JP2001134762A - Selector for bottles - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a selector for bottles for exactly and speedily selecting a desired bottle. SOLUTION: A bottle 1 is photographed by a CCD camera, a bottle type is identified from a distance Δx from a center G of the image (center of gravity) to a point A on the bottom o f the bottle, further, a reference point B0 is set to the prescribed point on the bottom of the bottle, and whether it is the desired bottle or not is judged from the relative positions of prescribed points B1-B6 on a contour with this reference position as an origin.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for selecting a desired bottle from bottles such as glass bottles and plastic bottles as waste.

[0002]

2. Description of the Related Art
2. Description of the Related Art There is known an apparatus that takes images of bottles conveyed on a conveyor with a CCD camera and identifies and sorts the colors of the bottles. For example, a beer bottle is identified and sorted as a brown bottle. Here, beer bottles can be reused and need to be collected. However, brown bottles other than the beer bottles are also sorted together with the beer bottles. Therefore, it is necessary to further search for and collect beer bottles from the selected brown bottles, resulting in poor collection efficiency. On the other hand, beer bottles are slightly different in shape from one maker to another, and if these are classified by maker, more time and labor are required. Thus, in the conventional bottle sorter,
It was not easy to accurately and quickly sort the desired bottle.

[0003] In view of the above-mentioned conventional problems, an object of the present invention is to provide a bottle sorting apparatus for accurately and quickly sorting desired bottles.

[0004]

According to the present invention, there is provided an apparatus for sorting bottles, comprising: a photographing apparatus for photographing bottles being conveyed as an object; and an image of the object photographed by the photographing apparatus. A reference position setting means for obtaining a center and using the centroid as a reference position; a position calculating means for calculating relative position information with respect to the reference position for a predetermined point on the contour of the image; Storage means for storing the above-mentioned position information measured in advance as selection data, and the position information calculated by the position calculation means, which is designated among the selection data stored in the storage means. When the position information calculated by the position calculating unit has the sameness as the specified sorting data, an output unit that outputs a command to take out the object. Those (claim 1).

In the bottle sorter configured as described above (claim 1), the centroid of the image of the object is determined by the reference position setting means and set as a reference position. The relative position information up to a predetermined point on the contour and the sorting data are collated by the collating means, and a desired bottle is selected.

In addition, a bottle sorting apparatus of the present invention obtains a photographing apparatus for photographing bottles conveyed as an object and a centroid of an image of the object photographed by the photographing apparatus. Reference position setting means for setting a reference position at a predetermined point on the bottle bottom of the image with reference to the centroid, and calculating relative position information with respect to the reference position for a plurality of predetermined points forming the outline of the image Position calculating means, storing means for storing the position information measured in advance for each bottle type to be sorted as sorting data, and position information calculated by the position calculating means, stored in the storage means. The collating means for collating with the specified data among the selected data for identification and all the position information calculated by the position calculating means have the same identity as the specified data for classification. It may be one and output means for outputting a take-out instruction of the object (claim 2).

In the bottle sorting apparatus configured as described above (claim 2), the reference position setting means refers to the centroid of the image of the object, and the predetermined point at the bottom of the bottle of the image is defined as the reference position. Then, the relative position information from the reference position to a plurality of predetermined points constituting the outline of the image and the sorting data are collated by the collating means, and a desired bottle is selected.

Further, the bottle sorting apparatus of the present invention obtains an image of a bottle to be transported as an object, and obtains a centroid of an image of the object captured by the image capturing apparatus. Reference position setting means for setting a reference position at a predetermined point at the bottom of the bottle with reference to the centroid;
First sorting means for calculating relative position information from the centroid to the bottom of the bottle of the image and comparing the calculated information with first sorting data stored in advance to determine the identity; Calculating relative position information up to a plurality of predetermined points constituting an outline of the image, comparing the calculated positional information with second pre-stored second selection data to determine identity. An output unit that outputs a command to take out the target object when the identity is recognized in the first sorting unit and the second sorting unit (claim 3). .

