JPH0694646A - Circular container inner surface inspecting device - Google Patents

Abstract

PURPOSE:To accurately detect the depressed area on the side surface of a container by scanning the inner surface image of the circular container with ring-shaped or spiral circular scanning rasters and then comparing the concentra tion difference of elements at a specific spacing on the scanning rasters. CONSTITUTION:A video signal obtained by raster-scanning the image of a TV camera is A/D-converted and then is input to a frame memory 1 as a light and shade signal P0 for storing as multiple-value screen data. A light and shade signal (pixel value) 1a is read from the memory 1 by a circular address generator 4 and then is input to a failure detection circuit 12 via a window gate circuit 7. A circuit 12 delays the signal by the amount of picture spacing. compares the pixel value of an aimed point and that of a background point, and detects the failure pixels where the difference exceeds a threshold and then a failure detection criterion 13 totals the number of failure pixels and then outputs the criterion result (depressed area failure) to a total criterion circuit 19. The circuit 19 totally judges the criterion result of a criterion result (contamination failure) of a failure detection criterion circuit 11 and an output circuit 20 outputs the quality according to the criterion signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention inspects the inner surface of a circular container such as a beer can conveyed by a conveyor or the like, and detects the inner surface of the circular container as an image processing apparatus for detecting deformation, foreign matter, dust, scratches, etc. of the container. Regarding the device. In the drawings below, the same reference numerals indicate the same or corresponding parts.

[0002]

2. Description of the Related Art FIG. 2 is an explanatory view of a high-intensity part when an aluminum can container for beer is observed from above, as an example. FIG. 2A is an upper surface (image) view of the can container and FIG. ) Is a side sectional view. 102 is a container, 101 is this container 10
A ring-shaped illuminator that illuminates 2 from above, 103 is a high-intensity part at the mouth, and 104 is a high-intensity part at the bottom. By thus irradiating the inner surface of the container 102 with light rays by using the ring illuminator 101, the mouth portion and the bottom portion of the inner surface of the container 10 are respectively irradiated.
High-luminance portions such as 3,104 occur. This is particularly noticeable when the inner surface of the container has a metallic luster such as a can.

FIG. 3B is a top view image of the container 102 (see FIG.
It shows the density change on the scanning line Q-Q1 with respect to (A)), but is classified into five regions W1 to W5 according to the characteristics of the density change. The first region W1 is the mouth high-intensity part 10
3, the second region W2 is the middle upper part of the side surface of the container where the change in concentration is relatively small, and the third region W3 is darker than the other regions because the light rays from the illumination 101 described in FIG. It is the lower part of the side surface, the fourth region W4 is the bottom high-intensity part 104, and the fifth region W5 is the bottom.

Conventionally, a window is provided in each of these areas W1 to W5, and a threshold value for detecting a defect of black stain (black spot) or white stain (white spot) is set according to the optical characteristics of the area. Was. As a method of detecting a defect, for example, an 8-bit multi-value grayscale image signal obtained by A / D conversion of an analog video signal (analog grayscale image signal) obtained by scanning a target image is set at a predetermined threshold value. A binarization method and a differentiation method of differentiating the video signal to extract a defect signal are known. For this differentiation method,
A differential signal is output even in the contour portion of the outer shape of the object, but in the contour portion, either a positive direction pulse or a negative direction pulse is generated by differentiation, whereas a minute defect portion simultaneously generates a positive direction pulse and a negative direction pulse. It is possible to extract the defective portion by utilizing this.

That is, a value P (i, j) at a target point (coordinate values X = i, Y = j) of a signal P (X, Y) obtained by differentiating an analog grayscale image signal based on raster scanning,
The value P (i- at a point distant by a predetermined minute α pixel and β pixel from the point of interest on the X-direction scanning line, respectively.
between α, j) and P (i + β, j), P (i, j) −P (i−α, j)> TH1 and P (i + β, j) −P (i, j) )> TH1 (if TH1 is a predetermined threshold value (positive value)), the binarization function value PD (i, j) for defect detection at the point of interest
= 1 is set as the point of black level failure, and in other cases PD (i, j) = 0 is set as the point of normality.

