JPH0458373A - Method and device for sorting class/grade of shellfishes - Google Patents

Abstract

PURPOSE:To sort the shellfish, finely ranking the class of the shellfish by installing the shellfish on a transparent belt conveyor and sorting the class and grade of the shellfish with a picture processor while taking in video with an image pickup camera. CONSTITUTION:Shellfishes 10 are continuously carried by a belt conveyor 12 equipped with light transmittence, and black-and-white gradation video is prepared while the shellfishes 10 on carrying is irradiated with the light of a illumination device 18 for class measurement and the reflection video of light is hotographed by a photographic camera 28. Next, the luminance frequency distribution of the prepared black-and-white gradation video is prepared by a picture processor 40, the distribution pattern is normalized to prescribed size, and the normalized distribution pattern and the rank pattern stored in the picture processor 40 in advance are collectively operated. A weighted average value is determined from each assembly area, and an output signal discriminating the class of the shellfish 10 is transmitted to a sorting device 48. Thus, the class of a clam can be finely ranked.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method and apparatus for sorting shellfish into equal classes, and in particular, to sorting shellfish, such as clams, with black and white striped patterns according to their grading based on light and shade, and the size of their external shape. The present invention relates to a method and apparatus for sorting shellfish into classes based on their size.

[Conventional technology]

Conventionally, classification of clams, that is, classification of whether the black and white striped pattern of clams is dark or light, has been carried out visually by an inspector.

Further, class sorting of clams, that is, sorting according to the external size of clams, is carried out using a rotating drum-type mechanical automatic sorting device.

[Issues that are raised but sought to be solved]

However, conventional grading of clams is carried out by visual inspection by inspectors as described above, which has the disadvantage that variations occur and it is extremely difficult to classify clams into three or more levels.

Recently, there has been a trend in grading clams to differentiate products by grading them not only based on the depth of the surface color, but also on the shading of the striped pattern. However, the conventional grading method has the disadvantage that it is not possible to classify the grades into fine ranks.

On the other hand, in a rotary drum-type grade sorting device, the clams are rotated by the rotary drum, so there is a drawback that the clams are damaged by contact with the rotary drum.

The present invention has been made in view of the above circumstances, and provides a method for sorting shellfish into equal classes, which can finely classify clams into ranks, and sort the clams by class without damaging the clams. The purpose is to provide equipment.

[Means to solve problems]

In order to achieve the above object, the present invention irradiates the shellfish with light,
A black and white grayscale image was created by photographing the reflected light image, and the created black and white grayscale image was digitally converted in terms of both spatial area size and image brightness to create a brightness frequency distribution of the black and white grayscale image. The distribution pattern of the brightness frequency distribution is normalized to a predetermined size, and the normalized distribution pattern and a plurality of rank patterns each composed of a predetermined fuzzy set are subjected to a set operation, and t is determined by the set operation. Find a weighted average value of the total aggregation area from each aggregation area, and select the grade of the shellfish based on this weighted average value, and
A light that provides a contrast difference between the shellfish and the background is irradiated toward the shellfish, a transmitted image of the light is photographed to create a black-and-white grayscale image, and a video signal of the created black-and-white grayscale image is set at a predetermined threshold. The method is characterized in that it is binarized and converted into a binarized image, and the shellfish are classified into classes based on the total number of pixels of the shellfish image in the binarized image.

[Effect]

According to the present invention, shellfish (1) are continuously transported by a light-transmitting belt conveyor (12).Next, the light is irradiated onto the shellfish (1) being transported by a lighting device (18) for grade measurement. , a reflected image of the light is photographed by an imaging camera (28) to create a black and white grayscale image.

Next, the created black and white grayscale image is processed by an image processing device (4).
First, both the spatial area size and image brightness are digitally converted to create a brightness frequency distribution of a black and white grayscale image. Next, the distribution pattern of the luminance frequency distribution is normalized to a predetermined size, and the normalized distribution pattern and a plurality of ranks composed of a predetermined fuzzy set stored in advance in the image processing device f (4) are Then, a weighted average value of the total aggregate area is calculated from each aggregate area obtained by the set operation, and based on this weighted average value, the sorting device (48) is given the amount (1). Sends an output signal to determine the grade of.