In the bottle sorter configured as described above, the first sorter is configured to determine whether the object is a desired bottle based on the relative position information from the centroid of the image of the object to the bottom of the bottle. Is determined. The second sorting means determines whether or not the target object is a desired bottle based on relative position information from a reference position at the bottom of the bottle to a plurality of predetermined points forming an outline of the image. Then, the output means outputs a command to take out the object when the first sorting means and the second sorting means recognize the identity, respectively.

[0010]

FIG. 2 is a perspective view showing a schematic configuration of a bottle sorting apparatus according to one embodiment of the present invention. In the figure, a bottle (glass bottle, plastic bottle, etc.) 1
The sheet is conveyed by the conveyor 2 from left to right in the figure.
The bottle 1 is aligned in the transport direction in the previous process,
It is assumed that the alignment posture is maintained by projections (not shown) provided continuously in the transport direction on the surface of the transport conveyor 2. Light transmission type photoelectric switches 3 are arranged at predetermined positions on both sides of the conveyor 2 to detect the arrival of the bottle 1. Further, a lamp 4 for irradiating the bottle 1 at a position detected by the photoelectric switch 3 with white light from the side, and an opposite side of the transport conveyor 2 opposed to the lamp 4, and a photographing device together with the lamp 4. Is provided. CCD camera 5
Captures the shape of the bottle 1 as a silhouette by the light of the lamp 4.

In the transport direction of the transport conveyor 2, a CCD
A robot 6 for taking out the bottle 1 is provided downstream of the camera 5. The robot 6 includes horizontally rotatable arms 6a and 6b and a vertically movable air suction device 6c. The robot 6 picks up the bottle 1 from the conveyor 2 and transfers it to a predetermined next process. The image signal output from the CCD camera 5 is input to an image processing device 7, where the shape of the bottle 1 is identified, and a take-out command is sent to a control device 8. The robot 6 is driven by the control device 8. The image processing device 7 includes a CPU,
It is configured to include a memory and the like, and performs predetermined processing shown in the flowchart of FIG.

Next, regarding the operation of the above-mentioned sorting device, the bottle 1
Is a beer bottle as an example. First, the CPU of the image processing device 7 (hereinafter, simply referred to as CPU) specifies a bottle to be sorted in step 100 of FIG. This is performed by the CPU reading the type of the bottle arbitrarily designated to the image processing apparatus 7 in advance. In FIG. 2, when the bottle 1 is transported on the transport conveyor 2, the tip of the bottle 1 is detected by the photoelectric switch 3. As a result, the CPU proceeds to step 101 in FIG.
To 102. In step 102, the CPU
The image signal of the bottle 1 is read from the CCD camera 5. Then C
In step 103, the PU performs binary processing of the image signal for each pixel of the CCD camera 5, and divides the image into white and black at a predetermined threshold. Thereby, a beer bottle that is relatively hard to transmit light is perceived as a black image in a white backlighting background, and the outline of the bottle 1 becomes clear.

FIG. 1 is a view when an image of a bottle (beer bottle) 1 obtained by subjecting an image signal of the CCD camera 5 to binary processing is viewed on a monitor screen. Although the bottle 1 shows only the outline, the bottle 1 is actually displayed in black. The image is formed in a matrix by the pixels of the CCD camera 5. Assuming that the horizontal direction in the figure is the X direction and the vertical direction is the Y direction, a straight line L extending in the X-axis direction in contact with the bottle 1 is the surface of the conveyor 2 in FIG. The conveyor 2 is also shown below the straight line L in the Y direction, but is not shown. Since the bottle 1 is always represented by a predetermined range in the X direction, it can be easily distinguished from the straight line L continuously shown in the X direction.

On the other hand, a straight line C parallel to the X axis is not based on an image taken from the CCD camera 5, but is an auxiliary line for image processing. This auxiliary line C is fixed regardless of the size of the bottle diameter. Not only in the case of a beer bottle as shown, but also in the case of a small bottle such as a drink, the Y coordinate position of the auxiliary line C is set so as to be located between the center line of the bottle in the horizontal direction and the straight line L. Is set. As a result, a predetermined point on the contour of the upper part of the bottle 1 described later (FIG. 1)
(C) in (c) can be reliably grasped. Further, since the straight line L is the surface of the conveyor 2, it is a substantially fixed fixed straight line on the image.

