JP2009187482A - Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium - Google Patents

Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium Download PDF

Info

Publication number
JP2009187482A
JP2009187482A JP2008029539A JP2008029539A JP2009187482A JP 2009187482 A JP2009187482 A JP 2009187482A JP 2008029539 A JP2008029539 A JP 2008029539A JP 2008029539 A JP2008029539 A JP 2008029539A JP 2009187482 A JP2009187482 A JP 2009187482A
Authority
JP
Japan
Prior art keywords
shelf allocation
product
shelf
image
program
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP2008029539A
Other languages
Japanese (ja)
Inventor
Sachiko Hirahara
Nobuo Hotaka
Sei Matsumoto
Ryuichi Miyajima
龍一 宮島
祥子 平原
聖 松本
伸夫 穂鷹
Original Assignee
Nippon Sogo System Kk
日本総合システム株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nippon Sogo System Kk, 日本総合システム株式会社 filed Critical Nippon Sogo System Kk
Priority to JP2008029539A priority Critical patent/JP2009187482A/en
Publication of JP2009187482A publication Critical patent/JP2009187482A/en
Application status is Pending legal-status Critical

Links

Abstract

<P>PROBLEM TO BE SOLVED: To provide a shelf allocation reproducing method for creating a shelf allocation model representing an article display state with high accuracy in a short period of time, and to provide a shelf allocation reproduction program, a shelf allocation evaluating method, a shelf allocation evaluation program, and a recording medium. <P>SOLUTION: A sales floor is imaged by a camera (step S1), and a computer receives the image data of the sales floor and stores it in an image storing part, or the like (step S2). The computer is used for color correction, size correction, distortion correction, or the like of the image of the sales floor (step S3). Matching is performed between each article in the image with an article master stored in the article master storing part of the computer, to identify each article (step S4). An article master selected by the matching is arranged at a position corresponding to a furniture model (step S5) to complete the shelf allocation model (step 6). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a shelf allocation reproduction method for creating a shelf allocation model representing the display state of a product, a program for implementing the shelf allocation model, a method for evaluating the product display state using the created shelf allocation model, and the method The present invention relates to a program and a recording medium on which these programs are recorded. More specifically, the present invention relates to a technique for creating a shelf allocation model of a current display state from photographs or image data obtained by photographing a sales floor.

  Since the assortment of goods and the display position of goods in a merchandise store greatly influence the sales of each store, the merchandise dealer performs a shelf replacement operation to periodically change the merchandise display state of the shelves. At that time, an operation for determining the product display state after the replacement of the shelf, that is, the shelf allocation is performed in advance. ) From the product information, etc., the types of products to be displayed and the display positions thereof must be examined, and the best selling products must be determined from a large number of product groups, which is troublesome. For this reason, generally, when performing a shelf allocation work, a system and a program for efficiently performing the work are used.

  However, the conventional shelf allocation support system and program have a problem that the burden on each store is heavy because information such as the type, number, and position of currently displayed products must be input. Therefore, a method has been developed and put into practical use by reading a barcode attached to a displayed product with a portable barcode reader for POS (Point of Sales) and taking the data into a computer. (For example, refer to Patent Document 1).

  In addition, a method of performing shelf allocation using an image obtained by photographing the product display state has been proposed (see Patent Documents 2 and 3). For example, in the shelf allocation table creation apparatus described in Patent Document 2, a product image is cut out from an image of a product display shelf photographed by a digital camera, and pasted into a shelf allocation table format prepared in advance, so that the shelf allocation operation is performed. To improve efficiency.

  Further, in the shelf allocation support system described in Patent Document 3, a digital image of the current product display state is obtained by taking a picture with a digital camera or taking a picture with a scanner in each store. Then, the image is partitioned for each product, and product data such as a product number is input to each partition and transferred to the server. The server performs diagnosis based on the transferred product data and position data, and sends a composite image obtained by combining the images of the products to be replaced with the transmitted image data to each store. Since the system described in Patent Document 3 inputs information about each product while looking at an image obtained by photographing the product display state, input errors can be reduced.

  On the other hand, in inventory work to check the number of products displayed on the shelf, the product display state is photographed with a camera, and the product is stored on the shelf by comparing the image data with pre-stored master data. In addition, a system for taking an inventory in a state where it is displayed is also proposed (see Patent Documents 4 and 5). For example, in the inventory system described in Patent Document 4, first, a target product is photographed with a camera, and image processing such as edge extraction processing, color boundary processing, pattern boundary processing, and character recognition processing is added to the digital image. Then, contour feature image data such as product shape, overlap, stacked shape and number of stacked stages is extracted. Then, the extracted data is compared with the outline feature data stored in the file, and the number of inventory of the target product is calculated.

