WO2016063484A1 - 画像処理装置、表示制御装置、画像処理方法、および、記録媒体 - Google Patents
画像処理装置、表示制御装置、画像処理方法、および、記録媒体 Download PDFInfo
- Publication number
- WO2016063484A1 WO2016063484A1 PCT/JP2015/005151 JP2015005151W WO2016063484A1 WO 2016063484 A1 WO2016063484 A1 WO 2016063484A1 JP 2015005151 W JP2015005151 W JP 2015005151W WO 2016063484 A1 WO2016063484 A1 WO 2016063484A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- article
- recognition
- fixture
- unit
- image processing
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 95
- 238000003672 processing method Methods 0.000 title claims description 4
- 238000001514 detection method Methods 0.000 claims abstract description 145
- 238000012937 correction Methods 0.000 description 44
- 238000000034 method Methods 0.000 description 25
- 238000010586 diagram Methods 0.000 description 18
- 238000003384 imaging method Methods 0.000 description 12
- 244000299461 Theobroma cacao Species 0.000 description 10
- 235000019219 chocolate Nutrition 0.000 description 10
- 230000010365 information processing Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 5
- 235000011888 snacks Nutrition 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000009434 installation Methods 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47F—SPECIAL FURNITURE, FITTINGS, OR ACCESSORIES FOR SHOPS, STOREHOUSES, BARS, RESTAURANTS OR THE LIKE; PAYING COUNTERS
- A47F10/00—Furniture or installations specially adapted to particular types of service systems, not otherwise provided for
- A47F10/02—Furniture or installations specially adapted to particular types of service systems, not otherwise provided for for self-service type systems, e.g. supermarkets
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
- B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
- B65G1/00—Storing articles, individually or in orderly arrangement, in warehouses or magazines
- B65G1/02—Storage devices
- B65G1/04—Storage devices mechanical
- B65G1/137—Storage devices mechanical with arrangements or automatic control means for selecting which articles are to be removed
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47F—SPECIAL FURNITURE, FITTINGS, OR ACCESSORIES FOR SHOPS, STOREHOUSES, BARS, RESTAURANTS OR THE LIKE; PAYING COUNTERS
- A47F10/00—Furniture or installations specially adapted to particular types of service systems, not otherwise provided for
- A47F10/02—Furniture or installations specially adapted to particular types of service systems, not otherwise provided for for self-service type systems, e.g. supermarkets
- A47F2010/025—Furniture or installations specially adapted to particular types of service systems, not otherwise provided for for self-service type systems, e.g. supermarkets using stock management systems
Definitions
- the present invention relates to an image processing device, a display control device, an image processing method, and a recording medium.
- Patent Document 1 describes a product management apparatus that recognizes a displayed product from an image obtained by imaging a state in which a plurality of types of products are displayed.
- Patent Document 2 describes a method of recognizing a product by cutting out a product image of the product from an image of the product displayed on the product display shelf.
- ⁇ Shooting images of shelves on which products are displayed are affected by various environments such as lighting position, shooting angle of view, and shielding.
- a boundary value indicating whether or not the product to be recognized is recognized as a specific product, and a similarity value with the specified product
- the recognition threshold value is set higher to prevent erroneous recognition.
- the recognition threshold is set lower in order to prevent recognition failure, there is a possibility that the occurrence rate of misrecognition that is recognized by other different products increases.
- the present invention has been made in view of the above problems, and an object of the present invention is to provide a technique capable of detecting a region with high probability of occurrence of recognition failure with higher accuracy.
- an image processing apparatus relates to a recognition unit that recognizes an article from a photographed image obtained by photographing the displayed article, and a fixture on which the article is displayed. Detecting means for detecting a region of an article that is included in the photographed image and not recognized by the recognition means based on fixture information.
- the display control device is an area in which the article is not recognized in a photographed image obtained by photographing the displayed article, and the article may be displayed. Is displayed on the screen as an area where the article is not recognized.
- the image processing method recognizes the article from a photographed image obtained by photographing the displayed article, and adds the photographed image to the photographed image based on furniture information related to the furniture on which the article is displayed. An area of an article that is included but not recognized is detected.
- FIG. 14 is a flowchart illustrating an example of an operation flow of the image processing apparatus according to the third exemplary embodiment of the present invention. It is a figure which illustrates illustartively the hardware constitutions of the computer (information processing apparatus) which can implement
- FIG. 1 is a functional block diagram illustrating an example of a functional configuration of the image processing apparatus 100 according to the present embodiment.