The CPU calculates the centroid G (centroid) of the bottle 1 in step 104 with respect to the image shown in FIG. The centroid G is obtained by calculating the center of mass of a set of black pixels representing the bottle 1 as a finite number of iso-mass systems. That is, the coordinates (Xg, Yg) of the centroid G
Represents the number of black pixels representing bottle 1 as n, and the coordinates of each pixel as (X1, Y1), (X2, Y2),. . . , (X
n, Yn), Xg = (X1 + X2 +... + Xn) / n Yg = (Y1 + Y2 +... + Yn) / n.

The centroid obtained as described above is highly accurate for an object having a small amount of light transmission such as a beer bottle or a wine bottle, but is high for an uncolored transparent bottle or a plastic bottle. Accuracy is low for objects. Subsequently, the CPU proceeds to step 105 to display the centroid G
Then, a pixel forming a boundary with the white area at the same Y coordinate as that of is searched. Although there are two pixels on the left and right, the CPU determines that the one closer to the center of gravity G is the bottom of the bottle.
Using the centroid G as a reference position, the distance Δx to the pixel at the point A at the bottom of the bottle 1 is determined (see FIG. 1B).

The value of Δx varies depending on the bottle type (bottle type). Therefore, when it is desired to sort large and medium beer bottles of a specific maker, Δx is measured in advance, and a predetermined range including an allowable error is set in the image processing device 7 as sorting data. CPU goes to step 1
At 06, it is determined whether or not the distance Δx is within a predetermined range set for the bottle specified in step 100 (first sorting). Here, if the determination result is no, the CPU jumps to step 110 and waits for the target object to pass because there is very little possibility that the target object is a desired beer bottle. When the object has passed, the CPU returns to step 101 and waits for the next object. On the other hand, if the decision result in the step 106 is YES, the CPU proceeds to the next selection process (step 107). The centroid G
Is accurately calculated, a desired bottle can be selected very accurately by the determination of YES in the above-described step 106. However, when the accuracy of the centroid G is low as described above, the determination result of step 106 may be YES even though the bottle is not the desired bottle. Therefore, a second sorting described below is further performed.

In step 107, the CPU determines a reference position B0 on the auxiliary line C having the same X coordinate as the point A at the bottom.
(See FIG. 1 (c)). Since the point A is obtained based on the centroid G and the reference position B0 is obtained based on the point A, the reference position B0 is also obtained with reference to the centroid G. However, since the Y coordinate of the reference position B0 is a constant value based on the auxiliary line C and the X coordinate is determined by the bottle bottom, the coordinates of the reference position B0 do not depend on the accuracy of the centroid G.

The CPU obtains the distance Δy1 of the point B1 represented by the pixel of the black region existing at the top of the X coordinate position at a predetermined distance Δx1 from the reference position B0 in the X direction with respect to the reference position B0. Similarly, predetermined distances Δx2, Δx3, Δx4, Δx5, Δ
For a plurality of points B2, B3, B4, B5, and B6 that have X coordinates that are x6 apart and that form the upper contour of the image of bottle 1, each distance Δy in the Y direction from reference position B0.
2, Δy3, Δy4, Δy5, Δy6 are determined. In the image processing device 7, the distance in the Y direction with respect to the distance in the X direction is measured in advance for each bottle to be sorted, and is stored in the image processing device 7 as sorting data. Therefore, the CPU
By comparing the distances Δy1 to Δy6 determined as described above with the stored values for the bottle specified in step 100, the distances Δy1 to Δy6 are set within the predetermined allowable error range with the stored values. It is determined whether they match (step 1
08) If all match, the identity is recognized.
The target object is specified as a specified bottle type. Therefore, C
The PU outputs a removal command to the control device 8 in order to remove the object (Step 109). On the other hand, if any one does not match, the CPU proceeds to step 10
Jump from 8 to 110 and wait for the passage of the object. If the object passes, the CPU proceeds to step 1
It returns to 01 and waits for the arrival of the next object.

