JP2023113282A - Number measurement device and number measurement method of granular product, as well as inspection device and inspection method of granular product - Google Patents

Abstract

To provide a measurement device which can correctly measure the number of granular products even when the shape of the granular product is in the square shape in a photographed image and the plurality of granular products are adjacent to each other.SOLUTION: A number measurement device of a granular product according to the present invention comprises: illumination means which irradiates a plurality of granular products with light; imaging means which images the plurality of granular products irradiated by the illumination means; and number measurement means which measures the number of granular products existing in a photographed image. The number measurement means executes processing of extracting a granular product region in which one granular product exists or the plurality of granular products exist so as to be adjacent to each other in the photographed image captured by the imaging means, processing of calculating an area of each granular product region, and processing of estimating the number of granular products existing in each granular product region on the basis of the area of each granular product region and the area distribution of the granular product region obtained in advance from the past photographed image.SELECTED DRAWING: Figure 6

Description

The present invention relates to an apparatus and method for counting granular products and an apparatus and method for inspecting granular products.

In the inspection of granular products, a method of taking an image of the granular product and judging the quality of the granular product from the captured image is widely used.

For example, there is an inspection apparatus and an inspection method for determining whether oxidative deterioration or thermal decomposition of granular resin is progressing, and in a captured image, granular resin whose brightness value is equal to or higher than a standard is judged as a defective product (Patent Document 1). In this inspection, an image of granular resin arranged on a plane is taken, and each granular resin is extracted from the taken image.

In inspection of granular products, it is important to accurately measure the number of granular products for quality control of granular products. For example, in many cases, there is a standard for judging the quality of granular products, such as "the number of defective products should be n or less for N granular products".

However, when the granular resins are arranged adjacently, it is difficult to distinguish and extract the granular products one by one from the captured image, which also makes it difficult to accurately measure the number of granular resins. Japanese Patent Laid-Open No. 2002-200003 does not disclose a method for coping with the case where granular resins are arranged adjacent to each other.

In Patent Document 2, even if the granular products are adjacent, a region of the granular product is extracted from the captured image of the granular product, and a plurality of reference points are distributed and arranged along the contour of the region, and each reference A method is disclosed for determining that a region contains a plurality of granular products when a straight line connecting the points intersects the contour of the region.

JP 2017-20790 A JP 2006-234519 A

However, in the method disclosed in Patent Document 2, when a plurality of granular products appearing in a square shape when photographed are lined up with their sides in contact with each other, the plurality of granular products are judged to be one. When photographing a granular product, it frequently appears as a square, so the method of Patent Document 2 cannot accurately measure the number of granular products.

The present invention provides a measuring device and method capable of accurately measuring the number of granular products even when the shape of the granular product is square in a captured image and a plurality of granular products are adjacent to each other, and the measuring device and measurement thereof. An inspection apparatus and inspection method using the method are provided.

The present invention for solving the above problems is a device for capturing an image of a granular product whose management surface looks like a square and counting the number of the granular product from the captured image,
illuminating means for irradiating light onto the plurality of granular products arranged without overlapping with the management surface facing the imaging direction;
an imaging means for imaging the plurality of granular products illuminated by the illumination means;
a number counting means for counting the number of the granular products present in the captured image;
The number counting means is
A process of extracting a granular product region in which one granular product exists or a plurality of granular products exist adjacently in the captured image captured by the imaging means;
a process of calculating the area of each granular product region;
a process of estimating the number of the granular products present in each granular product region based on the area of each granular product region and the area distribution of the granular product region obtained in advance from past captured images; conduct.

In the device for measuring the number of granular products of the present invention, the number measuring means comprises:
Extracting a plurality of mountain shapes in the area distribution of the granular product region, performing processing for making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of granular products, the number of granular products existing in each granular product region may be estimated as the number corresponding to the mountain shape to which the area of the granular product region belongs.

The apparatus for counting the number of granular products of the present invention may perform processing for calculating the number of the granular products present in the captured image based on the estimated number of granular products.

In the apparatus for counting the number of granular products of the present invention, resin granular products are targeted as the granular products, the illumination means irradiates light including an ultraviolet wavelength band, and the imaging means receives a wavelength band of 400 to 620 nm. A light-receiving element group including light-receiving elements provided in a region may be provided, and an image including the granular product fluorescent by the light irradiated by the illumination means may be captured.