- the image processing apparatus 100 according to the present embodiment includes a recognition unit 110 and a detection unit 120.
- the direction of the arrow in the drawings shows an example, and does not limit the direction of signals between blocks.
- the directions of the arrows in the drawings show an example and do not limit the direction of signals between the blocks.
- the recognition unit 110 recognizes an article included in the photographed image from a photographed image obtained by photographing the displayed article (product).
- the method by which the recognition unit 110 recognizes an article is not particularly limited, and may be a general recognition method.
- the recognition unit 110 outputs the captured image and information indicating the article recognized from the captured image to the detection unit 120.
- the detection unit 120 receives from the recognition unit 110 a captured image and information indicating an article recognized by the recognition unit 110 from the captured image. Then, the detection unit 120 detects an area of an article that is included in the received photographed image and is not recognized by the recognition unit 110 based on the fixture information related to the fixture on which the article is displayed.
- the detection unit 120 detects the region of the article that the recognition unit 110 has not recognized from the captured image. At this time, the detection unit 120 detects an area in which the article is not recognized from the captured image based on the fixture information. Thereby, the image processing apparatus 100 can detect a region having a high probability of occurrence of recognition failure with higher accuracy.
- FIG. 2 is a functional block diagram illustrating an example of a functional configuration of the image processing apparatus 200 according to the present embodiment.
- members having the same functions as those included in the drawings described in the first embodiment are denoted by the same reference numerals.
- the image processing apparatus 200 includes a recognition unit 110, a detection unit 120, and a reception unit 210. Further, the image processing apparatus 200 may further include a storage unit 220.
- the receiving unit 210 is means for receiving a photographed image obtained by photographing the displayed article.
- the captured image is, for example, an image captured by an imaging device such as a non-fixed point camera.
- the receiving unit 210 receives this captured image from, for example, an imaging device.
- the method for receiving unit 210 to receive a captured image is not particularly limited.
- the receiving unit 210 may receive a captured image from an imaging device connected to the image processing device 200 using a USB (Universal Serial Bus) cable or the like.
- the receiving unit 210 may receive a captured image from an imaging device connected to the image processing device 200 via a network.
- the receiving unit 210 may receive a captured image from, for example, a storage device that stores the captured image.
- the receiving unit 210 receives, together with the captured image, position information (referred to as captured image information) indicating the position where the captured image was captured and / or the position of the captured fixture.
- position information referred to as captured image information
- the reception unit 210 supplies the received captured image and the captured image information associated with the captured image to the recognition unit 110.
- the storage unit 220 stores information for recognizing articles included in the captured image. Specifically, in the storage unit 220, information on an image of an article and / or a feature amount included in the image of the article identifies the article (for example, an identifier for identifying the article, an article name, etc.). It is linked to and stored. The information stored in the storage unit 220 may be information necessary for recognizing an article. In addition, information indicating the type (category) of the article is associated with the information for identifying the article.
- the storage unit 220 stores fixture information related to fixtures on which articles are displayed.
- the fixture information includes installation position information indicating a location where the fixture is installed for each fixture.
- the fixture information includes the following (a) to (e) associated with the installation position information.
- B Information on shelf allocation recommended for furniture,
- C The number of stages of fixtures,
- D the number of articles that can be placed on each stage of the fixture,
- E Information indicating a condition related to display of an article displayed on a fixture.
- the number of articles that can be arranged in each stage of the fixture can be rephrased as the number of slots (number of slots) in which articles can be arranged in each stage of the fixture.
- (a) above is stored in the storage unit 220 by the recognition unit 110.
- the above (b) to (e) are stored in the storage unit 220 in advance.
- the storage unit 220 may store a fixture size as fixture information.
- (e) above includes, for example, information indicating a condition that articles are arranged in all slots. Further, (e) may include, for example, a condition that articles having the same article name are continuously arranged, and a condition that articles of the same type are displayed at close positions.
- the storage unit 220 stores information (referred to as ordering information) indicating the ordering (or purchase) of goods, and sales information managed by a POS (Point Of Sale) system or the like.
- the information for recognizing the article, the fixture information, the ordering information, and the sales information may be stored in the same storage device (for example, the storage unit 220), or may be stored in different storage devices. It may be a thing.