Japanese Patent Laid-Open No. 4-3699795 Japanese Patent Laid-Open No. 9-160993 JP 2002-163436 A

  However, the conventional techniques described above have the following problems. First, in the case of a method of sequentially reading bar codes and the like of displayed products as in the display state display device described in Patent Document 1, the product information input operation can be made efficient. In a store with a large number of products such as the above, there is a problem that it takes a long time to read. Moreover, when reading the goods stored in the refrigerator, there is a problem that the door opening time becomes long and the internal temperature rises.

  On the other hand, the methods described in Patent Documents 2 and 3 do not require visual confirmation of displayed products and barcode reading operations, but use an image obtained by photographing the product display state as it is. On the terminal, image data must be classified for each product, and individual product information must be input, which increases the burden on the operator.

  Therefore, the present invention provides a shelf allocation reproduction method, a shelf allocation reproduction program, a shelf allocation evaluation method, a shelf allocation evaluation program, and a recording medium capable of accurately creating a shelf allocation model representing the display state of a product in a short time. The main purpose is to provide.

The shelf allocation reproduction method according to the present invention is a method of creating a shelf allocation model that represents the display state of products using a computer, and identifies each product displayed on the sales floor from a digital image of the sales floor photo. And a step of arranging a product master provided with product information at a corresponding position of a fixture model created in advance based on the identification result.
In the shelf division reproduction method, the identification step includes a step of classifying each product in the digital image of the sales floor photo for each product, a feature extracted from the image of each classified product, and the product master. A matching step of comparing and calculating the degree of coincidence.
Further, in the step of classifying each product, the image color of the digital image of the sales floor photo may be binarized and the boundary of each product may be extracted.
In that case, it is also possible to correct the category for the category set based on the extracted boundary and the product in the digital image of the sales floor photo that does not match.
Furthermore, in the matching step, a product master to be compared may be limited by the product information.
As the product information, for example, at least one of a handling situation at a sales floor, inventory information, product classification, manufacturer name, brand, release date, update date, and product name can be adopted.
Moreover, in the said matching process, the magnitude | size of each classified goods can be calculated, and the goods master of a comparison object can also be limited based on the value.
Furthermore, in the matching step, it is possible to compare only a partial area of the image of each product.
Furthermore, the product master having the highest degree of coincidence may be selected from the product master group on which the matching process has been performed, and placed on the fixture model.
Furthermore, color correction and / or distortion correction can be performed on the digital image of the sales floor photograph before the identification step.

The shelf allocation reproduction program according to the present invention is for causing a computer to execute the above-described shelf allocation reproduction method.
This program can also be used in combination with a program that supports shelf allocation.

  The shelf allocation evaluation method according to the present invention compares the shelf allocation model created by the above-described shelf allocation reproduction method with previously executed shelf allocation data, and calculates the degree of coincidence.

The shelf allocation evaluation program according to the present invention causes a computer to execute the shelf allocation evaluation method.
This program can be used in combination with a program that supports shelf allocation creation and / or the above-described shelf allocation reproduction program.

  The recording medium according to the present invention is a computer-readable recording medium in which the above-described shelf allocation reproduction program and / or shelf allocation evaluation program is recorded.

(Glossary)
“Shelves allocation” is a simulation of changing the shelf, and examines the type and position of the product displayed on the shelf (displayed), the position of the shelf level and the hook level, etc., and the product display state after the shelf change Refers to determining. In addition, “shelf replacement” means moving the position of the product displayed on the shelf at the store, removing the currently displayed product, adding a new product, or if necessary. Accordingly, it means changing the merchandise display state of the shelf by performing various operations such as moving, removing or adding the shelf and / or hook steps.

  The “shelf allocation model” is a simulation model in which “product master” is arranged in the “furniture model” and reproduces the product display state of the sales floor. “Furniture model” is a device on which products such as shelves are displayed, and “Product master” is POS data, inventory data, product classification and manufacturer name, brand, release date, update date, product name, etc. It is a series of data in which various product information and product images are associated with each other.

  According to the present invention, each product is identified from the digital image of the sales floor photo, and the product master of the corresponding product is arranged at the corresponding position of the fixture model created in advance. It can be created with high accuracy.

