WO1999022331A1

WO1999022331A1 - Grade stamp reader for packed products and grade stamp judging method

Info

Publication number: WO1999022331A1
Application number: PCT/JP1997/004675
Authority: WO
Inventors: Toshiharu Shinto; Tadashi Motohashi; Shinichi Shiroma
Original assignee: Japan Tobacco Inc.
Priority date: 1997-10-24
Filing date: 1997-12-18
Publication date: 1999-05-06
Also published as: TW411429B

Abstract

A grade indication part (4) is printed on a side 10a of a corrugated box (10) in which agricultural products are packed. The color of the frame (41) and grade table (42) of the grade indication part (4) is red. The color of image reference marks (43) is black. The color of grade stamps (5 and 5') is also black. Illuminating light whose color is red which is the color of the grade table (42) is applied to the side (10a) of the corrugated box (10) by an illuminating device and the image of the side (10a) is picked up by a CCD camera. The image of the image reference marks (43) and the image of the grade stamps (5 and 5') are extracted from the image data. The positions of the grade stamps (5 and 5') are judged based on the positions of the image reference marks (43). Thus, the grade stamps can be read with a high precision automatically from the grade table of the corrugated box of agricultural products.

Description

Description Grade stamp reader for packaged products and method for class stamp determination

According to the present invention, in order to classify packaged products such as boxed products of agricultural products by the class of the packaged product, the packaged product which reads the class grade stamp marked on the packaged product and determines the presence or absence of the stamp is determined. The present invention relates to a first-class stamp reading device and a first-class stamp determining method. Background art

Traditionally, some boxed agricultural products have a cardboard box stamped to indicate the grade of the boxed agricultural product. For example, Figure

As shown in Fig. 24, on the side of the cardboard box 100, a grade table 200 with letters in the boxes is printed, and after packing the boxes, the letters (squares) corresponding to the grade of the agricultural product are displayed. One or several stamps 300 are imprinted on the part. Then, when such boxed products are shipped, etc., they are classified by judging the equivalent class by the stamp position of this stamp.

Sorting as described above is performed semi-automatically by a sorting device for cardboard boxes conveyed on a conveyor. In other words, the inspector visually reads the stamp marking position in the grade table of the cardboard box being conveyed on the conveyor within a predetermined time, and designates the read result to the sorting device as the judgment result. Input device manually.

As described above, conventionally, the inspector visually determines the stamp stamping position in the grade table, which not only places a burden on the work but also causes problems such as reading errors and input errors. . The present invention improves the grade table and automatically positions the stamp of the grade stamp in the grade table in order to reduce the burden of sorting products using the grade stamp. An object of the present invention is to accurately read and accurately determine the presence / absence of a class stamp in a cell of a class table. Disclosure of the invention

The present invention has been made to solve the above-mentioned problem. The apparatus according to the first aspect of the present invention is a packaged product class stamp reading device, comprising: a class table marked with a color that reflects visible light in a specific wavelength range; A packing product having a class indicator consisting of an identification mark marked in a color different from the table and having a class stamp of a different color from the class table imprinted on the class table. And an illumination means for irradiating the surface of the packaged product on which the equivalent class display section is marked with illumination light of the same color as the equivalent class table. Imaging means for capturing an image of the surface on which the class display section on which the illumination light has been irradiated by the illumination means is marked; and determining the position of the identification mark based on image data captured by the imaging means. Based on the position of the identification mark Determining means for extracting an image of the above-described pump and reading a position of the extracted image of the pump with respect to the equal rank table.

According to the apparatus for reading a graded stamp of a packaged product according to claim 1 configured as described above, the illuminating means irradiates the graded table with illumination light of the same color as the graded table. The identification mark and the rank stamp having a different color from the rank table are emphasized as a darker color than the rank table, so that the identification mark and the rank stamp can be easily extracted from the image data captured by the imaging means. In addition, as described in claim 2, the apparatus for reading the class-grade stamp of the packaged product is provided in a dark room covering a part of the transport path for transporting the packaged product. In the dark room, the surface of the printed product with the grade display on the transport path is imaged, and the position of the stamp image with respect to the grade table is read. Automatic sorting of products can be performed.

Further, as described in claim 3, when the lighting means irradiates the illumination light almost parallel to the surface of the packaged product on which the equivalent class display portion is marked, the illumination on the surface of the packaged product can be improved. The influence on the image due to the direct reflected light can be reduced. Further, as described in claim 4, when the illuminating means is constituted by the LED, it is possible to stably irradiate illumination light of the same color as that of the equal rank table.