- the storage unit 220 may be built in the image processing apparatus 200, or may be realized by a storage device separate from the image processing apparatus 200.
- the recognition unit 110 receives a captured image from the reception unit 210.
- the recognizing unit 110 refers to the information for recognizing the article stored in the storage unit 220, and recognizes the article included in the photographed image from the received photographed image.
- the method by which the recognition unit 110 recognizes an article is not particularly limited as in the recognition unit 110 of the image processing apparatus 100 according to the first embodiment, and may be a general recognition method.
- the recognition unit 110 outputs a captured image and information (recognition result) indicating an article recognized from the captured image to the detection unit 120.
- FIG. 3 is a diagram for explaining the operation of the recognition unit 110 of the image processing apparatus 200 according to the present embodiment, and is a diagram illustrating an example of a captured image.
- FIG. 4 is a diagram for explaining a recognition result by the recognition unit 110.
- the captured image includes images of a plurality of articles 31.
- the photographed image is an image obtained by photographing one whole fixture as shown in FIG. 3, but is not limited to this.
- the captured image may be a photograph of a plurality of fixtures. Further, the captured image may be an image of a part of one fixture.
- the character described in each article 31 indicates the article name.
- the fixture on which the article is displayed has three stages.
- the top row (referred to as the first row), three items with the item name “Snack A” and two items with the item name “Snack B” are displayed.
- two articles “ ⁇ A” to “ ⁇ C” are displayed.
- articles of “Chocolate A” to “Chocolate H” are displayed one by one.
- a broken-line frame (article image region 32) surrounding each item 31 indicates an image region of the item 31 recognized by the recognition unit 110. Therefore, in FIG. 3, it is assumed that the article whose name is “Chocolate C” and the two articles “ ⁇ B” are articles that have not been recognized by the recognition unit 110.
- the recognition unit 110 includes information indicating the recognized article (for example, an article name), image area information indicating the article image area 32 of the article 31, and a recognition score indicating the probability of the recognition result of the article 31. It outputs to the detection part 120 as a recognition result.
- An example of the recognition result is shown in FIG.
- the image area information is position information indicating the position of the article image area 32 of the article 31 in the captured image.
- the recognition result includes, for each article, an article name indicating the recognized article, a recognition score of the article 31, and position information (image area information) indicating the article image area 32 of the article 31. , Is included.
- the article name, the article score, and the image area information are associated with each other.
- one row includes information on one recognized article 31.
- the recognition unit 110 recognizes an article whose article name is “Snack A”.
- the coordinates of the four corners of the article image area 32 of this article are (x1, y1), (x2, y2), (x3, y3), (x4, y4), respectively, and the recognition score is “0.80”.
- the recognition score is “0.80”.
- the recognition unit 110 outputs “snack A, 0.80, x1, y1, x2, y2, x3, y3, x4, y4” as the recognition result of “snack A” as shown in FIG.
- the article image region 32 is preferably a rectangle that circumscribes the article recognized by the recognition unit 110, but is not limited thereto.
- the article image area 32 may be an area that matches the shape of the article.
- the recognition result is not limited to this, and may include, for example, an identifier indicating the item 31 in addition to the item name or instead of the item name. That is, the recognition unit 110 may include information indicating the recognized article 31 in the recognition result. Further, the image area information included in the recognition result is not the coordinates of the four corners of the article image area 32 indicated by a substantially rectangular shape, but, for example, the coordinates of one of the four corners, the width of the article, and the height of the article. May be. That is, the recognition unit 110 may include information indicating the article image region 32 of the recognized article 31 in the recognition result.
- the above-described recognition score has an upper limit of 1.0, and a value closer to 1.0 indicates higher reliability.
- the method of expressing the recognition score is not limited to this.
- the recognition unit 110 may include, for example, a JAN (Japan Article Number) code and information indicating the type of the item (for example, the type name or type code of the item) in the recognition result.
- a JAN Joint Article Number
- the recognition unit 110 may perform control such that information regarding an article having a recognition score smaller than a predetermined value is not included in the recognition result.
- the recognition unit 110 outputs such a recognition result to the detection unit 120 together with the captured image in which the article is recognized and the captured image information of the captured image.
- the recognition unit 110 receives the detection result from the detection unit 120 and recognizes the article based on the detection result. This operation will be described after the operation of the detection unit 120 is described.