  The best mode for carrying out the present invention will be described below with reference to the accompanying drawings. Note that the present invention is not limited to the embodiments described below.

  First, the shelf allocation reproduction method according to the first embodiment of the present invention will be described. The shelf allocation reproduction method of the present embodiment is a method of creating a shelf allocation model that represents the display state of products from a photo of a sales floor taken with a digital camera or a film camera using a computer. The computer used in the shelf layout reproduction method of the present embodiment stores at least an image storage unit that stores image data of sales floor photos, and images and product information (hereinafter collectively referred to as a product master) of each product. A product master storage unit, and a display unit for displaying photographic images, shelf allocation models, and the like.

  The product image in the product master is not particularly limited. For example, an image cut out from a photograph of the product, a product image provided by the manufacturer, an image captured from a catalog with a scanner, and a computer. Various images can be used. Moreover, since the direction of each product displayed on the shelf may not be constant, it is desirable that the product master is provided with images viewed from a plurality of directions. Thereby, the precision in the matching process mentioned later can be improved.

  Further, the product information in the product master includes, for example, handling status at each store or sales floor such as POS data, inventory data, product classification, manufacturer name, brand, release date, update date, product name, and the like. Furthermore, the images and product information of each product are associated with each other and stored as a series of data (product master) for each product.

  FIG. 1 is a flowchart showing the procedure of the shelf allocation reproduction method of this embodiment. As shown in FIG. 1, in the shelf layout reproduction method of this embodiment, first, a photo of a sales floor is taken with a camera (step S1). At that time, it is desirable to take a photo for each shelf or each span, but if all the products displayed in one shelf or one span cannot fit in one photo, divide it into multiple photos. May be. In addition, it is desirable to take a photograph from the front of the shelf, but in the shelf division reproduction method of the present embodiment, even a photograph taken from an oblique direction can be used. Furthermore, it is also possible to cut out and use a video image of a target sales area as a still image from a moving image taken by a surveillance camera or the like.

  Next, the image data of the sales floor photograph taken in step S1 is taken into a computer and stored in an image storage unit or the like (step S2). At this time, if the sales floor photo is taken with a digital camera, the image data can be directly captured and stored. However, when the photo is taken with a film camera, the printed photo is read with a scanner or the like and digitized. Remember the data.

  Next, in order to align the size and color tone of the image data (hereinafter also referred to as a photographic image) of the captured sales floor photo with the size and color tone of the product master image, the photographic image is corrected using a computer (step S3). ). Specifically, the photo image was adjusted so that the size of the photo image shelf was the same as the size of the prefabricated furniture model shelf, and the sales floor photo was taken from an oblique direction rather than from the front. If it is, the distortion and aspect ratio of the photographic image are also adjusted. Further, the color tone such as hue, saturation and luminance of the photographic image is adjusted in accordance with the image of the product master. Further, the photographic image is rotated as necessary. In addition, when one shelf or one span is divided and photographed, it is desirable to combine the photographic images into one image and then perform the above-described processes.

  Thereafter, a matching process is performed to compare each product in the photographic image with each product master stored in the product master storage unit of the computer, and individual products are identified (step S4). FIG. 2 is a flowchart showing the procedure of the product matching process. As shown in FIG. 2, when performing product matching processing, first, a plurality of products shown in a photographic image are classified for each product (step S41).

  In that case, it is desirable to classify according to the width and height of each product. Thereby, since the background part which is unrelated to goods can be decreased, the precision of matching can be improved. In addition, although the shape of the area | region (henceforth a division area | region) divided for every product can be made into a rectangular shape, for example, it is not limited to this, Product shape and the product image in a product master Depending on the shape and the like, it can be appropriately selected.

  The step S41 may be performed manually while the operator looks at the photographic image. For example, the image color is binarized by the discriminant analysis method, the boundary line is extracted, the position of the shelf and each product, It can also be automated by detecting the contour of each product. At this time, the binarization of the image color is, for example, for each pixel of the photographic image, a lower limit density, an upper limit density, a forward histogram integration frequency, a forward appearance probability integration frequency, a reverse histogram integration frequency, and a reverse appearance probability. It is only necessary to calculate the integration frequency and set the threshold based on the result. In addition, the boundary line is extracted, for example, by extracting the frequency of white pixels on the vertical or horizontal line from the binarized image, setting a threshold based on the standard deviation calculated therefrom, and then binarizing again. The retrieved image may be searched and both ends of the portion that is equal to or greater than the threshold may be used as boundaries.