Further, the present invention for solving the above problems is an apparatus for capturing an image of a granular product whose control surface looks like a square, and inspecting the granular product from the captured image,
Using the device for counting the number of granular products of the present invention,
The number measuring means is
performing a process of calculating the area occupied by the non-defective granular products and the area occupied by the defective granular products in each granular product region;
In the process of estimating the number of granular products,
estimating the number of the good granular products present in each granular product region based on the area occupied by the good granular products in each granular product region and the area distribution of the granular product regions;
The number of defective granular products present in each granular product region is estimated based on the area occupied by the defective granular products in each granular product region and the area distribution of the granular product region.

In the granular product inspection device of the present invention, the number counting means is
Extracting a plurality of mountain shapes in the area distribution of the granular product region, performing processing for making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of granular products,
estimating the number of the non-defective granular products present in each of the granular product regions as the number corresponding to the mountain shape to which the area occupied by the non-defective granular products belongs;
The number of the defective granular products present in the granular product region may be estimated as the number corresponding to the mountain shape to which the area occupied by the defective granular products belongs.

In the granular product inspection apparatus of the present invention, the number counting means measures the number of the acceptable and defective granular products present in the image based on the estimated number of acceptable and defective granular products. A process of calculating the number may be performed.

Further, the present invention for solving the above problems is a method for capturing an image of a granular product whose control surface looks like a square, and counting the number of the granular product from the captured image, comprising:
arranging the plurality of granular products so that the management surface faces the imaging direction without overlapping;
irradiating the plurality of granular products with light;
imaging the plurality of granular products irradiated with light;
extracting a granular product region in which one granular product exists or a plurality of granular products exist adjacently in the captured image;
The area of each granular product region is calculated, and the number of the granular products present in each granular product region is calculated based on the area of each granular product region and the area distribution of the granular product region obtained in advance from past captured images. to estimate

The method for counting the number of granular products of the present invention comprises:
Extracting a plurality of mountain shapes in the area distribution of the granular product region, making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of the granular products, the number of the granular products existing in the granular product region may be estimated as the number corresponding to the mountain shape to which the area of the granular product region belongs.

In the method for counting the number of granular products of the present invention, the number of granular products present in the captured image may be calculated based on the estimated number of granular products.

In the method for counting the number of granular products of the present invention, the granular products are made of resin, the plurality of granular products are irradiated with light containing an ultraviolet wavelength band, and a light receiving element having a wavelength band of 400 to 620 nm in a light receiving region is used. An image containing the granular product fluorescent by the irradiated light may be captured by using a group of light receiving elements including.

Further, the present invention for solving the above problems is a method for capturing an image of a granular product whose management surface looks like a square, and inspecting the granular product from the captured image, comprising:
Using the method for counting the number of granular products of the present invention,
calculating the area occupied by the good granular products and the area occupied by the defective granular products in the granular product region;
In the process of estimating the number of granular products,
estimating the number of the good granular products present in each granular product region based on the area occupied by the good granular products in each granular product region and the area distribution of the granular product regions;
The number of defective granular products present in each granular product region is estimated based on the area occupied by the defective granular products in each granular product region and the area distribution of the granular product region.

The method for inspecting granular products of the present invention comprises:
Extracting a plurality of mountain shapes in the area distribution of the granular product region, making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of granular products,
estimating the number of the non-defective granular products present in each of the granular product regions as the number corresponding to the mountain shape to which the area occupied by the non-defective granular products belongs;
The number of the defective granular products present in the granular product region may be estimated as the number corresponding to the mountain shape to which the area occupied by the defective granular products belongs.

In the granular product inspection method of the present invention, the number of non-defective and defective granular products existing in the captured image may be calculated based on the estimated number of non-defective and defective granular products. good.

The term "management surface" in the present invention means a surface that can be checked when the granular product is viewed from one direction, and is used when judging the quality of the granular product. For example, in a granular product, it is the surface with the largest area, etc., depending on the granular product.

In the present invention, "the management surface faces the imaging direction" means that when the management surface of the granular product is irradiated with light from the illumination means, the granular product is arranged so that the reflected light can be acquired by the imaging device. It is that the management aspect of

According to the present invention, even if the shape of a granular product is square in a captured image and a plurality of granular products are adjacent to each other, the number of granular products can be accurately measured, and the quality of the granular product can be accurately determined. can.

1 is a schematic configuration diagram of one embodiment of an inspection apparatus for granular products of the present invention, which is equipped with a device for counting the number of granular products of the present invention. An example of a granular product. It is an example of the flowchart which counts the number in this invention. It is an example of the region extraction processing of the granular product in the present invention. It is an example of the area distribution of the granular product region in the present invention. It is an example of number estimation processing in the present invention. It is an example of quality determination processing in the present invention.