In addition, at the same time as this grade selection, a lighting device for grade measurement (1
In step 4), the shellfish being transported (1) is irradiated with light that gives a contrast difference between the shellfish (1) and the background. Then, the transmitted image is photographed by the imaging camera (28) to create a black and white grayscale image. Then, the image processing device (4) binarizes this black-and-white gray video signal using a predetermined threshold value and converts it into a binary image. An output signal for determining the class of the shellfish (1) is transmitted to the sorting device (48) based on the total number of pixels of the shellfish (1).

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the method and apparatus for sorting shellfish into equal classes according to the present invention will be described in detail below with reference to the accompanying drawings.

FIG. 1 shows a system configuration diagram showing an embodiment of the equal class sorting device for shellfish according to the present invention.

According to FIG. 1, clams 10 are conveyed by a belt conveyor 12 in the direction of the arrow in the figure. The belt conveyor 12 is made of a translucent material such as vinyl chloride that transmits light. A pair of lower chamber lights 14.14 are installed below the belt conveyor 12. Said lower room light 14.1
4 uses a LOW fluorescent lamp which is a high frequency lighting method with 40 kHz modulation. In addition, the lower room light 14.14
A curved glue flap 16 is disposed below.

Thereby, the light emitted from the lower chamber light lamps 14.14 is reflected by the reflector 16 and is emitted in the direction of the belt conveyor 12, that is, upward.

On the other hand, the upper chamber light 18 is located above the belt conveyor 12 and across the belt conveyor 12 from the lower chamber light 14.
installed on the opposite side. The upper room light 18 is composed of a pair of fluorescent lights 20 and 22 having different lux. The lower fluorescent lamp 20 of the pair of fluorescent lamps 20 and 22 is
40W, a high frequency lighting method with 4QkHz modulation! Ignition fluorescent lamps are used. In addition, the upper fluorescent lamp 22 is a 32W high-frequency lighting method with the same <40kHz modulation.
Ring fluorescent lights are used. A ring diffuser plate 24 is disposed below the fluorescent lamp 20. Thereby, the light emitted from the fluorescent lamp 18 is transmitted to the belt conveyor 1.
The light is irradiated toward the clam 10 on top of the shell 2.

Incidentally, the irradiation timing of the fluorescent lamps 14 and 18 is controlled by a timer control device (not shown) so that the fluorescent lamps 14 and 18 are irradiated at the same time when the clams 10 being conveyed by the belt conveyor 12 are positioned above and below them, respectively.

Incidentally, a high-speed shutter camera 28 equipped with a lens 26 is fixedly installed above the upper chamber light 18.

The light receiving portion of the lens 26 is attached to face the clam 10 passing between the lower chamber light 14 and the upper chamber light 18.

Further, the high-speed shutter camera 28 is a high-speed shutter camera with a shutter speed of 1/600 seconds, and its photographing timing is controlled by the timing control device described above. Therefore, the light is controlled to be turned off at the same time as the light is emitted from the lower chamber light 14 and the upper chamber light 18. As a result, the high-speed shutter camera 28 captures the light from the lower chamber light 14, that is, the image (transparent image) including the background transmitted through the belt conveyor 12 and including the shadow of the clam 10, and the light from the upper chamber light 18. At the same time, an image (reflected image) of the black and white shading striped pattern of the clam 10 reflected by the camera can be photographed. It is important to note that in order to capture the two images mentioned above with the high-speed shutter camera 28, the lux ratio of the lower chamber light 14 and the upper chamber light 18 is set to 1.
: Must be about 1 to 10.

An analog gain amplifier 32 is connected to the high-speed shutter camera 28 via a cable 30. Further, a low pass filter 34 is connected to the analog gain amplifier 32. Further, the -1 bus filter 34 is connected to an image processing device 4 to which a CPU 38 is connected via a cable 36.
0 is connected. Furthermore, a TV monitor 44 is connected to the low-pass filter 34 via a cable 42 .

The image processing device 40 has a 256×256 screen development capability and a 6-bit (64 gradation) density resolution capability. The reason why the 6-bit density resolution capability is required is to smoothly perform density resolution for pattern measurement of a density histogram, which will be described later.