  In addition, when the product is automatically classified by the above-described method, the range of the segmented area is set so that the segmented area and the product are aligned with respect to the product in which the segmented area and the product in the photographic image are shifted. It is desirable to correct manually or automatically. Thereby, matching accuracy can be improved. Moreover, although the process of classifying for each product as described above can be performed for all the products at a time, it can also be performed for each shelf or for each designated area.

  Next, a region for matching (hereinafter referred to as a matching region) is set for each product classified for each product (step S42). At this time, all of the divided areas set in step S41 can be used as matching areas, but only a specific area of the divided areas can be used as a matching area.

  FIG. 3 is a diagram showing an example in which a part of the divided area is set as the matching area. For example, as shown in FIG. 3, some of the products 3 a to 3 e of the photographic image 1 are hidden by the price tag 7, product advertisements (POP advertisements) 5, 6, 8, and slip stoppers (not shown). In such a case, the regions without these may be set as the matching regions 9 and 11. Thereby, since elements other than goods 3a-3e can be excluded, the precision of matching processing can be improved. Similarly, when the background is included in the product image 10 and the photographic image segment areas 3a to 3e of the product master, and the background color is different between the product image 10 and the photographic images 3a to 3e, other than the background. By setting the portions in the matching regions 9 and 11, it is possible to eliminate the influence of the background portion unrelated to the products 3a to 3e and improve the matching accuracy.

  Next, the product master to be compared is narrowed down by the size of the product or the segment area and / or the product information (step S43). At that time, the size of the product or the divided area can be calculated based on the value of the shelf in the photographic image, for example, from the size of the furniture model. In addition, the range which narrows down a product master is not specifically limited, According to the number etc. of the product master used as a comparison object, it can set suitably.

  On the other hand, as narrowing down by product information, for example, a product master of a product that is not handled at a store or sales floor to be reproduced is excluded from comparison targets. By limiting the number of product masters to be compared with these methods, the number of objects to be subjected to a process described later is reduced, so that the processing time can be shortened and matching accuracy can be increased. In addition, this narrowing-down process may be performed by either size or product information, or both.

Next, the feature used for matching is extracted from the matching area of each product (step S44). At this time, the features used for matching include color features and shape features. The information used as the color feature includes RGB, XYZ, HSI, La * b *, and the like as the color system, but “hue”, “color” that is invariant to the gray value of the image, and the like. It is desirable to use an HSI color system composed of “Saturation” and “Luminance”.

  Next, matching with the product image of the product master is performed using the features extracted in step S44 (step S45). As the method, for example, a correlation coefficient method, SSDA method, least square matching, or the like can be applied. In addition, although these methods may be performed independently, it can also be performed combining several methods.

  In the shelf allocation reproduction method of the present embodiment, it is desirable to apply a correlation coefficient method that is less affected by the brightness of the image, among the above-described matching methods. The correlation coefficient method is a method of performing matching in units of pixels, and uses the value obtained by subtracting the average value of the pixel values of the entire image from the individual pixel values to calculate the correlation for all the pixels. Specifically, the consistency of the set matching region with the product image of the product master is determined based on the correlation coefficient R (M) obtained by the following formulas 1 to 3. In Equations 1 to 3 below, M is the product image of the product master, T is the matching area of the product to be evaluated, and M (x, y) is the pixel value at position (x, y). Further, the value of the correlation coefficient R (M) expressed by the following mathematical formula 1 decreases as the value approaches −1, and increases as the value approaches 1, the consistency increases.

  Based on the matching result of step S45, the corresponding product master is selected (step S46). Then, Steps S44 to S46 are repeated, the corresponding product master is selected for all products in the photographic image, and matching is completed (Step S47).

  Next, each merchandise master selected by the above-described matching process is arranged at a corresponding position on the furniture model (step S5). As a result, an image obtained by combining the product image of the product master with the fixture model is displayed on the display unit. In addition, matching results such as matching rate are also displayed on the display unit, and when there is something different from the product of the photographic image, the operator confirms the result of the matching process and It is also possible to perform correction work such as replacement. Note that the arrangement and display of the product master in step S5 is not limited to the completion of the matching process for all products, and the products for which matching has been completed may be sequentially arranged and displayed.

