

0010

Embodiments Hereinafter, embodiments of the present invention will be explained with reference to the drawings. 1 to 5 show an automatic bundled steel bar counting device as a first embodiment of the present invention. FIG. 1 is an overall configuration diagram thereof, FIG. 2 is a flowchart for explaining its operation, and FIGS. 5 is a schematic diagram showing an example of image data for explaining the image processing state.

In FIG. 1, reference numeral 1 indicates a bundled steel bar in which a plurality of round steel bars 1a are bundled and is transported on a conveyance line in the right direction in the figure, and 2 indicates a bound steel bar 1 (end face to be imaged) at a predetermined position on the conveyance line. A sensor 3 for detecting arrival at a (imaging point) is a plurality of illumination lights 3a arranged around the circumference (six in the figure) to uniformly illuminate the end face of the bundled steel bar 1 to be imaged. 4 is an imaging camera (imaging device) that images the end face of the bundled steel bars 1;
This is an image processing device that performs image processing on the imaging data from the imaging camera 4 at the time of detection and detects the number of round steel bars 1a of the bundled steel bar 1.

The image processing device 5 includes (1) a color monitor 5a for displaying the image processing status, and (2) an imaging camera 4.
A screen dividing means 5b that divides the imaged data from the screen into an appropriate number of screen parts (in this embodiment, three parts as shown in FIG. 3) according to the luminance; A binarization means 5c performs binarization (see FIG. 4) on each screen portion of the imaged data according to the brightness; An image dividing means 5d that divides into images for each round steel bar 1a (see FIG. 5), and a counting means 5e that counts, by labeling, images for each round steel bar 1a of the imaged data divided by this image dividing means 5d. It is composed of: Further, 6 is an input/output device (I/O) that interfaces between the image processing device 5 and the host computer 7.
).

The operation of the apparatus of this embodiment constructed as described above will be explained with reference to FIG. First, the sensor 2 determines whether or not the bundled steel bar 1 has reached the imaging point (step S1), and based on the image data (image of the end face of the bundled steel bar 1) from the imaging camera 4 at the time when the bundled steel bar 1 has reached the imaging point. Then, the image processing device 5 performs image processing (step S2
). Note that when imaging the end face of the bundled steel bar 1, a uniform image can be obtained by illumination from the multi-light illumination device 3.

The bundled steel bar 1 imaged by the imaging camera 4
The imaging data of the end face is sent to the screen dividing means 5 of the image processing device 5.
b, the screen is divided into three screen parts A, B, and C according to their luminance, as shown in FIG. 3 (step S3). This is because if the screen shown in Fig. 3 is binarized as it is, the difference in brightness at the end face of the bundled steel bar 1 will be too large (the lower right part of Fig. 3 is dark), so an appropriate threshold value for binarization should be set. This is because the settings cannot be made. Then, the binarization means 5c performs binarization (discriminant analysis method) on each divided screen portion according to its brightness (steps S4, S5). By performing binarization after screen division in this manner, even if there is a brightness difference in the captured image of the end face of the steel bar, it is possible to prevent the image of the round steel bar 1a from being hidden in the background due to the binarization.

After that, noise is removed from the binarized image data (step S6), and the divided screen is restored as shown in FIG. 4 (step S6).
7) In the image dividing means 5d, image division is performed for each round steel bar 1a by the circular separation method (step S8).
. The image after the image division is shown in FIG. Here, the circular separation method is a general method, and is also used, for example, when counting the number of microparticles observed with a microscope.

As shown in FIG. 5, a labeling process is applied to the divided image for each round steel bar 1a by the counting means 5e, and the counting means 5e counts the round steel bars 1a. It is output as the number of steel bars of the steel bar 1 (step S9). Here, the labeling process is a general method, for example, "Introduction to Computer Image Processing" published by Souken Publishing.
It is also described in the Japan Industrial Technology Center edited by Hideyuki Tanaka.

