JPH0611393A - Bottle color discriminating device - Google Patents

Abstract

PURPOSE:To obtain a bottle color discriminating device which can accurately discriminate each color so as to recycle used beverage bottles. CONSTITUTION:The title device is constituted of an image pickup means 3b which takes the color picture of a bottle 1, first discriminating means 30 which discriminates the presence/absence of a bottle shade after converting the color picture of the bottle 1 into a monotone picture and binarizing the monotone picture based on a first set value, and second discriminating means 31 which discriminates the presence/absence of a bottle shade after performing prescribed color emphasizing processing set at every color component of R, G, and B and binarizing the monotone picture based on second set values set at every color component.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to sorting and collecting non-combustible waste such as bottles containing a plurality of types of waste by each color such as white (transparent), brown, blue and green for recycling. The present invention relates to a bottle color identification device used in a collection device.

[0002]

2. Description of the Related Art In recent years, the importance of recycling industrial waste has been pointed out from the viewpoint of effective use of resources. Especially for bottles, if these bottles of multiple colors are mixed, the use of each bottle is limited.Therefore, it is desirable to sort and collect each color, and many bottles are sorted by color by hand. It was However, manpower is limited, and continuous work for a long time with great fatigue causes erroneous selection and is not preferable in terms of safety. Therefore, an automatic selection device has been proposed in which a color sensor is used to determine the color of a bottle and which constitutes a color identification means.

[0003]

The color of the bottle is slightly different depending on the type and amount of the contained elements.
Even if you say blue and green, the range is not clear. For example, green is chrome, blue is mixed with cobalt or copper, and tea is mixed with iron, manganese, nickel, or the like. However, since the color sensor compares the outputs of a plurality of photodetection sensors having different sensitivities of R (red), G (green), and B (blue) with respect to each wavelength by a logarithmic amplifier, a plurality of wavelength components The bottles and the like have a drawback in that it is difficult to uniquely distinguish subtly different colors. In other words, it cannot be individually attributed to different groups to be identified, and it is determined that they belong to a plurality of groups. For example, a dark green bottle will be identified not only as green but also as tea. An object of the present invention is to eliminate the above-mentioned conventional drawbacks.

[0004]

To achieve this object, the bottle color identification device according to the present invention is characterized by an image pickup means for picking up an image of a bottle and a color image of the bottle picked up by the image pickup means. , A first determination means for determining the presence or absence of a bottle shadow after converting to a monotone image and binarizing with a first set value, and a predetermined color enhancement process set for each color component of RGB to perform color It is composed of a second discriminating means for discriminating the presence / absence of a bottle shadow after performing a binarization process with a second set value set for each component.

[0005]

When the color image of the bottle picked up by the image pickup means is converted into a monotone image by the first discriminating means and binarization processing is performed with the first set value, a white image showing a relatively bright data value is obtained. It is possible to separate bottle shadows and bottle shadows of other colors that show dark data values, and if bottle shadows are not confirmed in the binarized data, it is a white bottle and bottle shadows are confirmed. If so, it can be identified as a bottle of a color other than the white bottle. Furthermore, if the color image is subjected to color enhancement processing for each component by the second discriminating means and binarization processing is performed with the second set value, the color of the bottle containing a large amount of the enhancement component and the color of other bottles are separated. The color of the bottle can be identified for each color by performing this process for the number of colors to be separated or for the number of colors to be separated by one.

[0006]

According to the present invention, it is possible to provide a bottle color identification device capable of reliably identifying a large amount of bottle waste for each color such as white, brown, blue, and green without human intervention. It has become possible to efficiently recycle glass resources.

[0007]

EXAMPLES Examples will be described below. As shown in FIGS. 1 and 2, the bottle sorting and collecting apparatus detects a reflected light from the transporting means 2 for transporting the transporting means 2 in which the bottle 1 is vertically aligned in a certain direction and is transported. Then, a color identification device 3 for identifying the color is provided, and the color identification device 3 detects the reflected light from the bottom portion 1b of the bottle 1 aligned in a certain direction. Sorting and collecting means 4 arranged on both sides and collecting bottles 1 for each color on the downstream side thereof
Is provided and configured.