According to the above-described second sorting, relative position information on the outline of the bottle is checked at a plurality of locations with reference to the reference position B0 at the bottom of the bottle. Once identified, the desired bottle can be sorted. In particular,
Since the shapes of the points B3 to B5 of the beer bottles are clearly different depending on the maker, a desired bottle can be accurately selected by the above check. In the above example, B1
I checked about 6 points of ~ B6, but not limited to 6 points,
Needless to say, it can be increased or decreased as needed. Also,
As described above, since the reference position B0 is not affected by the accuracy of the centroid G, even if the accuracy of the centroid G is not good, the bottles can be accurately sorted without being affected by the accuracy.

In the above embodiment (FIG. 1), the bottle 1 may be upside down, that is, the right side is the bottom of the bottle. The shorter distance from the centroid G is defined as the distance Δx. Thus, sorting of bottles can be performed in the same manner. In addition,
In this case, Δx, Δx1, Δx2, etc. are set in the opposite sign direction to that in the above embodiment.

In the above-described embodiment, the second sorting is performed in addition to the first sorting. However, for example, in a line where only beer bottles are conveyed, the accuracy of the centroid G is always high. Can be omitted. The flow chart in this case is as shown in FIG.
7 and 108 are omitted. On the other hand, it is also possible to omit the first sorting and perform only the second sorting. In this case, the CPU proceeds from step 104 to step 10 in FIG.
Jump to 7.

FIG. 5 is a block diagram showing the internal functions of the image processing apparatus 7 operating as described above. In the figure, the image processing apparatus 7 includes a position detecting unit 71, a collating unit 72, an output unit 73, a reference position setting unit 74, and a storage unit 75. The reference position setting means 74 obtains a centroid G of the image of the object photographed by the CCD camera 5 (step 104), and sets this centroid G as a reference position in the first sorting. Further, the reference position setting means 74 further sets a reference position B0 at a predetermined point on the bottle bottom of the image with reference to the centroid G (step 10).
7). In the first sorting, the position calculating means 71 calculates relative position information of the predetermined point A on the contour of the image with respect to the reference position (G) (step 105). Further, in the second sorting, the position detecting means 71 calculates relative position information with respect to the reference position B0 for a plurality of predetermined points B1 to B6 constituting the outline of the image (step 107). The storage means 75 stores the above-described position information measured in advance for each bottle type to be sorted as sorting data. The collation means 72 collates the position information calculated by the position calculation means 71 with the designated data among the sorting data stored in the storage means 75 (step 10).
6, 108). The output means 73 outputs the position information calculated by the position calculation means 71 in the first selection.
If it has the same identity as the specified sorting data, a command to take out the object is output (step 106 in FIG. 4,
109). In the second sorting, the output unit 73
If all the position information calculated by the position calculation means 71 has the same identity as the specified sorting data, a command to take out the object is output (step 10 in FIG. 3).
8, 109).

In the above embodiment, the bottle 1 is photographed as a silhouette from the side as shown in FIG. 2, but the CCD camera 5 is arranged above the conveyor 2 to photograph the bottle 1 from directly above. Is also good. In this case as well, sorting can be performed similarly by checking a predetermined point on the outline of the bottle 1 as shown in FIG. However, it is desirable to irradiate the bottle 1 with light from many directions so that the shape of the bottle 1 stands out.

[0025]

The present invention configured as described above has the following effects. According to the bottle sorter of the first aspect, the shape of the object is identified by checking the position information of a predetermined point on the contour with the centroid of the image of the object as the reference position, and the desired bottle is identified. Be sorted out. Since the position information up to the predetermined point with the centroid as the reference position differs depending on the bottle type, accurate and quick sorting can be performed.

According to the bottle sorter of the second aspect, relative position information up to a plurality of predetermined points forming the contour of the image is checked with the predetermined point at the bottom of the bottle of the image of the object as a reference position. Thereby, the shape of the object is identified, and a desired bottle is selected. Since the position information up to the plurality of predetermined points with the bottle bottom predetermined point as the reference position differs depending on the bottle type, accurate and quick sorting can be performed. In addition, although the centroid is referred to when setting the reference position, since the reference position is at a predetermined point on the bottom of the bottle, accurate sorting can be performed without being affected by the accuracy of the centroid.