- the detection unit 120 receives the captured image, the captured image information of the captured image, and the recognition result of the article with respect to the captured image from the recognition unit 110. Then, the detection unit 120 detects an area of an article that is included in the received photographed image and is not recognized by the recognition unit 110 based on the fixture information related to the fixture on which the article is displayed. At this time, the fixture information to be referred to is associated with the installation position information that matches or approximates the received captured image information.
- the detection unit 120 detects an area on the photographed image where an article is not recognized. Then, using the captured image and the recognition result, the detection unit 120 detects a candidate area (referred to as a candidate area) of the areas that may have a recognition failure among the detected areas.
- This candidate area is an area where an article is highly likely to be placed.
- the detection unit 120 calculates the distance between the upper and lower sides and the left and right sides of the article (referred to as the inter-article distance).
- the detection unit 120 determines whether or not the calculated distance is greater than a predetermined threshold, and when the calculated distance is greater than the predetermined threshold, the area where the distance is calculated is detected as a candidate area.
- the detection unit 120 is the upper end of the fixture and the upper end of the article image area 32 of each article recognized by the recognition unit 110, and the article on which no other article is arranged above the article.
- the distance from the upper end of the article image area 32 is calculated. That is, when all the articles are arranged in the first stage, which is the uppermost stage of the fixture, the detection unit 120 calculates the distance between the upper end of each article image region 32 of the first article and the upper end of the fixture. calculate. If an article is not placed in a slot in the first row, the distance between the article placed in the second (or second and later) slot immediately below this slot and the upper end of the fixture is calculated. To do. Then, the detection unit 120 determines whether or not the calculated distance is greater than a predetermined threshold (first predetermined threshold). If the calculated distance is greater than the predetermined threshold, the area in which the distance is calculated Are detected as candidate regions.
- a predetermined threshold first predetermined threshold
- the detection unit 120 calculates the distance from the lower end, the left end, and the right end of the fixture to the article, and when the calculated distance is greater than a predetermined threshold, detects the calculated area as a candidate area.
- a predetermined threshold detects the calculated area as a candidate area.
- the left end line which shows the left end of a fixture is shown using the dashed-dotted line.
- the detection unit 120 detects the candidate area by calculating the distance from the outer edge of the fixture to the article.
- the predetermined threshold value may be stored in the storage unit 220 as fixture information.
- the detection unit 120 detects the region (34) including the part of the bidirectional arrow (33) in FIG. 3, the region of the article “ ⁇ B”, and the region of the article “chocolate C” as candidate regions. .
- the number of slots in the first stage of the fixture shown in the captured image of FIG. 3 is 5, the number of slots in the second stage is 6, and the number of slots in the third stage is 8.
- the information indicating the number of slots is stored in the storage unit 220 as fixture information as described above.
- the detection part 120 detects the area
- the third row in FIG. 3 has 8 slots as described above.
- the detection unit 120 receives information on seven articles as a third-stage recognition result. Therefore, the third level does not satisfy the condition that articles are arranged in all slots. Therefore, the detection unit 120 detects that there is a recognition failure of one article in the third stage. And the detection part 120 detects the area
- the detection unit 120 detects that there are omissions in recognition of two articles based on the condition that articles are arranged in all slots, as in the third-tier region.
- the detection unit 120 determines whether or not the width and / or height of the region 34 exceeds a second predetermined threshold. Then, when the width and / or height of the region 34 exceeds the second predetermined threshold, the detection unit 120 detects the region as a region having a high possibility of recognition failure. In this example, it is assumed that the width and / or height of the region 34 does not exceed the second predetermined threshold. Therefore, the detection unit 120 determines that the region 34 is not a recognition failure region.
- the second predetermined threshold value may be, for example, an average size of articles displayed on a fixture or a fixed value. Further, the average size and the fixed value may be set for each fixture, or may be set for each stage or row of fixtures. In addition, the second predetermined threshold value may be stored in the storage unit 220 as fixture information.
- the average size may be obtained from the recognition result received by the detection unit 120 from the recognition unit 110. As a result, the detection unit 120 can detect an area where there is a high possibility of recognition failure based on the size of an article around the recognition-rejected article.