  Through the above steps, a shelf allocation model reproducing a photographic image is completed (step S6). Since the shelf allocation model created by this shelf allocation reproduction method includes product information, it can be used for, for example, evaluation of the product display state after the shelf allocation operation and the shelf replacement.

  As described above, in the shelf allocation reproduction method of the present embodiment, since the shelf allocation model is created from the sales floor photograph, the input work of the displayed product is not required, and the work load and the work time in the store are greatly reduced. Can do. In addition, since the image correction is performed before the matching process, the displayed product can be identified regardless of the shooting direction, shooting range, shooting size, and the like of the sales floor photo. Furthermore, in the shelf allocation reproduction method of the present embodiment, products are identified by matching processing. Therefore, each product of the created shelf allocation model is not a mere image but a product master associated with product information. Therefore, there is no need to input product information separately.

  Furthermore, in the shelf allocation reproduction method of the present embodiment, since the product master for comparison is narrowed down by size and / or product information when matching processing is performed, matching is performed using the extracted features. Matching processing can be performed with high accuracy in a short time. Furthermore, since the matching area can be set arbitrarily, even if a part of the product is hidden in the price tag or product advertisement, or when the background color is different between the product master and the photographic image, it is highly accurate. The product can be identified. As a result, the work efficiency in creating the shelf allocation model image can be improved.

  The shelf allocation reproduction method of the present embodiment can be realized by executing a program (a shelf allocation reproduction program) for executing the above-described work on a server or a terminal of each store. 4-16 is a figure which shows an example of the operation screen in the program which performs the shelf allocation reproduction method of this embodiment. When creating a shelf allocation model using the shelf allocation reproduction program, first, the program is started and the screen shown in FIG. 4 is displayed. Then, a “photo selection / correction” button is selected on the screen shown in FIG. 4 to display a “file selection” screen to call up the image file of the shelf space sales photo to be reproduced from the image storage unit. Thereby, the photograph image of the selected sales floor photograph is displayed as shown in FIG.

  Next, the displayed photographic image is corrected as necessary. For example, when correcting the color, the “image correction” screen shown in FIG. 6 is opened to adjust the brightness, contrast, hue, saturation, sharpness, and the like. When correcting the tilt of the image, the photograph is rotated on the “angle designation” screen shown in FIG. Further, in the case of correcting the size of the photographic image, if the sales floor photograph is taken from the front, the “Cropping Method” is set to “Rectangle Designation” on the screen shown in FIG. Specify the area to create the split model. When “Crop” is executed under this condition, the size of the photo image is corrected so that the size of the shelf and the fixture model in the photo image is the same, and the corrected image is displayed on the display unit.

  On the other hand, if the sales floor photograph is not taken from the front, as shown in FIG. 8, the “cutout method” is set to “4 point designation” and the area for creating the shelf allocation model is designated. When “Crop” is executed under these conditions, the distortion is corrected so that the designated area becomes rectangular as shown in FIG. 9, and the size of the shelf and the furniture model in the photographic image is the same. Thus, the size of the photographic image is corrected, and the corrected image is displayed on the display unit. When the above correction is completed, as shown in FIG. 10, the “photo shelf allocation reproduction” screen displays the corrected photographic image and the shelf allocation model format describing only the furniture model.

  Next, a range for performing the matching process is designated as necessary. For example, when matching processing is performed for each shelf, the shelf for matching is selected as shown in FIG. As a result, as shown in FIG. 12, for example, shading is also displayed at the corresponding position in the shelf allocation model format. Thereafter, when the matching process is executed, first, the products in the designated range are classified for each product as shown in FIG. At this time, the operator can check each set segment area, and manually correct those that have a deviation from the product.

  Then, matching is performed using all or a part of each divided region as a matching region. Here, if the matching area is limited to a part of the divided area, the influence of the price list, the POP advertisement, the background color, and the like can be excluded. In the matching process, the product master to be compared can be limited in advance using product information. Moreover, it is also possible to set so as to limit the product master to be compared according to the size of each segmented area.

  When the above-described matching process is completed, as shown in FIG. 14, for example, the matching result is displayed, and an image obtained by combining the product image of the product master having the highest consistency at the corresponding position of the fixture model is displayed. The

  In this program, when a product master different from the product of the photographic image is arranged in the created shelf allocation model, that is, when an incorrect product image is combined with the furniture model, as shown in FIG. In addition, the product master can be replaced by selecting the corresponding product from the list indicating the matching result. It is also possible to specify a range or individual products and perform matching processing again.