A mode for carrying out the present invention (hereinafter referred to as "this embodiment") will be described with reference to the drawings. The present invention is not limited by the embodiments described below.

FIG. 1 is a schematic diagram showing an example of an inspection apparatus according to this embodiment. The inspection apparatus A in this embodiment comprises an arrangement means 2, an illumination means 3, an imaging means 4 and a number counting means 5. As shown in FIG.

Granular product 1 is chips of polymer resin. By irradiating the granular product 1 with an ultraviolet wavelength band, the granular product 1 emits fluorescent light, and the quality of the granular product 1 itself can be judged based on the degree of this emission. This granular product 1 has the shape shown in FIG. 2, and its management surface is rectangular when viewed from the imaging direction.

After a large number of granular products 1 are placed on the placing means 2, a vibrating mechanism (not shown) provided for the placing means 2 is operated to vibrate the placing means 2 minutely. By doing so, a plurality of granular products 1 can be arranged so that their control surfaces face the imaging direction and do not overlap each other. It should be noted that the arrangement may be performed manually as described above without using the arrangement means 2 .

Light is emitted from the illumination means 3 toward the plurality of granular products 1 thus arranged. In this embodiment, the illumination means 3 emits light in the ultraviolet wavelength band.

Next, the imaging means 4 captures images of the plurality of granular products 1 irradiated with light to obtain captured images. In this embodiment, the imaging means 4 is an area camera including a light-receiving element having a light-receiving region of a wavelength band of 400 to 620 nm capable of detecting fluorescence emission of the granular product 1 with high sensitivity.

This captured image is transmitted to the number counting means 5, and the number counting means 5 counts the number of the granular products 1 in the captured image. The number measuring means 5 includes area extracting means, quality determining means, area calculating means, number estimating means, and number calculating means (all not shown).

First, in the region extracting means, a binarized image is obtained by binarizing the captured image so that the background is black and the granular product 1 is white. When a plurality of granular products 1 are adjacent to each other, individual granular products 1 cannot be distinguished from each other in the binarized image, and appear as a mass larger than the granular product 1.例文帳に追加Therefore, in the present invention, both a single granular product 1 and a mass of a plurality of adjacent granular products 1 are each referred to as a "granular product region".

Subsequently, the quality judgment means judges the granular product 1 having a luminance equal to or higher than the threshold value in the captured image in each granular product region as a defective product.

Subsequently, the area calculating means calculates the area of each granular product region determined to be non-defective by the quality determining means and the area of each granular product region determined to be defective.

Subsequently, the number estimating means estimates the number of non-defective or defective granular products 1 in each granular product region based on the area distribution of the granular product region calculated in advance from past inspection results.

Finally, the number calculating means calculates the total number of granular products 1 in the image, the total number of granular products 1 judged to be defective, and the total number of granular products 1 judged to be non-defective. If the total number of the granular products 1 determined as defective products is equal to or greater than the quality determination criteria, the inspection lot is determined as defective products.

The granular product 1 is not limited to a polymer resin chip, and may have a shape in which the management surface is a square when viewed from the imaging surface. For example, shapes such as those shown in FIG. 2, cylinders, rectangular parallelepipeds, and cubes having side surfaces as management surfaces are conceivable. Moreover, it is preferable that the standard deviation of the area of the control surface of the granular product 1 in the captured image is ±10% or less of the assumed reference area of the control surface. Here, the reference area is the area of one granular product 1 in the captured image when the granular product 1 designed to have a certain size is captured. By being ± 10% or less, when estimating the number of granular products in each granular product region from the area distribution according to the number measurement flowchart of Fig. 3 described later, the regions of 1 and 2 granular products are detected with 99% accuracy. It can distinguish between 2 and 3 granular products with 96% accuracy.

When placing the granular products 1, the placing means 2 places the granular products 1 by, for example, shaking the placing device 2 by machine vibration to place the granular products 1, or shaking the placing device 2 manually by an operator to place the granular products 1. good too. The arranging means 2 may be conveyed in one direction like a belt conveyor, or may remain in place.

The light emitted from the illumination means 3 may contain ultraviolet light. Also, for the lighting means 3, for example, a halogen lamp, an LED, a fluorescent lamp, a heating lamp, or the like may be used. Moreover, it is preferable that the light emitted from the illumination means 3 be uniformly emitted.