Further, the lighting device for imaging a packaged product according to claim 5 of the present invention is a packaged product imaging device for irradiating illumination light on the flat surface of the packaged product in order to image a chart marked on the flat surface. An illumination device according to claim 1, further comprising a pair of light diffusing surfaces disposed substantially at right angles to the flat surface portion of the packaged product and arranged close to the flat surface portion, wherein the light diffusing surface is A pair of light diffusing members disposed so as to face each other at an interval sandwiching the diagram, and a light emitting surface for emitting the illumination light between the pair of light diffusing surfaces of the light diffusing member in a direction parallel to the plane portion. And an illumination light source disposed opposite to the illumination device. The illumination light source irradiates the illumination light from the illumination light source nearly parallel to the plane portion, and diffuses the illumination light on the light diffusion surface. It is characterized by the following.

According to the illumination device for imaging a packaged product according to claim 5 configured as described above, the illumination light from the illumination light source is irradiated almost parallel to the flat portion of the packaged product. The effect of direct reflected light on the image on the image is reduced. Further, since the light diffusing surface of the light diffusing member is arranged at a right angle to the flat portion and close to the flat portion, uniform illumination light is applied to the flat portion.

As described in claim 6, when the lighting means is configured by an LED, -

Four

Illumination light of the same color as the above-mentioned rank table can be stably radiated.

In addition, the method for judging the rank of a packaged product according to claim 7 of the present invention has a grade display section comprising a grade table composed of a plurality of squares and an identification mark, A method for judging the presence or absence of a class stamp in the cell of the class table, wherein the class product has a class stamp imprinted therein. An image is captured, the captured entire image is divided into a plurality of partial images, and pixels are divided for each partial image based on an average luminance value of the entire image and an average luminance value of each of the partial images. A threshold value for binarizing the luminance is set. For each partial image, the luminance of the image is binarized based on the threshold value corresponding to the partial image, and the image data of the two-tone image is displayed. Converted to the evening, the image of the two-tone image Detecting the position of the identification mark, setting an equal-grade frame in the square of the equal-grade table based on the position of the identification mark, and determining the same-grade frame to be determined based on the image data of the two-tone image. The degree of concentration of the stamp on the surrounding class frame is determined, and a threshold value for the area ratio of the same luminance portion in the class frame for the judgment is set according to the concentration level. And determining whether there is a jump in the class frame to be determined based on the set threshold value and the area ratio in the class frame to be determined. .

According to the method for judging the rank of a packaged product according to claim 7 configured as described above, the image including the grade display is divided into partial images, and a threshold is set for each partial image. Therefore, it is possible to obtain a two-tone image in which the influence of uneven illumination during imaging is reduced.

In addition, a threshold value is set for the area ratio of the same luminance portion in the class frame to be determined in accordance with the degree of concentration of the stamp in consideration of the class frame around the class frame to be determined. The threshold value of the stamp in the Since the presence / absence is determined, it is possible to reduce the influence of the displacement of the stamp on the surrounding class frames.

In addition, the method of forming an equivalent class display portion according to claim 8 of the present invention is characterized in that an equivalent class table marked with a color reflecting light in a specific wavelength range and an identification mark marked with a color different from the grade table. Are formed in a packaged product.

According to the grade display unit formed by the method for forming a grade display unit according to claim 8 configured as described above, a stamp of a different color from the grade table is used, and light in a specific wavelength range is used. By simply illuminating with, the image of the identification mark and the sky can be easily and accurately extracted. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram of an apparatus for reading a class-grade lamp of a packaged product according to an embodiment of the present invention.

FIG. 2 is a front perspective view of an imaging unit in the reader.

FIG. 3 is a plan view of the imaging unit.

FIG. 4 is a diagram illustrating an example of a class display unit according to the embodiment.

FIG. 5 is a flowchart of a main flow of the determination logic in the embodiment.

FIG. 6 is a flowchart of the binarization processing of the determination logic in the embodiment.

FIG. 7 is a flowchart of the image reference mark detection process of the determination logic in the embodiment.

FIG. 8 is a flowchart of the stamp detection process of the determination logic in the embodiment.

FIG. 9 is a functional block diagram of the judgment port magic in the embodiment.

FIG. 10 illustrates the entire image average luminance calculation processing of the determination logic in the embodiment. FIG.

FIG. 11 is a diagram illustrating a partial image average luminance value calculation process of the determination logic in the embodiment.

FIG. 12 is a diagram showing a partial image binarization threshold value calculation process of the determination logic in the embodiment.

FIG. 13 is a diagram illustrating an example of a two-tone image of the determination logic in the embodiment.

FIG. 14 is a diagram showing a schedule setting process of the image reference mark detection area of the judgment logic in the embodiment.