- the detection unit 120 may detect an area that has a high possibility of recognition failure among the candidate areas based on whether or not the amount of feature points of the image of the candidate area is larger than a predetermined value. For example, since no articles are arranged in the region 34, the number of feature points is reduced. On the other hand, in the area of the article “ ⁇ B ”, since the article is arranged, the number of feature points is larger than that in the case where no article is arranged. Based on these results, the detection unit 120 determines that the region 34 is not a region with a high possibility of recognition failure, and determines that the region of the article “ ⁇ B” is a region with a high possibility of recognition failure. To do.
- the detection unit 120 is included in the distance between the recognized articles, the distance between the recognized article and the outer edge of the fixture, the size of the recognized article, and the area where the article is not recognized. Based on at least one of the number of feature points, an unrecognized article region is detected. Thereby, the detection part 120 can detect the area
- the detection unit 120 may detect an area where there is a high possibility of recognition failure using past display results as fixture information. For example, when the past display result indicates that all the articles shown in FIG. 3 have been recognized, the detection unit 120 compares the past display result with the recognition result output from the recognition unit 110. To do. Then, the detection unit 120 detects the difference area obtained by the comparison as an area where the possibility of recognition failure is high.
- the detection unit 120 uses the difference area obtained by comparing the recognition result with the shelf allocation information recommended for the fixture to be imaged as the fixture information in a region where there is a high possibility of recognition failure. You may detect as there being.
- the detection unit 120 outputs, to the recognition unit 110, information indicating a detected region having a high possibility of recognition failure as a detection result.
- the image processing apparatus 200 uses a photographed image obtained by photographing a fixture in which articles are displayed side by side on each stage.
- the present invention is not limited to this.
- the image processing apparatus 200 may use an image obtained by photographing a fixture in which articles are arranged vertically.
- the number of stages of fixtures is the number of slots that can be arranged in the row of fixtures. Therefore, the detection unit 120 can detect an area where there is a high possibility of recognition failure even using a captured image obtained by capturing such a fixture.
- the detection unit 120 further detects the article image region 32 of the misrecognized article. Below, the method to detect the misrecognized article which the detection part 120 detects is demonstrated.
- the detection unit 120 determines an article that does not satisfy this condition among recognized articles as an erroneously recognized article.
- the same type of articles is included in the condition that they are displayed at close positions. explain. Articles of the same type are often placed in close proximity within the same fixture.
- the recognition result includes information indicating the type of the article. Based on the information indicating the type of the article included in the recognition result and the image area information of the article, the detection unit 120 determines that an article that does not satisfy this condition among the recognized articles is an erroneously recognized article. To do.
- the detection unit 120 determines a misrecognized article based on order information prior to the time when the captured image was captured. Specifically, the detection unit 120 compares the recognized article with the ordered (purchased) article, and determines that the recognized article is an erroneously recognized article when the recognized article is not ordered. .
- the detection unit 120 determines a misrecognized article based on sales information before the time when the photographed image was photographed. Specifically, the detection unit 120 confirms sales information regarding the recognized article, and determines that the article is an erroneously recognized article if the article is a sold article.
- the detection unit 120 may determine the misrecognized article based on other fixture information. For example, when the article included in the past display result is significantly different from the recognized article (for example, in the case of an article having a different category), the detection unit 120 determines that the article is an erroneously recognized article. Also good. In addition, when the recommended shelf allocation information for a fixture that is a recognition target of an article is significantly different from the recognized article (for example, an article with a different category), the detection unit 120 may misrecognize the article. It may be determined that the article has been changed. Moreover, the detection part 120 may determine the articles
- the detection part 120 outputs the information which shows the goods image area
- FIG. 5 is a diagram for explaining the detection result output by the detection unit 120.
- the detection unit 120 detects a substantially rectangular region having (x1, y1), (x2, y2), (x3, y3), and (x4, y4) as the coordinates of the four corners as the recognition failure region. And, as the article image region 32 of the article in which the detection unit 120 is erroneously recognized, (x′1, y′1), (x′2, y′2), (x′3, y′3), (x Assume that a substantially rectangular region having coordinates of four corners of '4, y'4) is detected. At this time, as shown in FIG.
- the detection unit 120 includes, as a recognition failure detection result, a “recognition failure” that includes a character string “recognition failure” that indicates a recognition failure and information that indicates a region where the recognition failure has been detected. Leak, x1, y1, x2, y2, x3, y3, x4, y4 "are output.