On the other hand, a sorting device 48 is connected to the image processing device 40 via a cable 46 indicated by a dotted line in the figure. This sorting device 48 selects the clams 10 conveyed by the belt conveyor 12 based on the output signal transmitted from the image processing device 40.
It is established to perform class sorting.

Next, the operation of the image processing device 400 of the shellfish equal class sorting device configured as described above will be explained.

First, the image processing device 40 receives two images taken by the high-speed shutter camera 28, that is, a transmitted image by the lower chamber light 14 and a reflected image by the upper chamber light 18, through a cable 30 and an analog gain amplifier. 7'32, low pass filter 34 and cable 36. Further, the two images are connected to an analog gain amplifier 32,
By passing through the low-pass filter 34, both the spatial area size and image brightness are digitally converted into a video signal showing a black and white gray video, and this video signal is output to the image processing device 40 via the cable 36.

Of the two black and white gray images outputted to the image processing device 40, the processing steps for the reflected image of the clam 10 obtained by the upper chamber light 18 will be described first.

FIG. 2 shows a block diagram showing the image processing steps of the image processing device 40. According to FIG. 2, first, a brightness frequency distribution (brightness histogram) shown in FIG.
Create. Next, the distribution pattern of the luminance histogram is normalized so that the maximum peak value becomes 1 and patterned into a triangular shape, as shown in FIG. 3(B). Next, this normalized distribution pattern and the image processing device 40
The intermediate fuzzy patterns of the grade ranks of clams shown in FIG. 3(C) stored in advance are subjected to set calculations. That is, the theoretical calculation shown in FIG. 3(D) is performed on the good pattern and the normalized pattern of the intermediate fuzzy pattern. Next, the theoretical calculation shown in FIG. 3(E) is performed on the superior pattern and the normalized pattern of the intermediate fuzzy pattern. Note that this normalized pattern has no overlap with the intermediate fuzzy pattern shown in FIG. 3(C), so the theoretical calculation of "excellent/<turn" is not performed.

Then, find the center value of the overlapping part found in FIGS. 3(D) and (E) between the normalized pattern and the intermediate fuzzy pattern,
Figure 3 (
Obtain as shown in F).

Then, based on the weighted average value obtained, an output signal for determining the grade of the clam 10 is transmitted to the aforementioned sorting device 48. This allows the classification of the 10 clams to be divided into fine ranks and sorted.

FIG. 4 shows an explanatory diagram in which a plurality of clams 10 are graded.

According to FIG. 4, the clams of samples 1, 4, and 7 fall under rank 1 when a theoretical calculation is performed between the brightness histograms and the aforementioned fuzzy set.

That is, the clam of sample 1.4.7 corresponds to a black clam. Further, when the histogram of the clam of sample 2.5 is similarly subjected to theoretical calculation, the clam of sample 2.5 corresponds to rank 2, and corresponds to a clam with black and light stripes. Furthermore, sample 3.8
According to the clam histogram of sample 3.8, the clam of sample 3.8 corresponds to rank 3, which corresponds to a clam of intermediate concentration. sample 6
．． According to the clam histogram of 9, this sample is 6.9
The clam corresponds to rank 4, and is equivalent to a pale clam. In addition, clam 1
In the ranking of 0, the higher the rank number, the lower the quality of the product.

Next, a process for processing the transmitted image of the clam 10 obtained by the fluorescent lamp 14 will be explained.

The transmitted image is an image in which a large contrast difference occurs between the clam 10 and the surrounding area of the clam 10, as described above. The video signal of this transparent video is transmitted to the image processing device 4.
0, the image processing device 40 first binarizes the video signal using a predetermined threshold value and converts it into a binarized image. Next, the total number of pixels of the shellfish image in the image converted into a binarized image is calculated. Next, based on the calculated total number of pixels, an output signal for determining the class of shellfish is transmitted to the sorting device. In other words, if the total number of pixels of a shellfish is 30,000 pixels or less, it is considered a small clam because the longest length of the clam is about 50 mm, as shown in Figure 5, and if the total number of pixels is around 60,000 pixels, it is considered a small clam. Since the maximum length of the clam is about 70 mm, it corresponds to a medium-sized clam. Furthermore, if the total number of pixels is 70,000 pixels or more, the maximum length of the clam is 3Q mm or more, so it corresponds to a large-sized clam, and the respective outputs are sent to the sorting device 48. Send a signal. In this way, since the class of the clam is sorted based on the total number of pixels of the clam, the class of the clam can be sorted accurately and without damaging the clam.

