JPS62127616A - Method and device for detecting curved surface contour image of transparent container - Google Patents

Abstract

PURPOSE:To obtain the contour image of the curved surface part of a transparent container by emitting light from behind the transparent container when it is viewed from a sensor. CONSTITUTION:A bottle 16 on which a label 19 is stuck flows on a conveyor 18. The timing sensor 14 detects the bottle 16 and a timing generating circuit 5 which receives the detection timing generates a timing signal to turn on a strobe 15 as an intense light generating means instantaneously. The image of the contour of the bottle and the label is obtained the area opposite the strobe with the light. Thus, the contour image of the transparent container is obtained automatically to automate the process of inspection, etc., and reduce the cost, save the labor, and obtain high quality.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method and apparatus for capturing an outline image of a curved surface of a transparent container such as a transparent bottle.

(Prior Art) Conventionally, capturing a contour image of a curved surface of a container made of a transparent material has been carried out when a liquid or the like is placed in the container.

However, if the container is empty and there is nothing inside, no contour image has been captured.

(Problem to be solved by the invention) In order to capture the outline of the curved surface of a hollow transparent container, a sensor is fixed at a certain location relative to the container and illumination light is applied to the container from various directions. In certain directions, light reflected from the surface of the curved portion may cause an area that is not an outline to be mistaken as an outline image.

Also, when using diffused lighting, since the material of the container is transparent,
The contour part cannot be easily captured as a contour image.

(Means for solving the problem) In order to solve the problems mentioned above, light is emitted from behind the transparent container as seen from the sensor so that the outline of the curved surface of the transparent container is captured by the sensor. A transparent container curved surface contour detection method is adopted, and for this purpose, a transparent container curved surface contour image detection device is used, which is accompanied by means for emitting light from behind the transparent container when viewed from the sensor.

(Operation) As shown in Figure 1, the sensor and the light emitting area are placed facing each other with the transparent container in the center.

The principle by which the contour image of the curved surface of the transparent container can be captured by such an arrangement will be explained next.

FIG. 2 is a top view of the components of the present invention;
The transparent container was also sliced horizontally to reveal the optical path.

The light emitted by the light generator mainly creates five paths shown in FIG. One is the case shown in Figure 2 (a), where the light reaches the sensor as is. In addition, the light that reaches the curved surface of the transparent container is divided into two parts: the light reflected on the outer wall surface of the container as shown in Figure 2 (A), the light that propagates inside the container wall as shown in Figure 2 (T), and the light that propagates inside the container wall as shown in Figure 2 (T). As shown in Figure 2 (1), there is light that is reflected on the inner wall surface of the container, and as shown, the light is divided into three parts, and only a small amount of light reaches the sensor. In addition, light passing near the center of the container is
As shown in Figure 2 (e), the light is reflected by the outer surface, and while being reflected by the surface of the second wall, it travels almost straight and passes through the two walls to reach the sensor. Therefore, the amount of light that hits the curved surface reaches the sensor due to reflection etc. is small, and most of the light from outside the container or inside the container other than the curved surface reaches the sensor. Only that part will be perceived as a dark image.

(Example) As an example of the present invention, an example will be shown in which the present invention is used to inspect the position of a label on a bottle.

In this example, a sensor simultaneously captures the outline of the bottle and the label on the bottle, and determines whether the label is placed in the correct position.

In this example, a method and apparatus for detecting a contour image of a curved surface of a transparent container are used as a method and apparatus for determining a measurement standard for a label pasting position. − This will be explained below with reference to the figures. FIG. 3 is a block diagram of a bottle label position inspection device to which the transparent container curved portion contour image detection method and device of the present invention are applied. The bottle (16) with the label (19) attached is on the conveyor (18)
Among them, the flow first flows over an inspection timing generation means consisting of a timing sensor (14) and a timing generation circuit (5). A timing sensor (14) detects a bottle (16), and upon receiving this timing, a timing generating circuit (5) generates a timing and instantaneously lights up a strobe (15) which is a strong light generating means. With this light, an image as shown in FIG. 5 is obtained in the area shown in FIG. 4 on the opposite side of the bottle from the strobe. Therefore, the outline of the bottle and the image of the label are captured.

An image sensor/camera (1) serving as a detection means takes an image of the boundary of the bottle and the edge of the label. FIG. 6 shows the operation on the imaging plane. In the horizontal direction (×
direction) is sequentially performed in the Y direction to extract a one-dimensional next series of density signals.

An A/D converter performs A,/, and D conversion on the density signal of this image for each predetermined pixel and digitizes it. The data of the digital signal is stored one after another in an image memory (3), which is an image data storage means, by a DMA control circuit (4). This image memory forms a two-dimensional matrix G, IX.

The index XY of the matrix is the address of the image memory, and the element value GXY is the density data of the image. Further, the index XY is the density data of the actual image. Also, index XY
corresponds to the coordinates (x,,y) of the actual image. Fifth
The contents of the image memory that captured the image shown in the figure are schematically shown in FIG. 7. In the figure, each square represents one image.