  On the other hand, if there is no difference between the selected product masters, the “decision” button is selected and the matching process is terminated. As a result, as shown in FIG. 16, an overall view reflecting the matching processing result is displayed. A shelf allocation model is completed by performing the above-described processing for all the shelves.

  According to the above-described program, most of the shelf allocation reproduction method according to the present embodiment can be automatically performed, so that a shelf allocation model can be created in a short time without manpower.

  The above-described shelf allocation reproduction program may be used in combination with the existing shelf allocation creation support program, or may be added to the existing shelf allocation creation support program.

  Next, the shelf allocation evaluation method according to the second embodiment of the present invention will be described. The shelf allocation evaluation method of the present embodiment is a method of evaluating the product display state of each store using the shelf allocation model created by the shelf allocation reproduction method of the first embodiment described above. In general, in chain stores such as convenience stores, drug stores, and supermarkets, the preparation of shelf allocation plans is performed collectively at the headquarters or outsourced to suppliers. In this way, if the person who changes the shelf and the shelf creator who creates the shelf plan are different, the shelf planner will check how much the proposed shelf plan is realized at the store. Sometimes.

  In such a case, by using the shelf allocation model created by the shelf allocation reproduction method described above, it is possible to easily compare the shelf allocation plan created by the shelf allocation creator with the product display state after the replacement. it can. Specifically, the shelf allocation creator obtains a photo of the sales floor after changing the shelf from the store. Then, the shelf allocation model is created from the sales floor photo by the shelf allocation reproduction method according to the first embodiment of the present invention, and matching processing is performed with the image of the shelf allocation plan created by the shelf allocation creator.

  Conventionally, the person in charge checked the shelf allocation plan proposed to each dealer and the sales floor after the replacement of the shelf or a photograph thereof by visually checking the shelf allocation evaluation. In the method, a shelf allocation model is created by the shelf allocation reproduction method according to the first embodiment of the present invention, and the shelf allocation model is compared with the image of the shelf allocation plan proposed to each store. In addition, the degree of realization of the shelf allocation plan can be evaluated easily. Thereby, compared with the past, the work efficiency of the shelf allocation evaluation work can be greatly improved.

  The shelf allocation evaluation method of the present embodiment can be realized by executing a program (a shelf allocation evaluation program) for executing the above-described work on the server or the shelf creator's terminal. Such a shelf allocation evaluation program may be used in combination with the existing shelf allocation creation support program and the aforementioned shelf allocation reproduction program, or added to the existing shelf allocation creation support program and the aforementioned shelf allocation reproduction program. You can also

  The above-described shelf allocation reproduction program and shelf allocation evaluation program include, for example, ROM (Read-Only Memory), nonvolatile memory, DVD (Digital Versatile Disk), MO (Magneto-Optical disk), MD (Mini Disc), It can be provided by being recorded on various recording media such as a CD (Compact Disc), a magnetic tape, and a magnetic disk. Alternatively, the program stored in the storage device of the server computer may be executed via a communication network such as a network. In that case, the storage device of this server computer is also included in the recording medium of the present invention.

It is a flowchart figure which shows the procedure of the shelf allocation reproduction method which concerns on the 1st Embodiment of this invention. It is a flowchart figure which shows the procedure of the matching process of goods. It is a figure which shows the example which set a part of division area to the matching area | region. It is a figure which shows an example of the operation screen at the time of starting in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the photograph selection and correction screen in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the image correction screen in the program which performs the shelf allocation reproduction | regeneration method of the 1st Embodiment of this invention. It is a figure which shows an example of the angle designation | designated screen in the program which performs the shelf allocation reproduction | regeneration method of the 1st Embodiment of this invention. It is a figure which shows an example of the shelf allocation creation area designation | designated screen in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the operation screen after distortion correction in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the photo shelf allocation reproduction screen in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the matching process implementation range designation | designated screen in the program which performs the shelf allocation reproduction | regeneration method of the 1st Embodiment of this invention. It is a figure which shows an example of the screen after the matching process implementation range designation | designated in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the screen after goods division in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the matching result display screen in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the matching result correction screen in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention. It is a figure which shows an example of the screen at the time of the completion of the matching process in the program which performs the shelf allocation reproduction method of the 1st Embodiment of this invention.