[0006]

However, the above method is advantageous for defects such as minute black spots, but in the case of the dent of the container, local contrast cannot be obtained and detection is difficult. Therefore, it is an object of the present invention to provide a circular container inner surface inspection device that can solve this problem.

[0007]

In order to solve the above-mentioned problems, the apparatus for inspecting the inner surface of a circular container according to a first aspect of the present invention is arranged such that an axially symmetric circular container (102 or the like) is moved from the axial direction (ring illuminator 1).
The inner surface side of this circular container is illuminated, and the illumination surface of this circular container is imaged from this axial direction via a TV camera, and the captured image is analyzed to deform or stain the circular container. In a device for inspecting the inner surface of a circular container, an image of the inner surface of the circular container is formed into a ring shape or a spiral shape whose center is the center of the image and whose radius is sequentially changed at every one revolution at one or a plurality of predetermined pixel intervals. Scanning line (L
(Hereinafter referred to as “circular scanning line”), and the density of the pixel of interest (P0, etc.) as pixels lined up at one or a predetermined plurality of pixel intervals on the circular scanning line, and the density of the pixel of interest before or after the pixel. The absolute value (ΔP or the like) of the density difference from the density of the background pixel (P1 or the like) separated by a predetermined number of pixels (d or the like) in the direction of is determined by the threshold value ( THP, etc.) is sequentially repeated for each of the target pixels, and when an absolute value of the density difference larger than the threshold value is detected, it is determined that the inner surface of the container is defective (defect detection circuit 12, etc.). ).

According to a second aspect of the invention, there is provided a circular container inner surface inspection apparatus according to the first aspect of the invention.
The threshold value is set every rotation of the circular scanning line (via the scanning line counter 22, the threshold value table 27, etc.). A circular container inner surface inspection device according to claim 3 is the circular container inner surface inspection device according to claim 1 or 2, wherein the density of the background pixel is centered on this background pixel (via an averaging circuit 26 or the like). The average density of the pixels in a predetermined one-dimensional or two-dimensional local area.

[0009]

The inner surface of the circular container (such as the ring illuminator 101) is illuminated from the axial direction of the axisymmetric circular container (such as 102), and the circular container is illuminated from the axial direction through the TV camera. When the illumination surface is imaged and the image is analyzed, pay attention to the fact that the light and shade distribution on the inner surface of the container occurs on concentric circles as shown in Fig. 3, and scans the inner surface of the container with a circular or spiral circular scanning line. By doing so, a pixel row with a small grayscale change is obtained, and when a dent occurs, it appears as a grayscale change of the scanned pixel row. The absolute value ΔP of the density difference is calculated from the density P1 of the background pixel as ΔP = | P0-P1 |, while the threshold value THP is given to each of the windows W1 to W5 as shown in FIG. Compares the absolute value [Delta] P between the this threshold THP, [Delta] P
Is larger than THP, the pixel of interest is detected as an internal defective pixel such as a dent.

However, the threshold value THP may be set according to the position of the circular scanning line, for example, every one rotation of this scanning line. Further, the density P1 of the background pixel may be the average density of the pixels of the one-dimensional local area (that is, n pixels continuous on the scanning line) or the two-dimensional local area (n pixels × n pixels of the local area) centered on this pixel. .

[0011]

FIG. 1 is a block diagram of hardware as an embodiment of the present invention. In the figure, PO is a TV not shown
A video signal obtained by raster scanning the camera screen is
A grayscale image signal (for example, 8 bits) obtained by D / D conversion,
Reference numeral 1 is a frame memory for inputting the multi-value grayscale image signal PO and storing it as multi-valued screen data. Reference numeral 3 is a raster address generator for generating a raster address for the frame memory.

Reference numeral 4 is a circular address generator for generating an address for the frame memory 1 to perform image scanning in a ring shape, and 2 is a ring-shaped mask pattern for each window as shown in FIG. 3B. The stored window memory, 5 is a raster address generator for generating a raster address for the window memory 2 in the same manner as the generator 3, and 6 is a ring scan for the window memory 2 similarly to the generator 4. It is a circular address generator that generates an address for. It is possible to switch the generation of the raster address and the circular address by switching between the generators 3, 5 and 4, 6.