The imaging means 4 may be, for example, a monochrome camera composed of a light receiving element having a single light receiving sensitivity, or a color camera composed of a plurality of light receiving elements having different light receiving sensitivities. Also, a line sensor camera, an area sensor camera, or the like may be used. Also, it is preferable that the distortion due to the angle of view of the camera is small.

A detailed processing flow in the number counting means 5 will be described with reference to the flow chart of FIG. 3 and FIG.

First, a captured image is acquired by the imaging device 4 (step S1).

Subsequently, the granular product area is extracted from the picked-up image by the area extracting means (step S2). In step S2, for example, the granular product area may be binarized with white and the background black to extract the granular product area. FIG. 4 shows an example of extraction of the granular product area. FIG. 4 shows the result of extracting the granular product area by binarization when one granular product is distributed and two or three granular products are arranged adjacently. there is Reference numeral 1a is the granular product area of one granular product. As can be seen from the binarized image of FIG. 4, when a plurality of granular products are adjacent to each other, it is difficult to distinguish individual granular products in the binarized image.

Subsequently, the quality determination means determines whether the granular product region is good or bad (step S3). In step S3, for example, the binarized image obtained in step 2 may be masked on the original image to extract a mask image in which the background is black and the granular product region is the original luminance value. In addition, when the brightness on the imaged image of a defective granular product is a feature that is higher than that of a non-defective product, a brightness threshold that can distinguish between a non-defective product and a defective product is used. A granular product region may be extracted as a non-defective product, and a granular product region with a luminance threshold value or higher may be extracted as a defective product. FIG. 5 shows the determination result of non-defective and defective granular products. The left side of FIG. 5 shows the results of non-defective products, and the right side shows the results of defective products.

Subsequently, the area of each extracted granular product region is calculated by the area calculating means (step S4). In step S4, the area of each granular product region can be obtained by, for example, performing particle analysis on each of the extracted non-defective and defective granular product regions.

Subsequently, the number estimating means calculates the area distribution obtained in advance from the past captured images, and the area of each granular product region determined as a non-defective product and each granular product region determined as a defective product calculated in step S4. Based on this, the number of non-defective and defective granular products 1 constituting each granular product region is estimated (step S5).

First, referring to FIG. 6, a method of setting an area range necessary for estimating the number of objects using an area distribution obtained in advance from past captured images will be described. FIG. 6 is an example of a graph showing the area distribution of each granular product region in past inspection results, where the vertical axis indicates the number of granular product regions and the horizontal axis indicates the area of the granular product region. A mountain shape is extracted from this area distribution. A mountain shape G1 in FIG. 5 indicates a granular product region in the case of one granular product 1, that is, an area distribution per one granular product 1. As shown in FIG. The mountain shape G2 shows the area distribution of the granular product region when two granular products 1 are adjacent, and the mountain shape G3 shows the area distribution of the granular product region when three granular products 1 are adjacent. The mountain shapes G1 to G3 can be obtained by fitting a mixed normal distribution to the area distribution of the granular product region. Next, the mountain shapes G1 to G3 are made to correspond to the number of granular products 1, respectively. Specifically, the area range of the mountain shapes G1 to G3 corresponding to the number of granular products 1 is set as follows. The area range of one granular product 1 is set as the average area ±3σ of the mountain shape G1. The area range of the two granular products 1 is set such that the lower limit is the average area of the mountain-shaped G1+3σ, and the upper limit is the average area of G2+2σ. The area range of the three granular products is set as greater than the average area of the chevron G2 + 2σ.

Next, referring to FIG. 7, a method of determining the number of granular products 1 included in the granular product regions 1b, 1c, and 1d is described assuming that the number is unknown. The upper part of FIG. 7 shows images of the granular product areas determined as non-defective products and defective products obtained in step 4. The lower part of FIG. 7 shows a graph of the area distribution and area range determined from the past inspection results. A vertical dotted line in the graph indicates the area of each of the granular product regions 1b, 1c, and 1d for the total number of pixels. Since the area corresponding to the total number of pixels of the granular product 1b determined to be a non-defective product is within the area range of one granular product, the number of non-defective granular products 1 is estimated to be one. Since the total pixel area of the granular product region 1c determined to be defective exists within the area of two granular products, the number of defective granular products 1 is estimated to be two. Similarly, since the total pixel area of the granular product region 1d determined as a non-defective product exists within the area of three granular products, the number of non-defective granular products 1 is estimated to be three.