According to the bottle sorting apparatus of the third aspect, the first sorting is performed based on the relative position information from the centroid of the image of the object to the bottom of the bottle, and at the same time, from the reference position of the bottom of the bottle. The second sorting is performed based on relative position information up to a plurality of predetermined points forming the outline of the image. Therefore, a desired bottle can be sorted very accurately and quickly by performing two types of shape sorting with different sorting criteria.

[Brief description of the drawings]

FIG. 1 is a view showing a case where an image of a bottle (beer bottle) obtained by subjecting an image signal of a CCD camera to binary processing in a bottle sorting apparatus of the present invention is viewed on a monitor screen.

FIG. 2 is a perspective view showing a schematic configuration of the bottle sorter.

FIG. 3 is a flowchart showing processing contents of an image processing apparatus in the bottle sorter.

FIG. 4 is a flowchart showing another processing content of the image processing device in the bottle sorter.

FIG. 5 is a block diagram showing an internal function of an image processing device in the bottle sorter.

[Explanation of symbols]

 Reference Signs List 1 bottle 4 lamp 5 CCD camera 6 lamp 7 image processing device 71 position calculating means 72 collating means 73 output means 74 reference position setting means 75 storage means B0 reference position G centroid (centroid)

 ──────────────────────────────────────────────────続 き Continued on the front page F-term (reference) 2F065 AA17 AA52 BB05 DD03 DD06 FF02 FF42 HH02 HH13 HH17 JJ03 JJ26 PP15 QQ05 QQ21 QQ24 QQ32 QQ38 RR07 TT03 3F079 AD12 CA18 CA21 CA23 CB25 CB29 DC02 ACO3 DC16 DC33 5L096 BA08 CA02 EA43 FA06 FA60 FA66 HA09

Claims (3)

[Claims] 1. A photographing apparatus for photographing bottles conveyed as an object, and a reference position of an image of the object photographed by the photographing device, and using the centroid as a reference position. Setting means; position calculating means for calculating relative position information with respect to the reference position for a predetermined point on the contour of the image; and storing the above-mentioned position information measured in advance for each bottle type to be sorted as sorting data. A storage unit that performs the position information calculation performed by the position calculation unit; and a comparison unit that compares the position information calculated by the position calculation unit with a designated one of the selection data stored in the storage unit. An output unit that outputs a command to take out the object when the positional information has the same identity as the specified sorting data. 2. An imaging device for photographing bottles conveyed as an object, and a centroid of an image of the object photographed by the imaging device is determined, and the centroid of the image is determined with reference to the centroid. Reference position setting means for setting a reference position at a predetermined point on the bottom of the bottle; position calculating means for calculating relative position information with respect to the reference position for a plurality of predetermined points forming the contour of the image; Storage means for storing the position information measured in advance for each of the pieces of data as selection data, and the position information calculated by the position calculation means, specified in the selection data stored in the storage means A collating means for collating the object with the object; and outputting, when all the position information calculated by the position calculating means have the sameness as the designated sorting data, a command to take out the object. Sorting apparatus of bottles, characterized in that an output means. 3. A photographing device for photographing bottles conveyed as an object, and a centroid of an image of the object photographed by the photographing device is determined, and the image is referred to with reference to the centroid. Reference position setting means for setting a reference position at a predetermined point on the bottom of the bottle, and calculating relative position information from the centroid to the bottom of the bottle of the image, and comparing the calculated position information with the first sorting data stored in advance. A first selection unit for determining the identity and relative position information from the reference position to a plurality of predetermined points constituting the contour of the image, and calculating them based on a second selection stored in advance. A second sorting unit that determines the identity by collating with the data for use, and when the first sorting unit and the second sorting unit determine the identity, respectively, outputs a command to take out the object. Output means. Sorting apparatus of bottles to be.