FIG. 15 is a diagram showing a labeling process of the decision logic in the embodiment.

FIG. 16 is a diagram illustrating an image reference mark detection process of the determination logic in the embodiment.

FIG. 17 is a diagram illustrating an image reference mark position calculation process of the determination logic in the embodiment.

FIG. 18 is a diagram showing a stamp area detection frame calculation process of the determination logic in the embodiment.

FIG. 19 is a diagram showing a modification of the image of the rank table, a stamp area detection frame, and a stamp area emphasis frame in the embodiment.

FIG. 20 is a diagram showing expansion / compression processing of the determination logic in the embodiment.

FIG. 21 is a diagram illustrating an example of an equal rank frame to be determined by the determination logic in the embodiment and an adjacent rank frame.

FIG. 22 is a diagram for explaining the degree of concentration of stamps in the embodiment.

FIG. 23 is a diagram showing the light emission characteristics of the light emitting diode in the example. FIG. 24 is a diagram showing an example of a conventional rank table. 1

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a device for reading an equal-grade stamp of a packaged product according to an embodiment of the present invention, FIG. 2 is a front perspective view of an imaging unit in the reading device, and FIG. 3 is a plan view of the imaging unit. The reader comprises an imaging system A and a data processing system B.

The imaging system A includes a light-shielding dark room 1 that covers a part of the conveyor 20 that transports a cardboard box 10 to be read, a lighting device 2 that is disposed in the light-shielding dark room 1 and that is disposed near the conveyor 20. , And a CCD camera 3 as imaging means disposed on the opposite side of the luminaire 2 from the comparator 20. The image processing system B is also capable of receiving image data from the CCD camera 3 and outputting monochrome digital image data of 256 gradations. The personal computer 12 is configured to determine the stamping position of the equal-grade stamp in the equal-grade table of the cardboard box 10 by taking into account the pre-determined logic. The result of the determination by the personal computer 12 is output to the next step (not shown) (such as a sorter).

The illuminating device 2 includes two opposing light diffusion plates 2 2-1, 2 2-as light diffusion members disposed along the conveyor 20 between the base 21 and the top plate 24. 2. Illumination light sources 2 3 arranged between the light diffusing plates 22-1 and 22-2 at both ends of the conveyor 20 of the light diffusing plates 22-1 and 22-2 in the conveying direction. — It has 1, 2 3— 2. The distance between the two light diffusing plates 2 2-1 and 2 2-2 should be such that it covers at least the area that includes at least the rectangular equivalent class display section 4 printed on the side of the cardboard box 10. The opposing surfaces of the light diffusing plates 2 2-1 and 2 2-2 have the property of reflecting and diffusing light. ing. Illumination light sources 23-1, 23-2 have a large number of red light-emitting diodes inside, and each light-emitting surface 23-la, 23-2a has transmission and diffusion properties for light. ing. Also, the light sources 23-1 and 23-2 face the light emitting surfaces 23-1 a and 23-2 a to the space between the light diffusers 22-1 and 22-2. They are arranged facing each other.

The CCD camera 3 has the imaging lens 31 facing the gap between the light diffusion plates 22-1, 22-2, and the optical axis of the imaging lens 31 is adjusted to the light diffusion plates 22-1, 22-1, 22. And at a right angle to the conveying direction of the conveyor 20. Thus, the CCD camera 3 captures an image of the side surface of the cardboard box 10 on the conveyor 20 through the gap between the light diffusion plates 22-1, 22-2.

With the above configuration, the illumination light sources 23-1 and 23-2 diffuse and emit red light on the light-emitting surfaces 23-1a and 23-2a by the radiation of the internal light-emitting diode. The emitted red light is further reflected and diffused by the opposing surfaces of the light diffusing plates 22-1 and 22-2. As a result, the side surface 10a of the cardboard box 10 conveyed by the conveyor 20 is uniformly illuminated with red light, and the side surface 10a of the cardboard box 10 illuminated with the red light is illuminated by the CCD camera 3. It is imaged. Then, the stamp position of the equivalent class stamp on the equivalent class display unit 4 is determined based on the captured image data. The light emitting diodes of the light sources 23-1 and 231-2 have light-emitting characteristics as shown in FIG.

By the way, in this embodiment, the cardboard box is white. Even if the cardboard box is made of paper, its surface may be glossy, and depending on the way of illuminating light, reflected light on the surface may cause extreme brightness unevenness in the captured image. This may affect the determination result.