Further, since the image processing device 40 performs grade sorting and class sorting of the clams 10 at the same time, it is possible to accurately sort many types of clams by combining grades and classes.

〔Effect of the invention〕

As explained above, according to the method and apparatus for sorting shellfish into equal classes according to the present invention, the illumination device for equal class measurement and the illumination device for class measurement are installed with a transparent belt conveyor in between, and an imaging camera is used. Since the two images obtained by the two lighting devices are taken in and the images are used to sort out the grade and class of the shellfish using the image processing device, the grade and class of the shellfish can be sorted by finely ranking them. Furthermore, since the classification of shellfish is performed using image processing technology, classification can be performed without damaging the shellfish.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram showing an embodiment of the apparatus for sorting equal classes of shellfish according to the present invention, FIG. 2 is a block diagram of an image processing device of the apparatus for sorting equal classes of shellfish according to the present invention, and FIG. 3 is a diagram of the present invention. FIG. 4 is an explanatory diagram showing the image processing process of the image processing device of the shellfish equal class sorting device according to the present invention, FIG. FIG. 3 is an explanatory diagram showing clam classification performed using the shellfish equal class classification apparatus according to the present invention. 10...Clam, 12...Belt conveyor, 1
4.18.20.22... Fluorescent lamp, 28... High-speed shutter camera, 32... Analog gain amplifier, 34... Low pass filter, 40... Image processing device, 48... Sorting device . Applicant: Hitachi Plant Construction Co., Ltd. Figure 4 (Concentration histogram “ram pattern”)

Claims (2)

[Claims] (1) Shine light on the shellfish and take a picture of the reflection of the light to create a black-and-white grayscale image.The created black-and-white grayscale image is digitally converted in terms of both spatial area size and image brightness to create a black-and-white grayscale image. create a brightness frequency distribution, normalize the created distribution pattern of the brightness frequency distribution to a predetermined size, and collect the normalized distribution pattern and a plurality of rank patterns each composed of a predetermined fuzzy set. A weighted average value of the total aggregation area is calculated from each aggregation area obtained by the aggregation area, and the grade of the shellfish is selected based on this weighted average value. Light that provides a contrast difference is irradiated, a transmitted image of the light is photographed to create a black and white grayscale image, and the video signal of the created black and white grayscale image is binarized using a predetermined threshold value to create a binary image. 1. A method for classifying shellfish into equal classes, characterized by classifying shellfish into classes based on the total number of pixels of a shellfish image in a binarized image. (2) A belt conveyor that transports shellfish and is made of a light-transmitting material, a grade measurement irradiation device that is installed above or below the belt conveyor and irradiates light onto the shellfish being transported, and a belt conveyor that is sandwiched between a lighting device for class measurement that is installed on the opposite side of the lighting device for class measurement and irradiates the shellfish with light, a first image of the shellfish reflected by the lighting device for class measurement, and the lighting device for class measurement a camera capable of simultaneously capturing a second image of the shellfish including the background transmitted by the device; an image processing device that processes the video signal from the camera; and an image processing device that processes the video signal from the camera; a sorting device for sorting equal classes; the image processing device digitally converts the first video in terms of both the spatial area size and the brightness of the video to create a brightness frequency distribution of the black and white gray video; The distribution pattern of this luminance frequency distribution is normalized to a predetermined size, and the normalized distribution pattern and a plurality of rank patterns composed of pre-stored predetermined fuzzy sets are each subjected to a set operation, and are determined by set operations. A weighted average value of the total aggregation area is determined from each aggregation area, and an output signal for determining the grade of the shellfish is generated for the sorting device based on this weighted average value, and
The second video signal is binarized using a predetermined threshold value and converted into a binarized image, and the class of the shellfish is assigned to the sorting device based on the total number of pixels of the shellfish image in the binarized image. A device for sorting shellfish into equal classes, characterized in that it creates an output signal for discrimination.