A window gate circuit 7 masks the multi-value grayscale signal PO or the image signal 1a read from the frame memory 1 with the mask pattern data from the window memory 2 to generate an image signal only for a designated window area. It is a circuit for passing the image signal 1a read from the PO or the frame memory 1. An image edge detection circuit 8 has a function of detecting an edge of an image, specifically, an outer end (outer peripheral point) and an inner end (inner peripheral point) of a ring-shaped high-intensity part. The signal is binarized with a predetermined threshold value for position detection and circularity inspection of the target image, and the coordinates of the rising point and the falling point of this binarized signal as an image edge are internally set. Stored in memory. 9
Is a circuit for inspecting the circularity of the coordinate values of the outer peripheral points or the inner peripheral points detected by the image edge detecting circuit 8.

Reference numeral 10 is a defect for inputting an image signal 1a read from the frame memory 1 by raster scanning of the frame memory 1 through a window gate circuit 7 and detecting a defective pixel such as a black dot using a difference or the like. A detection circuit 11 is a defect detection determination circuit that aggregates the detected defective pixels and determines a defect. Reference numeral 12 denotes a defect detection circuit for inputting the image signal 1a read from the frame memory 1 by the ring-shaped scanning of the frame memory 1 through the window gate circuit 7 and detecting defective pixels due to dents, as will be described later. Is a defect detection determination circuit that aggregates the detected defective pixels and determines a defect. Reference numeral 19 is a circuit for inputting the determination results of the circularity determination circuit 9 and the defect detection determination circuits 11 and 13 to make a comprehensive determination, 20
Is an output circuit that outputs a pass / fail according to the output determination signal of the comprehensive determination circuit 19.

Next, 16 is an X projection circuit for obtaining the X-direction projection circuit pattern of the target image using the latest multi-valued image signal PO that has passed through the window gate circuit 7, and 17 is also for obtaining the Y-direction projection pattern of the target image. Y projection circuit, 18
Is a circuit that uses the output data of these two projection circuits 16 and 17 to obtain only the region of the target container image that is not connected to another container image.

That is, when the circular container to be inspected passes a predetermined position, a stationary image to be inspected is taken in from the upper part of the container by stroboscopic light emission or exposure limitation by the shutter. During this capturing period, the X projection circuit 16 and the Y projection circuit 1
The image is sent to 7 to get the projection features. In the case of projection calculation of a fixed binary image, the projection circuits 16 and 17 have a binarizing function.

FIG. 4 shows a method of detecting a circumscribed rectangle (that is, a processing area of a shaded area) of a container by using a fixed binary image 201 of the container and its projected amount 203 in the X direction and projected amount 204 in the Y direction. When moving along the transport path while being connected to each other, the processing area determining circuit 18 separately performs processing of connection separation to determine the processing area 202. This connection separation is, for example, Japanese Patent Application No. 3-2 which is a prior application of the present applicant.
It can be performed by the process described in No. 49946.

Next, reference numerals 14 and 15 are an image edge detecting circuit and a positional deviation amount determining circuit for generating a window at a correct position with respect to the target image. FIG. 5 is an explanatory diagram of the operation of the image edge detection circuit 14. That is, the edge detection circuit 14 inputs the latest multi-value grayscale image signal PO in the same period as the above-mentioned projection circuits 16 and 17, and first, for example, to fix the center of the target image easily, for example, to fix the bottom high bright portion 2. Create a value image 210. However, the threshold used in this case is generally different from the threshold used by the image edge detection circuit 8 described above.

Next, an edge within a predetermined area of the fixed binary image 210 (in this example, the upper edge inspection area 21
1 and the coordinates of the upper end edge in 1 and the left end edge in the left end edge inspection area 212, and thus the position reference point 213.
Detect the coordinates of. The position shift amount determination circuit 15 compares the coordinates of the detected position reference point 213 with preset reference coordinates, and the shift, and accordingly the detected center position of the actual target image, and the preset window. The deviation from the center position of is detected.