1, the light source 23-1a's light emitting surface 23-1a and the light source 23-2's light emitting surface 23-2a have a positional relationship parallel to the conveying direction of the conveyor 20. Direction, that is, the grade class display part 4 of the cardboard box 10 faces in the direction parallel to the printed side face 10a, so that the direct light from the illumination light sources 2 3-1, 2 Light is incident at a low angle to 0a (the angle between the light beam and the side surface is a small angle). For this reason, even if the surface of the cardboard box 10 is glossy, the direct light from the illumination light sources 23-1, 2-3-2 is reflected by the side face 10a when viewed from the position of the CCD camera 3. This makes it possible to obtain image data without extreme brightness unevenness without being seen.

When taking an image with the CCD camera 3, for example, a cardboard box 10 is detected by a switch or the like placed near the transport path of the conveyor 20, and the class of the side of the cardboard box 10 is classified. The timing is set so that the display unit 4 is within a predetermined field of view of the CCD camera 3.

FIG. 4 is a view showing an example of the equal-grade display section 4 printed on the side face 10a of the cardboard box 10 in the embodiment, and FIG. 4 (A) is a multi-grade display of FIG. 4 (B). It is of a lower class. The equivalent class display section 4 is printed on the side face 10a of the cardboard box 10 and includes a rectangular outer frame 41, squares in the outer frame 41, and circled numbers attached to each square. It is composed of a class table 42, and image reference marks 43 as identification marks located on both sides in the longitudinal direction of the class table 42. Note that the image reference mark 43 has a vertically elongated rectangular shape on the outer periphery and a vertically elongated hollow portion inside.

In this embodiment, the outer frame 41 and the grade table 42 are printed in red so that the red light of the illumination light sources 23-1, 23-2 is almost reflected. The image reference mark 43 is black. Printed. Then, the rank stamps 5 and 5 ′ are stamped with a color (for example, black) having a lower luminance value than the red color, and ideally, as shown in the illustrated rank stamp 5, inside the square of the rank table 42. It is stamped. Also, the stamp may be stamped off the square as in the class stamp 5 ', and even in such a case, the stamp is stamped on any square, that is, on which square. Is determined by a determination logic described later.

Here, the cardboard box 10, the lighting device 2, and the CCD camera 3 are in the shading darkroom 1, and the cardboard box 10 is illuminated by the lighting device 2 with only red light. For this reason, when the side surface 10a of the cardboard box 10 is imaged by the CCD camera 3, the outer frame 4 1 of the equivalent class display section 4 and the class table 4 The portion 2 has almost the same luminance value as the side surface 10a, and the black image reference mark 43 and the equal-grade stamps 5 and 5 'have remarkable luminance values different from the surroundings. Therefore, by binarizing the luminance value of this image data with a predetermined threshold value, it is possible to extract the shapes of the image reference mark 43 and the equivalent class stamps 5, 5 'and the positions over the image data. it can. In addition, since the rank table 42 and the image reference mark 43 are in a predetermined positional relationship, the position of the image reference mark 43 (the position on the image data) indicates the rank of the rank table 42. The position of the squares can be uniquely determined. The personal computer 12 is a CPU that executes programs, a ROM that stores basic programs (BIOS, etc.) for input / output control, etc., a video memory, etc. RAM, a hard disk that is a large-capacity memory, a floppy disk drive, and the like. A processing program that constitutes the determination logic is input from, for example, a floppy disk drive and is expanded on the RAM or the hard disk. Performed by

5 to 8 are main routines and subroutine flowcharts showing the flow of processing of the judgment logic in the personal computer 12, and FIG. 9 is a functional block diagram of the judgment logic. The judgment port will be described. The image data can be stored in a working area or the like of the RAM. Therefore, since the position and size of each part in the captured image can be obtained by known image calculation processing and calculation processing, detailed description of such image calculation processing and calculation processing is omitted.

FIG. 5 is a flow chart of the main flow showing a large flow of the determination logic. First, in step S1, the image data of the 256-tone image is captured, and in step S2, the binary values in FIG. Then, the image data of the 256-tone image is converted into black or white (0 or 1) 2-tone image data for each pixel. Next, in step S3, the image reference mark detection processing of FIG. 7 is performed to obtain the position and area of the image reference mark from the two-tone image, and in step S4, the skip detection processing of FIG. 8 is performed. Then, it is determined whether or not there is a class jump in each cell of the class table.

(Binarization processing)

In the binarization process shown in FIG. 6, in step S21, the average luminance value of the entire image of the 256-tone image (referred to as the "entire image average luminance value") is obtained. Calculate the average luminance value (referred to as “partial image average luminance value”) for each partial image obtained by dividing the whole into multiple regions. Next, in step S23, a threshold value for binarizing each partial image based on the average luminance value of the entire image and the average luminance value of the partial images (referred to as “partial image binarization threshold”) ) Is calculated, and in step S24, each pixel is binarized in units of each partial image based on the partial image binarization threshold value, and is converted into image data of a two-tone image. Then, the process returns to the main routine.