More specifically, the conveying means 2 is composed of a belt type conveying mechanism 2a having a concave portion having an inverted triangular cross section as a bottle holding portion 2b on the upper surface thereof. In addition, each holding portion 2b is separately held and conveyed so as to be in a posture perpendicular to the conveying direction.

The color discriminating device 3 includes a light source 3a for irradiating the bottle 1 with white light and an image pickup means 3b formed by a CCD camera for picking up an image of the bottle shadow by the reflected light from the bottle 1 and arranged side by side with a shield plate 3c interposed therebetween. Then, an image processing unit 3d for performing various conversion processes on the image captured by the image capturing unit 3b is provided. The image processing means 3d includes the image pickup means 3b.
The color image of the bottle 1 picked up by is converted into a monotone image, binarized by the first set value, and then the first discriminating means 30 for discriminating the presence or absence of the bottle shadow, and each of the RGB color components are set. A predetermined color emphasis process is performed for each color component, and the second discriminating unit 31 for discriminating the presence or absence of a bottle shadow after performing a binarization process with a second setting value set for each component.

The color discriminating apparatus 3 is mounted at a position facing the mouth portion 1a or the bottom portion 1b of the bottle 1 to be conveyed, and is mounted on both sides of the conveying means 2 at different positions along the conveying direction one by one. A white reflecting plate 3e is provided on the front of each color discriminating device 3 with the conveying means 2 interposed therebetween. And
By detecting the bottom portion 1b of the bottle 1, the color discrimination means 3 attached to the side facing the bottom portion 1b is selectively operated.

As shown in FIG. 3, the first discriminating means 30.
Is a storage unit 30a that stores color image data represented by 8-bit RGB digital values for each pixel, which is output data from the image pickup unit 3b, and converts the color image data into a luminance signal for each pixel. Then, the first calculation unit 30b that binarizes the monotone image represented by the brightness from 0 to 255 with the first setting value 200 as a threshold, and whether or not there is any bottle shadow in the binarized image. And a first discriminating unit 30c for discriminating That is, it should be noted that the data showing the white bottle shadow and the data showing the white reflector 3e in the background are larger than the first set value 200, and the data of the other bottle colors are the first set value 200 or less. It is a thing.

The second discriminating means 31 includes the storage section 30.
Predetermined color enhancement processing set for each of the RGB color components for the color image data stored in a. n × C1- (C2 + C3) (where n is a positive number, and C1 to C3 is any of RGB. Image data corresponding to colors, for example, red R is C1, green G
If C2 is C2 and blue B is C3, to emphasize red R, use C
The value determined by subtracting the addition value of C2 and C3 from the value obtained by multiplying 1 by n is the data of the pixel. ) Is performed for each color component, and a second calculation unit 31b that performs a binarization process with a second set value (determined for each color component) set for each component, and some bottle shadow is present in the binarized image. It is configured with a second discriminating unit 30c that discriminates whether or not it exists. That is, the second discriminating unit 31 is the bottle of the color that has been emphasized when the presence of the bottle shadow is confirmed in the binarized image, and the color that has been emphasized if the presence of the bottle shadow is not confirmed. It is determined that the bottle is not a bottle, and as shown in Table 1, the above-mentioned processing is performed on each of RGB.

[0013]

[Table 1]

The sorting and collecting means 4 has a bottle collecting container 4a for each color arranged downstream of the carrying means 2 along the carrying direction, and is opposite to the bottle collecting container 4a with the carrying means 2 in between. On the side, a pushing means 4b for accommodating the bottle 1 held by the holding portion 2b in the bottle collecting container 4a is arranged, and the pushing means 4b is selected based on the discrimination result by the color discrimination device 3. It is configured to operate automatically. That is, the transporting means 2 transports the bottle 1 to the position determined by the color determining means 3,
The color discriminating means 3 discriminates the color of the bottle 1, and the bottle 1 which has been discriminated earlier in the downstream thereof is bottle-collecting container 4 of the corresponding color.
The bottle 1 is intermittently transported such that it is stopped while the pushing means 4b is operated to be accommodated in a.