Raster address generators 3 and 5 are used for each processing of the image edge detection circuit 8, the defect detection circuit 10, the image edge detection circuit 14, the X projection circuit 16 and the Y projection circuit 17 excluding the defect detection circuit 12 described above. Then, the circular address generators 4 and 6 are switched to generate a circular address, and the defect detection circuit 12 operates. Figure 6
FIG. 3 is an explanatory diagram of a method of generating a circular address. In this example, 1 is set for the side surface portion of the container 102, and thus for the image area between the mouth high brightness portion 103 and the bottom high brightness portion 104 in FIG.
Alternatively, a concentric ring-shaped scanning line (circular scanning line) L (L1 to Ln) is generated at a predetermined interval of a plurality of pixels (however, this pixel interval is determined according to the size of the dent to be detected), and scanning is performed. The circular scanning lines are switched clockwise for each circular scanning line L, and the circular scanning lines are switched in the order of L1 → L2 ... Lk. The circular address is such a circular scan line L as will be described later with reference to FIG.
X, Y of the pixels arranged at intervals of the above 1 or a predetermined plurality of pixels
It is generated by sequentially specifying coordinates in the scanning order.

As the circular scanning line, it is possible to use a spiral scanning line in which the diameter of the scanning circle sequentially changes (in this case, gradually decreases) instead of the above-mentioned concentric scanning line. 7 (A) is a container 10 shown in FIG. 7 (B).
2 shows an example of a grayscale image when there is a dent 112 on the side surface, and in this case, between A and B on the kth scan line Lk of the circular scan line, and between A'and B'on the k'th scan line Lk '. There is a dent 112 in. The scanning lines Lk, Lk '
The upper points Q and Q'represent the scanning start points on the circular scanning line, respectively.

FIG. 8 is a diagram for explaining the principle of the circular type defect detection circuit 12 of FIG. 1, and PLk of FIG. 7 is shown in FIG.
An example of a change in shading of image data on the k-th scanning line Lk is shown. The symbols Q, A, and B on the horizontal axis of FIG. 8 are shown in FIG.
(A) Corresponds to the same sign as above. Here, P0 is a target point and its pixel value, and P1 is a predetermined pixel interval d from this target point.
A background point and its pixel value.

The defect detection circuit 12 uses a predetermined threshold value THP for a window including the kth scan line Lk.
Is compared with the absolute value ΔP = | P0−P1 | of the difference between the pixel values P0 and P1. If ΔP> THP, it is determined that the pixel of interest P0 is an internal defective pixel such as a dent.

FIG. 9 is another diagram for explaining the principle of the defect detecting circuit 12, and PLk 'in the same figure shows the K'th scanning line L in FIG. 7A.
An example of a grayscale change of pixel data on k'is shown. Note that the symbols Q ', A', and B'on the horizontal axis in FIG. 9 also correspond to the same symbols in FIG. 7A. The K'th scanning line Lk 'in FIG.
Is in a low-luminance portion where light generated on the outer periphery of the bottom high-luminance portion 104 is hard to hit, but when a dent occurs in this portion, the portion corresponding to the dent becomes bright as shown in FIG. 9, and as described in FIG. Signs of similar pixel density differences P0-P1 are the opposite of those in FIG. However, in the defect detection circuit 12 in the case of FIG. 9 as well as in the case of FIG. 8, the threshold value THP determined in advance for the window including the K′th scanning line Lk ′ and the absolute value ΔP = | P0 of the above-mentioned density difference. -P1 |
If P> THP, it is determined that the pixel of interest P0 is an internal defective pixel such as a dent.

FIG. 10 is a block diagram showing an embodiment of the configuration of the circular address generator 4 and the defect detection circuit 12 of FIG. The counter control circuit 21 in the circular address generator 4 uses the X address counter 2 in the generator 4.
3, the Y address counter 24 is controlled so that the scanning lines L1, L2, ..., L described in FIG.
The X address and the Y address as the circular address of each pixel to be scanned are sequentially output to the frame memory 1 in the order of n. A scanning line counter 22 is provided in the circular address generator 4, and the value of the counter 22 is incremented by 1 each time one scanning rotation is completed.