The processing of the above steps S21 to S24 will be described with reference to FIGS. FIG. 10 shows the whole image average luminance calculation processing (step S 21), and the luminance distribution of the whole image V of the 256-tone image shown in FIG. The distribution of the number of pixels with respect to the luminance value) is, for example, as shown in FIG. 10 (B). Therefore, for example, the sum of the luminance values of each pixel is divided by the total number of pixels. -1

12

Then, the average luminance value (G) of the entire image is obtained as shown by the broken line in FIG. 10 (B).

Fig. 11 shows the partial image average brightness value calculation processing (step S22). As shown in Fig. 11 (A), the whole image V is divided into 624 partial images of 13 rows x 48 columns. I do. Then, the average value of the partial images (Ln) is obtained by the same averaging process as the average luminance value of the entire image for all the partial images. For example, the luminance distributions of partial images a, b, and c are as shown in Figs. 11 (B) to (D), and the average luminance values of each partial image (La, Lb, Lc) are indicated by the broken lines. Ask for ο

FIG. 12 shows the partial image binarization threshold value calculation processing (step S23). For example, as shown in FIGS. 12 (A) to 12 (C), the average luminance value (G ) And the partial image average luminance value (Ln), the partial image binarization threshold (THDn) is obtained by the following equation (1).

THDn = K l x f (G, L n) (1)

Where n is the index of the partial image

K 1: Binarization threshold rate (Default is 0.75) f (X, y): Evaluation function (Default is (x + y) / 2) Note that Fig. 12 (A) to (C) For partial images a, b, and c, the average luminance value of the whole image (G), the average luminance value of the partial images (L a, Lb, L c), and the threshold values of the partial image binarization (THD a, THD b, THD c ).

Fig. 13 shows an example of a two-tone image obtained in the partial image binarization process (step S24). For all partial images n, the partial image binarization threshold (THDn) is used. Then, binarization processing is performed by the following equation (2) to generate a two-tone image. 1; I (x, y) ≥ THD n

0 (x, y) (2)

0; I (x ₃ ) <THDn where 1: luminance value of 256 gradation image

0: Luminance value of two-tone image

X, y: coordinates in partial image n

(Image reference mark detection processing)

In the image reference mark detection process of FIG. 7, in step S31, the inspection order (schedule) of the image reference mark detection area is set according to the type of the class display unit 4 printed on the cardboard box 10 described later. Then, in step S32, for each image reference mark detection area, a labeling process of assigning a different label to each block of pixels having a luminance value of "0" (black) in the image is performed. Next, in step S33, an image reference mark is detected based on a plurality of inspection items corresponding to the features of the image reference mark, and an image reference mark detection process for obtaining an area of the image reference mark is performed. Then, in step S34, an image reference mark position calculation process is performed to determine the positions (center of gravity, upper end position, lower end position) of the feature points of the image reference mark. The processes of the above steps S31 to S34 are shown in FIG. This will be described with reference to FIGS. FIG. 14 shows the schedule setting processing (step S31) of the image reference mark detection error. In this embodiment, the above-mentioned FIG. The class rank display section 4 in Fig. 4 (A) is for the case where there are many grade evaluation stages.The number of squares in the class scale table 42 is larger than that in Fig. 4 (B). The interval between the reference marks 43 is also wide. Therefore, as shown in FIG. 14, three image reference mark detection areas E 1, E 2, E 3 at the left end, the center, and the right end with respect to the entire image V Is set in advance. Then, the inspection order of the image reference mark detection area is set in the order of E1 → Έ3 → Ε2 for the rank display section 4 of FIG. 4 (A), and the rank order display section of FIG. 4 (() is set. Set E 1 E 2 Ε 3 in order for 4 ο

FIG. 15 shows the labeling process (step S32). This figure shows a case where dirt is present above and below the image reference mark 43. The dirt above and below the image reference mark 43 is the part where the pixel luminance value is "0" (black). A cluster of pixels having the same luminance value is formed. Therefore, different labels are defined, for example, label 1, label 2, and lapel 3 in the order of the upper dirt part, the image reference mark part, and the lower dirt part.

FIG. 16 shows a plurality of inspection items corresponding to the features of the image reference mark in the image reference mark detection processing (step S33). For each of the above labels, whether or not the block of the image is an image reference mark is determined. In order to determine whether or not each inspection item is calculated from the coordinates of the pixels of the block of the image of each label. Table 1 shows the inspection items and the inspection conditions for satisfying each item.