Another embodiment will be described below. The first calculation unit and the second calculation unit may be configured to appropriately repeat the contraction / expansion process to remove noise at the boundary between the white reflector 3e and the bottle shadow after the binarization process. . For example, in this example, it is effective to repeat the 8-continuous contraction treatment twice and then repeat the 8-continuous expansion treatment twice. In addition, the image processing in the first arithmetic unit and the second arithmetic unit is configured to handle a plurality of pixels collectively as one block in addition to the one to be performed for each pixel, thereby improving the processing speed.

In the above embodiment, the operation formula and the binarization process as shown in Table 1 are explained, but the present invention is not limited to these values, and it is assumed that some bottles cannot be identified by these. Then, the value of the coefficient n of the color enhancement processing and the threshold value for binarization may be automatically changed. That is, when it is not possible to identify any color for the first time, the coefficient n of the color enhancement processing is appropriately adjusted and the determination is performed again. For example, it is possible to increase the coefficient of the color component of RGB which makes the bottle shadow the thinnest after binarization. These coefficient values may be automatically set with a so-called learning function based on past data.

Although it has been described that the bottom portion 1b of the bottle 1 is examined by the color discriminating device 3, the mouth portion 1a of the bottle 1 and other portions may be examined. However, in this case,
It is indispensable that the bottle 1 is not capped and that no label or the like is attached to other parts. further,
In the above embodiment, the color discrimination is performed using the reflected light of the bottle 1, but the color discrimination may be performed using the transmitted light of the bottle 1.

In the above embodiment, the color discriminating device 3 is provided at a position facing the bottle 1 on both sides of the conveying means 2 so as to detect the reflected light from the bottom portion 1b of the bottle 1 aligned in a certain direction. However, the installation position of the color discriminating means 3 is not limited to this, and it goes without saying that it can be installed at any position using an optical system such as a reflecting mirror or a lens.

In the above embodiment, color image processing is performed by the CCD.
Although the digital numerical value processing has been described corresponding to each pixel imaged by the camera, the sensor for detecting the light amount at one point of the bottle obtained through each RGB filter and the output signal of the sensor are described above. Depending on the method, it may be configured by a signal processing unit that performs analog signal processing to specify a color.

It should be noted that although reference numerals are given in the claims to make the objects of the drawings convenient, the present invention is not limited to the configurations of the accompanying drawings by the entry.

[Brief description of drawings]

[Figure 1] Overall configuration diagram of a bottle sorting and collecting apparatus

FIG. 2 is a sectional view of a main part

FIG. 3 is a configuration diagram of a main part

[Explanation of symbols]

 1 Bottle 3b Imaging means 30 First discriminating means 31 Second discriminating means

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Shinichi Kitano 100-1 Sakado, Takatsu-ku, Kawasaki-shi, Kanagawa Kanagawa Science Park (KSP) B Building 11F 1113 Kubota Electronic Technology Center Co., Ltd.

Claims (2)

[Claims] 1. An imaging means (3b) for imaging a bottle (1)
And a first determination for determining the presence or absence of a bottle shadow after converting the color image of the bottle (1) captured by the image capturing means (3b) into a monotone image and binarizing the image with a first set value. The means (30) and a predetermined color enhancement process set for each color component of R (red) G (green) B (blue) are binarized by the second set value set for each color component. A bottle color identification device comprising a second determination means (31) for determining the presence or absence of a bottle shadow after processing. 2. Imaging means (3b) for imaging the bottle (1)
And a predetermined color enhancement processing set for each color component of R (red) G (green) B (blue) on the color image of the bottle (1) captured by the image capturing means (3b). And a bottle color discriminating device configured by discriminating means (31) for discriminating the presence or absence of a bottle shadow after binarization processing with a set value set for each color component.