Explanation of symbols

1 Photo image 2a-2e Product 3a-3e Classification area 4 Shelf 5, 6, 8 POP advertisement 7 Price tag 9, 11 Matching area 10 Master product

Claims (16)

  1. A method of creating a shelf allocation model that represents a display state of a product using a computer,
    Identifying each product displayed on the sales floor from a digital image of the sales floor photo,
    Based on the identification result, placing a product master provided with product information at a corresponding position in a fixture model created in advance;
    The shelf division reproduction method which has.
  2. The identification step includes
    A step of dividing each product in the digital image of the sales floor photo into one product;
    A matching step of comparing the feature extracted from the image of each classified product with the product master and calculating a degree of coincidence thereof;
    The shelf allocation reproduction method according to claim 1, wherein:
  3.   3. The shelf layout reproduction method according to claim 2, wherein in the step of classifying each product, the image color of the digital image of the sales floor photograph is binarized and a boundary of each product is extracted.
  4.   The shelf division reproduction according to claim 3, wherein the division is corrected for a category set based on the extracted boundary and a product in the digital image of the sales floor photo that does not match. Method.
  5.   The shelf allocation reproduction method according to any one of claims 2 to 4, wherein in the matching step, a product master to be compared is limited by the product information.
  6.   The shelf according to claim 5, wherein the product information is at least one of handling status in a sales floor, inventory information, product classification, manufacturer name, brand, release date, update date, and product name. Split reproduction method.
  7.   The shelf allocation according to any one of claims 2 to 6, wherein, in the matching step, a size of each classified product is calculated, and a product master to be compared is limited based on the value. Reproduction method.
  8.   The shelf allocation reproduction method according to any one of claims 2 to 7, wherein in the matching step, only a partial region of an image of each product is compared.
  9.   The shelf allocation according to any one of claims 2 to 8, wherein a product master having the highest degree of coincidence is selected from the product master group on which the matching process has been performed, and is arranged in the fixture model. Reproduction method.
  10.   The shelf layout reproduction method according to any one of claims 1 to 9, wherein color correction and / or distortion correction are performed on the digital image of the sales floor photograph before the identification step.
  11.   The shelf allocation reproduction program which makes a computer perform the shelf allocation reproduction method of any one of Claims 1 thru | or 10.
  12.   12. The shelf replacement reproduction program according to claim 11, which is used in combination with a program that supports shelf allocation creation.
  13.   A shelf allocation evaluation method that compares a shelf allocation model created by the shelf allocation reproduction method according to any one of claims 1 to 10 with data of a previously implemented shelf allocation and calculates a degree of coincidence thereof.
  14.   A shelf allocation evaluation program for causing a computer to execute the shelf allocation evaluation method according to claim 13.
  15.   15. The shelf allocation evaluation program according to claim 14, wherein the shelf allocation evaluation program is used in combination with a program for supporting shelf allocation creation and / or a shelf allocation reproduction program according to claim 9.
  16. A computer-readable recording medium in which the shelf allocation reproduction program according to claim 11 and / or the shelf allocation evaluation program according to claim 14 is recorded.
JP2008029539A 2008-02-08 2008-02-08 Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium Pending JP2009187482A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2008029539A JP2009187482A (en) 2008-02-08 2008-02-08 Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2008029539A JP2009187482A (en) 2008-02-08 2008-02-08 Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium

Publications (1)

Publication Number Publication Date
JP2009187482A true JP2009187482A (en) 2009-08-20

Family

ID=41070621

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2008029539A Pending JP2009187482A (en) 2008-02-08 2008-02-08 Shelf allocation reproducing method, shelf allocation reproduction program, shelf allocation evaluating method, shelf allocation evaluation program, and recording medium

Country Status (1)

Country Link
JP (1) JP2009187482A (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009265998A (en) * 2008-04-25 2009-11-12 Lambda Systems Inc Image processing system, image processing method, and image processing program
JP2013528869A (en) * 2010-06-01 2013-07-11 ミレニウム エスパシオ ソフト,ソシエダ アノニマ Procedure for recognizing objects
JP2013250647A (en) * 2012-05-30 2013-12-12 Dongjoo Shin Merchandise display information tabulation method and data structure and recording medium for recording the same
JP2014002722A (en) * 2012-05-11 2014-01-09 Dassault Systemes Comparing virtual and real images in shopping experience
JP2014044478A (en) * 2012-08-24 2014-03-13 Fujifilm Corp Article collating device, article collating method and article collating program
JP2014044481A (en) * 2012-08-24 2014-03-13 Fujifilm Corp Article collation device and method and program
JP2014511537A (en) * 2011-03-01 2014-05-15 ザ プロクター アンド ギャンブル カンパニー Display of data about physical retail environment on virtual illustration of physical retail environment
JP2014153894A (en) * 2013-02-07 2014-08-25 Toshiba Tec Corp Information processor and program
JP2014229246A (en) * 2013-05-27 2014-12-08 日本電気株式会社 Detection device, method, and program
JP2016071782A (en) * 2014-10-01 2016-05-09 辛東主 Merchandise display information totalization system
WO2016147612A1 (en) * 2015-03-16 2016-09-22 日本電気株式会社 Image recognition device, system, image recognition method, and recording medium
WO2018016214A1 (en) * 2016-07-21 2018-01-25 日本電気株式会社 Image processing device, image processing method, and program
US10074121B2 (en) 2013-06-20 2018-09-11 Dassault Systemes Shopper helper