Image reference Image reference mark Image reference discount inspection conditions

Mark Inspection Inspection contents (te, fault)

Item

Lower limit Upper limit

1) Main shaft Long distance of main shaft 1 30 pixels

Long distance (image capture range

Height of the enclosure)

2) Main shaft Main shaft angle (tan 6> 75 degree ~ 1 15 degree long angle evaluation)

3) Main axis Longitudinal distance of main axis 0.5 .20.0

Long distance ratio Distance ratio

4) Main shaft Short distance 0.5.22.0 of main shaft, short distance of sub shaft and sub shaft

Short distance ratio Distance ratio to separation

5) Area The area between 0.5 and 2.0 and the hole area in the hole area

Specific area ratio

6) Number of holes

A label that satisfies all the inspection conditions of the above inspection items is defined as an image reference mark. At this time, the area of the block corresponding to the label defined as the image reference mark is determined as the image reference mark area (M m). Here, m is the image reference mark number, 1 indicates the left side, and 2 indicates the right side.

Figure 17 shows the positions of the feature points of the image reference mark in the image reference mark position calculation processing (step S34), that is, the center of gravity, the upper end position, and the lower end position. The positions (Pum, P gm, P dm) of these feature points are calculated from the chunk corresponding to the label defined as the image reference mark. However, Pu indicates the top of the mark, Pg indicates the center of gravity of the mark, Pd indicates the position of the bottom of the mark, and m indicates the image reference mark number.

(Stamp detection processing)

In the stamp detection process shown in Fig. 8, in step S41, an image of a correctly stamped class stamp is taken from the image reference mark area (Mm) and the actual image reference mark area (known). Calculate the area above (“stamp logic area”). Next, in step S42, the position on the image of the cell of the class table (hereinafter referred to as "class class frame") is calculated from the detected position (Pum, Pm, Pdm) of the image reference mark. . In the class display section 4, since the image reference mark 43 and the class table 42 have a fixed positional relationship, the position of the class table on the image can be calculated.

Next, in step S43, based on the position of the class table on the image, the inscribed frame of the class box of the class table is obtained as a stamp area detection frame, and the stamp area detection frame is obtained. Calculate the stamp area detection frame that sets a rectangular sweep area emphasizing frame of the same size as that of the stamp.

Next, in step S44, in order to remove noise of the image in the stamp area detection frame, the black portion of the image is enlarged by one pixel around and the white portion is reduced to make the image thinner. An expansion / compression process is performed in which the image and the process of making the black portion thinner by enlarging the white portion by one pixel around the image twice each.

Next, in step S45, the stamp area (S) for each stamp area detection frame is calculated from the following equation (3). Note that the stamp expansion / compression normalization coefficient does not return to its original size when a circular stamp is subjected to expansion / compression processing. The default value at this time is hi = 1.07. S a

s + S b x? (3) a

Where S a is the area within the detection area

S b: Area within the stamp area emphasis frame

: Stamp expansion / compression normalization coefficient

5: Sump area emphasis load factor

Next, in step S46, the stamp area ratio (R) is calculated by the following equation (4).

S

R (4)

S L

Where S is the sump area

S L: Sump logical area

Then, in step S47, it is determined whether or not there is a jump in each cell of the equal rank table based on the stamp presence / absence determination condition.

The processing of the above steps S43, 44 and step S47 will be described with reference to FIGS. Fig. 18 shows the calculation process of the detection area of the stamp area (step S43). The inscribed frame of the equal-grade frame is stamped from the position of the equal-grade table calculated in step S42. The area detection frame is used, and its position is calculated. In addition, a rectangular stamp area emphasizing frame of the same size as the stamp is set inside the stamp area detecting frame by calculation.

Note that the image of the class table is shown in Fig. 19, for example.Fig. 19 (A) shows the case where the surface of the cardboard box on which the class frame is printed is not inclined, and Fig. 19 (B) shows the cardboard box. This is an example in the case where the surface is inclined. Even if the class table is deformed as shown in Fig. 19 (B), the image reference mark is also deformed as such, so the position of the class table calculated in step S42 is transformed. It is required immediately. Therefore, the stamp area, which is the inscribed frame of the class The frame and the stamp area emphasis frame are also deformed in the same way, but even with such a deformation, the presence or absence of the stamp is detected by the ratio of the area, so there is no effect.

FIG. 20 shows the result of the expansion / compression processing (step S44). As shown in the figure, the white noise as shown in FIG. That is, white noise in the black part is removed by thickening the black part of the image around, and black noise in the white part is reduced by thinning the white part of the image. Removed.