After step S5, the number of times of estimation is increased by 1 (step S6).

After step S6, the estimated number of times is compared with the number of all granular product areas in the captured image (step S7). If the estimated number is less than the number of granular product regions (step S7: No), the process returns to step S5 to repeat the process of estimating the number. By repeating steps S5 and S6, the number of non-defective and defective granular products 1 contained in all granular product regions can be estimated.

When the estimated number of times reaches the number of all the granular product areas in the captured image (step S7: Yes), the number calculating means adds up the estimated number of granular products 1 included in each granular product area (step S8). . Table 1 below shows an example of the number of non-defective products and defective products of the granular product 1 estimated for each granular product region 1b, 1c, 1d, and the total number of granular products.

After step S8, the same number calculating means calculates the estimated number of granular products 1 included in the captured image (step S9).

When extracting the granular product area 1a from the captured image in step S2 described above, an area that is less than a predetermined set area may be removed as noise.

The area distribution of each granular product region in the past inspection results in step S5 described above is not limited to the area distributions G1 to G3. There is

In step S5 described above, the setting of the area range may be determined by a person.

In the estimation of the number of granular products 1 in step S5 described above, a granular product area smaller than the lower limit of the area range of one granular product 1 may be determined as noise and removed.

When estimating the number from the area of each granular product region 1a in step S6 described above, if many granular products 1 are adjacent to each other, that is, if the area of the granular product region 1a is too large, the management surface of the granular products 1 The standard deviation of the area influences the number estimation of the granular products 1 and increases the possibility of erroneous estimation. Therefore, it is preferable to arrange the granular products 1 in the arrangement means such that the number of adjacent granular products 1 is reduced.

Further, in the processing flow of the number measuring means 5 described above, the area extracting means performs step S2, the area calculating means performs steps S3 and S4, the quality determining means performs step S5, the number estimating means performs step S6, and the number calculating means performs the processing of steps S9 and S10, but is not limited to this, and a series of processing in which the number measuring means 5 counts the number of non-defective and defective granular products 1 existing in the captured image. should be done. For example, instead of the area calculation means, quality determination means may perform the processing of step S4.

Since the inspection apparatus A of the present invention estimates the number of granular products 1 included in the captured image in step S9, it also functions as a number estimation apparatus. In the description of the processing flow of the number measuring means 5, the number of each of the non-defective and defective granular products 1 contained in the granular product region is estimated through steps S3 and S5. If the only thing is the number of granular products 1 present in the captured image, there is no need to estimate the number of each of good and bad granular products 1 . In this case, step S3 may be skipped. Then, in step S5 described above, the number of granular products 1 existing in the granular product region may be estimated based on the total number of pixels occupied by the granular product region.

In the pass/fail determination means in the number counting means 5 described above, the pass/fail determination is not limited to the magnitude of luminance in the image. For example, if the surface of a defective granular product differs in color from that of a non-defective product, a specific color region in the image may be determined to be defective, and the rest may be determined to be non-defective.

INDUSTRIAL APPLICABILITY The inspection apparatus and inspection method for granular products of the present invention can be used to inspect granular products whose management surface is rectangular in the captured image when the control surface for quality control is imaged.

A Inspection device (counting device)
1 Granular Products 1a, 1b, 1c Granular Product Region 2 Placement Means 3 Illuminating Means 4 Imaging Means 5 Counting Means G1, G2, G3 Mountain Shape in Area Distribution

Claims (14)