In this embodiment, in parallel with the counting process in step S9, counting is also performed by comparing the areas of the imaged images of the round steel bar 1a (step S10). For example, find the area of the end face of the steel bar from the number of imaging pixels,
The measured value is ±2 of the set value (value calculated from upper data)
If it is within 0%, it is counted as one, and if it is above it, it is counted as two. By performing such counting methods in parallel, the reliability of counting can be further improved.

In this embodiment, the counting result from the counting means 5e is compared with the data from the host computer 7, and if they match, normal output is performed (step S12), while if they do not match, the data is output normally. If so, an error is output (step S13), and the data is transmitted to the host computer 7 (step S14).

[0019] As a result, if the counting results do not match, the conveyance line can be stopped and re-inspected by an error output, and the quality of the final product can be reliably guaranteed and the reliability of the system can be improved. It can also improve sex.

As described above, according to the automatic bundled steel bar counting device of the first embodiment, the number of bundled steel bars 1 can be accurately and automatically detected by image processing, and all the bundled steel bars can be counted instead of the conventional sampling inspection. It is possible to count the number of bundled steel bars 1, and the product quality is greatly improved. Additionally, the counting work can be automated, resulting in significant labor savings.

Next, an automatic counting device for bundled steel bars as a second embodiment of the present invention will be explained with reference to FIGS. 6 to 9.
FIG. 6 is an overall configuration diagram, FIG. 7 is a graph for explaining the differential processing in this device, FIG. 8 is a schematic diagram showing an example of an image obtained by the differential processing, and FIG. 9 is a diagram showing a ring shape for pattern matching. It is.

As shown in FIG. 6, this embodiment also has the above-mentioned first
Although the configuration is almost the same as that of the second embodiment, the configuration of the image processing device 5A is different in this second embodiment. That is,
The image processing device 5A of this embodiment performs differential processing on imaging data from (1) a color monitor 5a that displays the image processing status, and (2) an imaging camera 4, and uses the absolute value output as a differential image (Fig. 7, (see 8);
(2) A counting means 5g that performs a ring pattern matching process (described later with reference to FIG. 9) on the differential image from the differential processing means 5f and performs counting.

By the way, when there is a relationship between the image brightness and the axial direction as shown in FIG. 7(a), if the threshold value for binarization is set to D1, the steel bar ■ will be hidden in the background, but if the threshold value is set to D2 In this case, it becomes impossible to distinguish between steel bars ■ and ■. In the first embodiment, the screen is divided according to the brightness to prevent this from happening, but in the present embodiment,
Focusing on the fact that there is a gradient of brightness and darkness in the edge portion of the steel bar image, the differential processing means 5f of the image processing device 5A is applied to the image data.
By performing differential processing, a differential image is obtained in which the features of each steel bar image are extracted.

FIG. 7(b) shows the result of differentiation of the image shown in FIG. 7(a). By performing differentiation, the differential value becomes large at a location where the brightness changes rapidly, so that a steel bar can be detected even at a location where the brightness itself is small. Actually, as shown in FIG. 7(c), the absolute value of the differential value is taken and binarization is performed at a luminance change rate of a certain value or more.

Although the image obtained by such processing has some defects, it becomes a figure with a clear outline as shown in FIG. 8, for example. By performing ring-shaped pattern matching processing on the differential image using the counting means 5g,
The number of steel bars in the bundled steel bar 1 is obtained.

Specifically, the black part in FIG. 9 is centered (xc
, yc) (xn, yn) [n=1 to 8]
The coordinates of each part are entered in a table in advance. Figure 8
If the matching ring of FIG. 9 is scanned from the upper left of the screen on the differential image of The outline of the circle (differential image) matches to a large extent. Count the number of matches and erase the target area one by one. By performing such scanning to the bottom right of the screen, the number of round steel bars 1a can be accurately counted.

[0027] Table 1 below shows (xn, yn) [n=1~
8] and an example of matching processing is shown in Table 2 below. Here, the number of matches is calculated at certain coordinates (x, y). Further, r is the ring radius.