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009265998A (en) * 2008-04-25 2009-11-12 Lambda Systems Inc Image processing system, image processing method, and image processing program
JP2013528869A (en) * 2010-06-01 2013-07-11 ミレニウム エスパシオ ソフト,ソシエダ アノニマ Procedure for recognizing objects
JP2014511537A (en) * 2011-03-01 2014-05-15 ザ プロクター アンド ギャンブル カンパニー Display of data about physical retail environment on virtual illustration of physical retail environment
JP2014002722A (en) * 2012-05-11 2014-01-09 Dassault Systemes Comparing virtual and real images in shopping experience
JP2013250647A (en) * 2012-05-30 2013-12-12 Dongjoo Shin Merchandise display information tabulation method and data structure and recording medium for recording the same
JP2014044478A (en) * 2012-08-24 2014-03-13 Fujifilm Corp Article collating device, article collating method and article collating program
JP2014044481A (en) * 2012-08-24 2014-03-13 Fujifilm Corp Article collation device and method and program
US9292748B2 (en) 2013-02-07 2016-03-22 Toshiba Tec Kabushiki Kaisha Information processing apparatus and information processing method
JP2014153894A (en) * 2013-02-07 2014-08-25 Toshiba Tec Corp Information processor and program
JP2014229246A (en) * 2013-05-27 2014-12-08 日本電気株式会社 Detection device, method, and program
US10074121B2 (en) 2013-06-20 2018-09-11 Dassault Systemes Shopper helper
JP2016071782A (en) * 2014-10-01 2016-05-09 辛東主 Merchandise display information totalization system
WO2016147612A1 (en) * 2015-03-16 2016-09-22 日本電気株式会社 Image recognition device, system, image recognition method, and recording medium
WO2018016214A1 (en) * 2016-07-21 2018-01-25 日本電気株式会社 Image processing device, image processing method, and program

Similar Documents

Publication Publication Date Title
CN101809601B (en) Planogram extraction based on image processing
CN103975342B (en) The system and method for capturing and handling for mobile image
US7362918B2 (en) System and method for de-noising multiple copies of a signal
US9269013B2 (en) Using extracted image text
EP1215626B1 (en) Automatically producing an image of a portion of a photographic image
CN102132312B (en) With the method for label labelling image and computing equipment
JP4160305B2 (en) System and method for determining image similarity
US20030048950A1 (en) Retrieval and browsing of database images based on image emphasis and appeal
US20030018631A1 (en) Information search and retrieval system
US20020085219A1 (en) Method of and system for generating and viewing multi-dimensional images
EP1109132A2 (en) Method for automatic assessment of emphasis and appeal in consumer images
CN102855906B (en) Image processing equipment and image processing method
CN105701519B (en) The practical planogram scene analysis of image based on super-pixel
JP2007272684A (en) Device, method, and program for image arrangement
US20080002914A1 (en) Enhancing text in images
EP1347417A2 (en) Face detection computer program product for redeye correction
US6535636B1 (en) Method for automatically detecting digital images that are undesirable for placing in albums
US8160366B2 (en) Object recognition device, object recognition method, program for object recognition method, and recording medium having recorded thereon program for object recognition method
JP2007249588A (en) Face image registration device, method, program, and storage medium
CN102132244B (en) Image tagging user interface
US20030072486A1 (en) Albuming method with automatic page layout
JP4285290B2 (en) Image processing apparatus, image processing method, and program
JP2010045819A (en) Face detecting camera and method
US20070100713A1 (en) Identifying inventory items within an image
US7243101B2 (en) Program, image managing apparatus and image managing method