A device for capturing an image of a granular product whose management surface looks like a square and counting the number of the granular product from the captured image,
illuminating means for irradiating light onto a plurality of the granular products arranged without overlapping with the management surface facing the imaging direction;
an imaging means for imaging the plurality of granular products illuminated by the illumination means;
a number counting means for counting the number of the granular products present in the captured image;
The number counting means is
A process of extracting a granular product region in which one granular product exists or a plurality of granular products exist adjacently in the captured image captured by the imaging means;
The granular product existing in each granular product region based on the processing of calculating the area of each granular product region and the area distribution of the granular product region obtained in advance from the area of each granular product region and the past captured image perform a process of estimating the number of products,
Counting device for granular products. The number counting means is
Extracting a plurality of mountain shapes in the area distribution of the granular product region, performing processing for making each of the extracted mountain shapes correspond to the number of the granular products,
in the process of estimating the number of granular products, estimating the number of granular products present in each granular product region as the number corresponding to the mountain shape to which the area of the granular product region belongs;
The device for counting the number of granular products according to claim 1. 3. The apparatus for counting the number of granular products according to claim 1, wherein said number counting means performs a process of calculating the number of said granular products existing in said captured image based on said estimated number of granular products. . Targeting resin granular products as the granular products,
The illumination means irradiates light containing an ultraviolet wavelength band,
The imaging means has a light-receiving element group including a light-receiving element having a light-receiving region of a wavelength band of 400 to 620 nm, and captures an image containing the granular product fluorescent by the light irradiated by the illumination means.
The device for counting the number of granular products according to any one of claims 1 to 3. An apparatus for capturing an image of a granular product whose management surface looks like a square and inspecting the granular product from the captured image,
Using the device for counting the number of granular products according to any one of claims 1 to 4,
The number counting means is
performing a process of calculating the area occupied by the non-defective granular products and the area occupied by the defective granular products in each granular product region;
In the process of estimating the number of granular products,
estimating the number of the good granular products present in each granular product region based on the area occupied by the good granular products in each granular product region and the area distribution of the granular product regions;
estimating the number of defective granular products present in each granular product region based on the area occupied by the defective granular products in each granular product region and the area distribution of the granular product region;
Inspection equipment for granular products. The number counting means is
Extracting a plurality of mountain shapes in the area distribution of the granular product region, performing processing for making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of granular products,
estimating the number of the non-defective granular products present in each of the granular product regions as the number corresponding to the mountain shape to which the area occupied by the non-defective granular products belongs;
estimating the number of the defective granular products present in the granular product region as the number corresponding to the mountain shape to which the area occupied by the defective granular products belongs;
The inspection device for granular products according to claim 5. 3. The number counting means performs a process of calculating the number of the non-defective and defective granular products existing in the captured image based on the estimated numbers of the non-defective and defective granular products. 5 or 6 granular product inspection device. A method of capturing an image of a granular product whose management surface looks like a square and counting the number of the granular product from the captured image,
arranging a plurality of the granular products so that the management surface faces the imaging direction without overlapping;
irradiating the plurality of granular products with light;
imaging the plurality of granular products irradiated with light;
Extracting a granular product region in which one granular product exists or a plurality of granular products exist adjacently in the captured image,
calculating the area of the granular product region;
estimating the number of the granular products present in each granular product region based on the area of each granular product region and the area distribution of the granular product region obtained in advance from past captured images;
A method for counting the number of granular products. Extracting a plurality of mountain shapes in the area distribution of the granular product region, making each of the extracted mountain shapes correspond to the number of the granular products,
in the process of estimating the number of the granular products, estimating the number of the granular products present in the granular product region as the number corresponding to the mountain shape to which the area of the granular product region belongs;
The method for counting the number of granular products according to claim 8. 10. The method for counting the number of granular products according to claim 8 or 9, wherein the number of the granular products present in the captured image is calculated based on the estimated number of granular products. The granular product is a resin,
irradiating the plurality of granular products with light containing an ultraviolet wavelength band;
using a light receiving element group including light receiving elements having a light receiving region of a wavelength band of 400 to 620 nm, and capturing an image including the granular product fluorescent by the irradiated light;
The method for counting the number of granular products according to any one of claims 8 to 10. A method for capturing an image of a granular product whose management surface looks like a square and inspecting the granular product from the captured image, comprising:
Using the method for counting the number of granular products according to any one of claims 8 to 11,
calculating the area occupied by the good granular products and the area occupied by the defective granular products in the granular product region;
In the process of estimating the number of granular products,
estimating the number of the good granular products present in each granular product region based on the area occupied by the good granular products in each granular product region and the area distribution of the granular product regions;
estimating the number of defective granular products present in each granular product region based on the area occupied by the defective granular products in each granular product region and the area distribution of the granular product region; Inspection method for granular products. Extracting a plurality of mountain shapes in the area distribution of the granular product region, making each of the extracted mountain shapes correspond to the number of the granular products,
In the process of estimating the number of granular products,
estimating the number of the non-defective granular products present in each of the granular product regions as the number corresponding to the mountain shape to which the area occupied by the non-defective granular products belongs;
estimating the number of the defective granular products present in the granular product region as the number corresponding to the mountain shape to which the area occupied by the defective granular products belongs;
The method for inspecting granular products according to claim 12. 14. The granular product inspection method according to claim 12 or 13, wherein the number of the non-defective and defective granular products present in the captured image is calculated based on the estimated numbers of the non-defective and defective granular products. .

Images