Squares with diagonal lines correspond to dark pixels in the actual image, and the content of the image memory is a binary value of "1".Similarly, squares without diagonal lines correspond to bright pixels in the actual image. The contents of the image memory are binary [0].

-Next, when the DM, A control circuit (4) finishes storing the image data in the image memory, the interrupt control circuit (6) issues an interrupt to the CPU (7). C, PUl, (
7) is operated by the program 6-1 loaded in the ROM (8), and uses the RAM (9) as a working area. The interrupted CPU (7) runs the test program. The distance to the edge of the bottle is determined by the number of pixels in the image memory, and this is designated as d. The minimum allowable value of d is set as drn:n, and the maximum allowable value of d is set as drnax, and these are set in advance by a program or externally. d is d, I;. If < d < d vrlay, it is considered normal; otherwise, it is considered defective. This test result may be displayed on the display (11), and if it is defective, an alarm (12) may be issued. Alternatively, an abnormality signal may be generated by the abnormality signal generation circuit (13) to blow defective products off the line with air or to stop the production line.

Next, specific operations will be explained based on the flowchart shown in FIG.

The SL-PO is emitted at the timing of the timing sensor (14), the image is transferred, the density signal of the image is A/D converted by the A/D converter (2), digitized into data, and transferred by DMA. and stores the data in the image memory (3). Immediately after that, an interrupt occurs and the inspection program runs. When accessing the image memory as shown in Figure 7,
> detects the image data corresponding to the outline of the bottle, and extracts the X coordinate of the matrix (the coordinate indicates the address) Xa
Memorize t;, . 44. Store the X-coordinate x- which corresponds to the X-coordinate of the outline 1 of the bottle in the actual image. This corresponds to the X coordinate of the edge of the label in the actual image. And (S3)
The value d corresponding to the distance from the contour of the bottle to the edge of the label is calculated using the following formula 6%.The minimum allowable value dH for d in the entire device or program
in and the maximum allowable value d,. Meals are given in advance. and(
If d measured in S4) is dmITV<d<dmay, the process proceeds to (S5); otherwise, the process proceeds to (S6).

In (S5), since the label pasting position is normal, this fact is displayed on the display (11) and the process returns to the main program.

In (S6), since the label pasting position is defective, the inspection result indicates that the product is defective on the display (11);
In (S7), the alarm (12) is issued to stop the production line, or
Alternatively, it outputs a signal to remove defective products and returns to the main program.

In order to increase accuracy, it is best to carry out measurements like this in the same way on other parts.

In addition, in the line where the next bottle is pasted, the position of the label on the bottle flowing on the conveyor was previously inspected by human eyes, but the quality of this inspection has been improved by using the device of the present invention. Workers are freed from monotonous inspection work and labor is eliminated from inspections, leading to a reduction in costs.

Note that this example shows an example in which the transparent container curved surface contour image detection method and its device are applied as a bottle label position detection method and device, but this is not limited to this example. It can be widely applied not only to recognition devices, etc.

(Effects of the Invention) 10-1001 According to the present invention, it has become possible to capture the contour image of the curved surface of a transparent container, which was difficult to capture with a conventional sensor.

By automatically capturing the contour image of a transparent container, it is possible to automate processes such as inspection, making it possible to reduce costs, save labor, and improve quality.

[Brief explanation of drawings]

FIG. 1 is a perspective view showing the constituent elements of the present invention, and FIG.
3 is a block diagram showing an embodiment of the present invention, FIG. 4 is a front view showing the image area of the object, and FIG. 6 is an explanatory diagram showing the obtained image, FIG. 6 is an explanatory diagram showing the scanning, FIG. 7 is a map showing the contents of the image memory, and FIG. 8 is an explanatory diagram showing the same scanning. There is also one flowch that shows the same action. (1)...Image sensor-camera (2)...A/D converter (3)...Image memory (4)...DMA control circuit (5)...Timing generation circuit (6) ...Interrupt control circuit (7)...CPU (8)...ROM (9)...RAM (10)...I10 interface (11)...
Display (12)... Alarm (13)... Abnormal signal generation (14)... Timing sensor (15)... Strobe (16)... Bottle (17)... Pass line (18) )...Conveyor (19)...Label (20)...Image limit frame

Claims (1)

[Claims] 1) A method for detecting a contour image of a curved surface of a transparent container in which a sensor captures the contour of a curved surface of a transparent container, characterized in that light is emitted from behind the transparent container as viewed from the sensor. A method for detecting contour images of curved surfaces of transparent containers. 2) A transparent container curved surface contour image detection device comprising a sensor that captures the contour of a transparent container's curved surface portion as an image, including means for emitting light from behind the transparent container as viewed from the sensor. A device for detecting a contour image of a curved surface of a transparent container.