On the other hand, in the defect detection circuit 12, the grayscale image signal (pixel value) 1a read from the frame memory 1 by the above circular address is sequentially input to the subtraction circuit 29 and the FiFo 28. This FiFo 28
Outputs the input image signal 1a to the subtraction circuit 29 after delaying the input image signal 1a by a pixel corresponding to the pixel interval d described in FIG.
Therefore, the subtraction circuit 29 directly sets the image signal 1a as the pixel value P0 of the point of interest described in FIG. 8, and sets the output pixel value of the FiFo 28 to the pixel of the background point that is separated from the point of interest described in FIG. The values P1 and P1 are input respectively to obtain the difference P0-P1 between them. The absolute value conversion circuit 30 uses this difference P0.
The absolute value of −P1 | P0−P1 | = ΔP is obtained.

The threshold value table 27 in the defect detection circuit 12 outputs the threshold value THP by using the output of the scanning line counter 22 described above as an address (that is, for each scanning line L1, L2, ..., Ln). . In this case, the threshold value THP may be set to the same value within the same window. The comparison circuit 31 compares the absolute value ΔP of the density difference with the threshold value THP, and indicates whether or not the pixel of interest is a defective pixel on the inner surface of the container such as a dent according to whether or not ΔP> THP. Output a digitized signal.

FIG. 11 shows an embodiment of the averaging circuit 26 which performs a moving average on the output pixel value P1 of the FiFo 28 of FIG. This circuit 26 is provided in series with the FiFo 28 of FIG. 10 (however, in this example, between the frame memory 1 and the FiFo 28). The averaging circuit 26 is a circuit for obtaining a moving average of four pixels that are successively input, and in this circuit 26, the input image signal (pixel value) 1a is sequentially input.
While moving to the latches 32, 33, 34, and 35, the adder 36 adds the image data of the latches 32 and 33, and the adder 37 adds the image data of the latches 34 and 35. The addition result is added by the adder 38 to complete the addition of 4 pixels, and the value obtained by shifting the addition value by 2 bits to the right by the shifter 39 (that is, the addition value is divided by 4) is averaged. Calculated output.

FIG. 12 shows an example in which the CPU 41 performs the functions of the circuits excluding the frame memory 1 and the raster address generator 3 of FIG. 1 by software. In this case, since the circular address generating circuits 4 and 6 as hardware for accessing the frame memory 1 do not exist independently, it is important to shorten the processing time of the CPU 41. For this purpose, it is effective to appropriately widen the intervals between the circular scanning lines L or appropriately widen the intervals between the pixels to be operated on the scanning lines L.

FIG. 13 shows an example of such shortening processing.
First, the data of the pixel of interest P0 on the circular scanning line L is read, then the data of the background pixel P1 which is separated from the pixel P0 by d pixels is read, and the absolute value of the density difference between these pixels | P0-P1 | Is calculated, the absolute value of the density difference is compared with a predetermined threshold value THP, and the pixel of interest P0 is binarized to determine whether it is good or bad. If good, P separated by θ
The pixel of interest is moved to 0 ′, the same processing is performed, and the entire circumference is ended, thereby ending the processing of the scanning line.

As a method of judging the quality of the defect detection judgment circuit 13, the number of defective pixels is accumulated in units of one scanning line or a certain conical circular scanning line, and the number of defective pixels is set to a predetermined area. The quality of the inner surface of the container is finally judged by comparing with the threshold value. The background pixel P used in the present invention
The pixel values of 1 are arranged on the circular scanning line L centering on the background pixel P1 as described in FIG. 11 (that is, 1
Pixels of a predetermined two-dimensional local area (for example, an area of (3 pixels) × (3 pixels) etc.) centered on the background pixel P1 instead of using the average density of a plurality of predetermined pixels of the three-dimensional local area) You may use the average density | concentration of.

[0032]

According to the present invention, an image of a circular container is sequentially scanned by a ring-shaped or spiral-shaped circular scanning line, and a density difference between pixels having a predetermined interval on this scanning line is obtained. When the absolute value exceeds a predetermined threshold value according to the position of this circular scanning line, it is determined that the inner surface of the container is defective, such as dents. It became possible to detect with high accuracy.