FIG. 21 shows an example of a class frame to be determined in the stamp presence / absence determination process (step S47). The sum of the stamp area ratios (R) of the class Azusa (the shaded area) to be judged and the class class adjacent to the class Judgment to be judged (in this example, the bottom and left and right three) Then, it is determined how many stamps are concentrated based on the total value (“the total value of the adjacent stamp area ratios”), and the threshold value for determining whether or not the stamp is present is determined based on the degree of concentration.

In other words, as shown in Fig. 22, even when the stamp is not pushed in the judgment target class frame (shaded area), the stamp pushed in the surrounding class class frame is determined by this judgment target class. In some cases, there is a deviation within the frame, and in this case, the protruding portion is detected as the stamp area ratio of the equal-grade frame to be determined. Also, the area of such a portion that protrudes into the class frame to be determined depends on how many stamps are stamped on the adjacent class frame. For example, as shown in Fig. 22 (A), when no stamp is stamped on the surrounding class, the area of the protruding part is 0, but as shown in Figs. 22 (B) to (D). However, as the number of surrounding stamps becomes 1, 2, or 3, the area of the part protruding into the judgment target class frame tends to increase. In other words, as the number of surrounding stamps increases, the stamp area ratio of the equal class frame to be determined tends to increase. Therefore, the class class to be judged and its surroundings From the total value of the adjacent stamp area ratios by the surrounding class frames, it is first determined how many stamps (concentration degree) have been pressed in the judgment class frame and the surrounding class frames. Based on the degree, the judgment threshold (stamp presence / absence judgment threshold) for the stamp area ratio of the target class frame is determined. Then, the presence / absence of a stamp is determined based on the threshold value for determining the presence / absence of the pump depending on the case. The adjacent class frames are the upper, lower, left, and right class frames. If the class frame to be judged is at both ends, there are two, and the other classes are three. Also, since the total value of the adjacent stamp area ratio is the total value, it may be 100% or more.

Table 2 shows an example of the correspondence between the degree of concentration, the threshold value for judging the presence / absence of stamps, and the result of judging the presence / absence of stamps, corresponding to the total value of the area ratios of adjacent lamps.

Adjacent stanf. Area Stan: 7 ° collection Stanf ° presence / absence Stanf ° presence / absence ratio Total value Medium threshold value Judgment result Lower limit Upper limit Lower limit Upper limit

50% 0

Zhu ί¾

50% 134% 1 piece Less than 50%

No more 未

50% or more and less than 268% Yes

134% 225% 2 pieces Less than 67%

More

67% or more and less than 268% Yes

225% 770% 3 4 less than 75%

More vermilion ^ j

75% or more and less than 268% Yes

770% 5 or more undecidable or more Table 2

Through the above processing, it is possible to obtain image data with reduced effects such as uneven lighting, and to automatically and accurately read the stamp position of the equal-grade stamp in the equal-grade table, and further determine Since the influence of the displacement of the stamp around the target class frame can be reduced, the presence or absence of the class stamp in the square of the class table can be determined with high accuracy. Therefore, by using this result for a sorting device or the like, it is not necessary for the inspector or the like to input a visual / visual determination result, and the burden of the sorting operation can be reduced. As described above, the processing program shown in the flowchart of FIG. 5 to FIG. 8 and the function block diagram of FIG. 2 That is, the program is being executed overnight, and the above effects can be obtained. This processing program can be supplied to a computer via various media such as a floppy disk, a hard disk, an MD, a CD-ROM, a magnetic tape, a ROM, and a RAM. The medium is equivalent to the medium of claim 2 of the present invention.o Industrial applicability

As described above, according to the packing product equivalent class stamp reading device described in claim 1 of the present invention, the stamp position of the class sorting stamp in the class table can be automatically and accurately read. This can reduce the burden of sorting products using a class-grade lamp.

According to the apparatus for reading a stamp of equal rank of a packaged product according to the second aspect of the present invention, it is possible to automatically sort a packaged product while reading an image of a stamp with high accuracy.

Further, according to the apparatus for reading a rank stamp of a packaged product according to claim 3 of the present invention, it is possible to reduce the influence on the image due to the direct reflection light on the surface of the packaged product, and the rank stamp in the rank table. It can automatically and accurately read the stamping position of the product, and can reduce the burden of sorting products using a graded stamp.

Further, according to the apparatus for reading a rank of a packaging product according to claim 4 of the present invention, it is possible to stably irradiate illumination light of the same color as the rank of the rank, and the rank in the rank of the rank is displayed. The stamping position of the stamp can be automatically read with high accuracy, and the burden of classifying packing products using graded stamps can be reduced.