[0028]

[Table 1]

[0029]

[Table 2]

[0030] In Table 2 above, 〇 indicates when (xn, yn) matches the counting target, and × indicates when (xn, yn) does not match the counting target. For example, if there are 6 or more 〇s, If we assume that pattern matching is established in Table 2.
Among them, Example 1 is determined to be a match, and Example 2 is determined not to be a match.

As described above, according to the automatic bundled steel bar counting device of the second embodiment, the bar counting is performed by applying pattern matching processing to the differential image in which the features of the image of the round steel bar 1a are extracted. , the number of bundled steel bars 1 can be detected very accurately and automatically, and the same effects as in the first embodiment can be obtained.

In the second embodiment, the case where the matching ring is approximated by eight points has been described, but the present invention is not limited to this.

[0033]

As described in detail above, according to the automatic bundled steel bar counting device (claim 1) of the present invention, the image data of the end face of the bundled steel bars is divided into an appropriate number of screens according to the brightness by the screen dividing means. The image is divided into parts, each divided screen is binarized according to the brightness by a binarization means, the image division means is divided into images for each bar by the circular separation method, and then the counting means is used for labeling processing. By counting the number of bundled steel bars, it is possible to accurately and automatically detect the number of bundled steel bars, and it is possible to count all the bundled steel bars instead of the conventional sampling inspection.
In addition to greatly improving product quality, counting work can be automated, resulting in significant labor savings.

Further, according to the automatic bundled steel bar counting device (claim 2) of the present invention, the counting means performs a pattern matching process on the differential image obtained by extracting the features of the image of the steel bar by the differential processing means. By counting the steel bars, the number of bundled steel bars 1 can be detected very accurately and automatically, and the same effect as the apparatus of claim 1 can be obtained.

[Brief explanation of the drawing]

FIG. 1 is an overall configuration diagram showing an automatic bundled steel bar counting device as a first embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the apparatus of the first embodiment.

FIG. 3 is a schematic diagram showing an example of image data for explaining an image processing state in the apparatus of the first embodiment.

FIG. 4 is a schematic diagram showing an example of image data for explaining an image processing state in the apparatus of the first embodiment.

FIG. 5 is a schematic diagram showing an example of image data for explaining an image processing state in the apparatus of the first embodiment.

FIG. 6 is an overall configuration diagram showing an automatic bundled steel bar counting device as a second embodiment of the present invention.

FIG. 7 is a graph for explaining differential processing in the apparatus of the second embodiment.

FIG. 8 is a schematic diagram showing an example of an image obtained by differential processing.

FIG. 9 is a diagram showing a ring shape for pattern matching.

[Explanation of symbols]

1 Bundled steel bar 1a Round steel bar 2 Sensor 3 Multi-light lighting device 3a Lighting light 4 Imaging camera (imaging device) 5, 5A
Image processing device 5a Color monitor 5b Screen dividing means 5c Binarizing means 5d Image dividing means 5e Counting means 5f Differential processing means 5g Counting means 6 Input/output device 7 Host computer

Claims (2)

[Claims] 1. An imaging device that images an end face of a bundled steel bar; and an image processing device that performs image processing on imaged data from the imaging device to detect the number of bars in the bundled steel bar; The apparatus includes a screen dividing means for dividing the imaged data from the imaging device into an appropriate number of screen parts according to the brightness thereof, and a screen part for each screen part of the imaged data divided by the screen dividing means according to the brightness. a binarization means for performing binarization; an image division means for dividing the imaged data binarized by the binarization means into images for each steel bar by a circular separation method; An automatic bundled steel bar counting device comprising: a counting means for counting images of each of the steel bars of the imaging data by labeling. 2. An imaging device configured to image an end face of a bundled steel bar, and an image processing device configured to perform image processing on imaged data from the imaging device to detect the number of bars in the bundled steel bar, the image processing device The apparatus includes a differential processing means for performing differential processing on the imaging data from the imaging device to obtain a differential image, and a counting unit for performing a ring-shaped pattern matching process on the differential image from the differential processing means to perform counting. An automatic counting device for bundled steel bars, characterized by comprising: means.