[Brief description of drawings]

FIG. 1 is a block diagram showing a hardware configuration as an embodiment of the present invention.

FIG. 2 is a diagram showing a high-intensity part on the inner surface of a circular container.

FIG. 3 is an explanatory view of a change in concentration on a horizontal scanning line on the inner surface of a circular container and conventional window division.

FIG. 4 is an explanatory diagram of a method for determining a container processing area.

FIG. 5 is an explanatory diagram of a method for detecting a container positional deviation amount.

FIG. 6 is an explanatory diagram of a circular address generation method according to the present invention.

FIG. 7 is a diagram showing an example of an image of a container having a dent.

8 is a diagram showing an embodiment of density change on a scanning line for explaining the principle of the circular type defect detection circuit of FIG.

9 is a diagram showing another embodiment of the density change on the scanning line for explaining the principle of the circular type defect detection circuit of FIG.

10 is a diagram showing an embodiment of the configuration of the circular address generator and the circular type defect detection circuit of FIG.

FIG. 11 is a diagram showing an embodiment of a circuit for performing a moving average on the background pixel value of FIG.

12 is a block diagram showing an embodiment of a configuration in which the circular address generation and the like shown in FIG. 1 are performed by software.

13 is an explanatory diagram of an embodiment of processing of the CPU in FIG.

[Explanation of symbols]

 PO multi-value grayscale image signal 1a multi-value grayscale image signal 1 frame memory 2 window memory 3 raster address generator 4 circular address generator 5 raster address generator 6 circular address generator 7 window gate circuit 8 image edge detection circuit 9 circularity Judgment circuit 10 Defect detection circuit 11 Defect detection judgment circuit 12 Defect detection circuit 13 Defect detection judgment circuit 14 Image edge detection circuit 15 Position shift amount decision circuit 16 X projection circuit 17 Y projection circuit 18 Processing area decision circuit 19 Overall decision circuit 20 Output Circuit 21 Counter control circuit 22 Scanning line counter 23 X address counter 24 Y address counter 26 Averaging circuit 27 Threshold table 28 FiFo 29 Subtraction circuit 30 Absolute value conversion circuit 31 Comparison circuit 32 Latch 33 Latch 34 Lara Chi 35 Latch 36 Adder 37 Adder 38 Adder 39 Shifter 41 CPU L (L1 to Ln) Circular scan line P0 Target pixel (value) P1 Background pixel (value) d Pixel interval ΔP Absolute difference absolute value THP threshold 101 Ring illuminator 102 Container 103 Mouth high brightness part 104 Bottom high brightness part 112 Hex 201 201 Fixed binarized image of container 202 Processing area 203 X-direction projection amount 204 Y-direction projection amount 210 Fixed binarized image of container 211 Top edge Inspection area 212 Left edge inspection area 213 Position reference point

Claims (3)

[Claims] 1. An inner surface side of this circular container is illuminated from the axial direction of an axially symmetric circular container, and an illumination surface of this circular container is imaged from this axial direction via a TV camera. In a circular container inner surface inspection device for analyzing deformation and dirt of the circular container, an image of the inner surface of the circular container is centered around the center of the image and has a radius of 1 or a predetermined number of pixel intervals. Scanning is sequentially performed with a ring-shaped or spiral scanning line (hereinafter referred to as a circular scanning line) that sequentially changes for each pixel, and the density of the pixel of interest as a pixel arranged at one or a plurality of predetermined pixel intervals on this circular scanning line, The absolute value of the density difference from the density of the background pixel that is a predetermined number of pixels away from the pixel of interest in a predetermined direction in either direction is compared with a threshold value determined according to the position of the circular scanning line. The above is sequentially repeated for each of the pixels of interest, and when the absolute value of the density difference larger than the threshold value is detected, the container inner surface is detected as defective. apparatus. 2. The circular container inner surface inspection apparatus according to claim 1, wherein the threshold value is set for each rotation of the circular scanning line. 3. The circular container inner surface inspection apparatus according to claim 1, wherein the density of the background pixel is an average of pixels in a predetermined one-dimensional or two-dimensional local area centered on the background pixel. Circular container inner surface inspection device characterized by the concentration.