Further, according to the lighting device for imaging a packaged product according to claim 5 of the present invention. For example, the influence of the direct reflection light on the flat part of the packaged product on the image is reduced, and the flat part can be uniformly illuminated by the light diffusing surface of the light diffusing member.

Further, according to the illumination device for imaging a packaged product described in claim 6 of the present invention, illumination light of the same color as that of the grade table can be stably emitted.

According to the method for judging the rank of a packaged product according to claim 7 of the present invention, it is possible to obtain a two-tone image in which the influence of uneven illumination at the time of imaging is reduced and obtain a surrounding graded frame. Can reduce the effects of misalignment of stamps, etc., and can accurately determine the presence or absence of equal-grade stamps in the squares of the equivalent rank table, and sort packaging products using the same-grade stamps The burden on the user can be reduced.

According to the grade display unit formed by the method for forming a grade display according to claim 8 of the present invention, a stamp having a different color from the grade table is used, and light of a specific wavelength range is used. By simply illuminating, the image of the identification mark and the stamp can be easily and accurately extracted, and the presence / absence of an equal-grade stamp in a cell of the equivalent-grade table can be accurately determined. It is possible to reduce the burden of the sorting operation of the packaged products using the wrapper.

Claims

The scope of the claims

1. It has a grade table that is composed of a grade table that is marked with a color that reflects light in a specific wavelength range and an identification mark that is marked with a color different from the grade table. An apparatus for reading a packaging product on which a packaging grade stamped with a classification stamp having a color different from that of the packaging grade table, wherein the packaging product reads the location of the packaging stamp of the packaging grade stamp,

Illuminating means for irradiating illumination light of the same color as the grade table to a surface of the packaged product on which the grade display section is marked;

Imaging means for imaging the surface on which the equal-grade display section irradiated with illumination light by the illumination means is marked;

Determining the position of the identification mark based on the image data imaged by the imaging means; extracting the image of the scan based on the position of the identification mark; Determining means for reading a position with respect to the rank table;

An apparatus for reading a rank stamp of a packaged product, comprising:

2. The package is arranged in a dark room that covers a part of a transport path for transporting the packaged product, and captures an image of a surface of the packaged product transported on the transport path on which the equivalent class display section is printed, and performs the scan. Reading the position stamp of the packaged product according to claim 1, wherein the position of the image is read with respect to the position table.

3. The packaged product according to claim 1 or 2, wherein the illuminating means irradiates the illumination light nearly parallel to a surface of the packaged product on which the equivalent class display portion is marked. Grade stamp reader.

4. The apparatus according to claim 1, 2, 3 or 4, wherein the lighting means is an LED.

5. An illumination device for imaging a packaged product, which irradiates illumination light onto the flat surface to image a chart marked on the flat surface of the packaged product,

The packaged product has a pair of light diffusing surfaces arranged at a right angle to the flat surface portion and close to the flat surface portion, and the light diffusing surfaces are arranged at an interval sandwiching the diagram of the flat surface portion. A pair of light diffusing members disposed to face each other,

An illumination light source disposed between the pair of light diffusion surfaces of the light diffusion member and a light emitting surface that emits the illumination light in a direction parallel to the plane portion;

With

An illumination device for imaging a packaged product, characterized in that illumination light from the illumination light source is radiated almost parallel to the plane portion, and the illumination light is diffused on the light diffusion surface.

6. The illumination device for imaging a packaged product according to claim 5, wherein the illumination means is an LED.

7. A packaged product that has a class rank display section consisting of a class scale table composed of a plurality of squares and an identification mark, and that has a class rank stamp stamped on the class rank table. A packaging method for determining the presence or absence of a rank stamp in the square of the rank table,

Take an image of the class display,

Dividing the captured whole image into a plurality of partial images,

Setting a threshold value for binarizing the luminance of the pixel for each partial image based on the average luminance value of the entire image and the average luminance value of each of the partial images;

For each of the partial images, the brightness of the image is binarized based on the threshold value corresponding to the partial image and converted into image data of a two-tone image,

The position of the identification mark in the image of the two-tone image is detected, and an equal-grade frame is set in a cell of the equal-grade table based on the position of the identification mark. ,

Based on the image data of the two-tone image, the degree of concentration of the jump on the class frame to be determined and the surrounding class frame is determined,

A threshold value for the area ratio of the same luminance portion in the equal-grade frame is set in accordance with the above-mentioned concentration level for the equal-grade frame to be determined.

A packaged product characterized in that the presence or absence of a jump in the class frame to be determined is determined based on the set threshold value and the area ratio in the class frame to be determined. A method for determining the rank of a class.

8. A grade table marked with a color that reflects light in a specific wavelength range and an identification mark marked with a color different from the grade table are formed on a packaged product. Forming method.