- the detection unit 120 detects, as a misrecognition detection result, a character string “misrecognition” indicating misrecognition and information (position information) indicating the article image region 32 of the misrecognized article. Recognition, x′1, y′1, x′2, y′2, x′3, y′3, x′4, y′4 ”are output. As described above, in FIG. 5, one row includes information of one detected area.
- the area detected by the detection unit 120 is not limited to a substantially rectangular shape, and may have any shape.
- the information indicating the region included in the detection result output by the detection unit 120 is not limited to the coordinates of the four corners, and may be any information that represents the region detected by the detection unit 120.
- the detection result shown in FIG. 5 is an example, and is not limited to this format.
- the recognition unit 110 receives the detection result from the detection unit 120 and recognizes the article based on the detection result.
- the recognition unit 110 When the recognition unit 110 receives the detection result from the detection unit 120, the recognition unit 110 recognizes the article again on the region indicated by the received detection result on the captured image. At this time, the recognition unit 110 recognizes the article by changing a recognition parameter set when the article is recognized based on the following (A) to (C).
- A) Recognition result by the recognition unit 110 (recognition result including information on an article already recognized by the recognition unit 110),
- B Past display results,
- C Information on shelf allocation recommended for furniture.
- the recognition parameter set when the article is recognized is, for example, the likelihood of the article indicated by the information used for recognition stored in the storage unit 220.
- the recognition unit 110 calculates the likelihood for an article that is determined to have a high possibility of recognition failure for at least one of the recognition failure areas based on at least one of the above (A) to (C).
- the case where the recognition unit 110 calculates the likelihood is described based on (A), for example.
- chocolates A, B, and D to H are displayed in the same level (third level) as “Chocolate C” that has not been recognized.
- the recognition unit 110 recognizes an article having an article name similar to the article name of the article displayed on the same stage as the stage where the area where the article is not recognized exists (in this case, “chocolate” ”Is added to the item name).
- the recognition unit 110 applies this article to an article of the same type as the article type with respect to the recognized article arranged around the recognition failure area. Increase the likelihood of.
- the recognition unit 110 calculates the likelihood based on (B), for example.
- the recognition unit 110 increases the likelihood of an article that is included in the past display result and is arranged at the same position and / or around the position where the recognition omission is present.
- the recognition unit 110 is an article included in the past display result, and is arranged at the same position as the recognition failure area and / or around the position. The likelihood of this article is increased for an article of the same type as that of the article.
- the recognition unit 110 calculates the likelihood based on (C), for example.
- the recognizing unit 110 increases the likelihood of an article that is included in the recommended shelf allocation information and that is arranged at the same position as the area where recognition is not performed and / or around the position.
- the recognition unit 110 is the article included in the recommended shelf allocation information, and is the same position as the area where the recognition is omitted and / or around the position. The likelihood of this article is increased with respect to an article of the same type as that of the article placed in the box.
- the recognition unit 110 may set the likelihood (recognition parameter) depending on whether the region in which the article is recognized is an erroneously recognized region or an unrecognized region. For example, when the area for recognizing the article is the article image area 32 of the misrecognized article, the recognition unit 110 performs the misrecognition so that the article included in this area is not recognized again as the misrecognized article. Reduce the likelihood of the item being made.
- the recognition unit 110 can narrow down the information used for recognition stored in the storage unit 220 by changing the recognition parameters. Thereby, the recognition part 110 can shorten the time of a recognition process.
- the recognition unit 110 recognizes an article with respect to the region indicated by the detection result based on the calculated likelihood. Thereby, the recognition part 110 can suppress misrecognition.
- the recognition unit 110 outputs the recognition result to the detection unit 120.
- the recognition unit 110 adds information indicating the article recognized for the unrecognized region to the previous recognition result. Further, the recognition unit 110 deletes information on the article determined to be erroneously recognized by the detection unit 120 from the previous recognition result, and the article recognized this time for the article image area 32 of the article determined to be erroneously recognized. Is added to the previous recognition result. Thereby, the recognition unit 110 can output a new recognition result. Then, the recognition unit 110 outputs the recognition result to the detection unit 120 until it receives a detection result from the detection unit 120 that there is no region indicating recognition failure or erroneous recognition.
- the recognition unit 110 When the recognition unit 110 receives a detection result from the detection unit 120 indicating that there is no region indicating a recognition failure or misrecognition, the recognition result output to the detection unit 120 immediately before receiving the detection result is taken by performing article recognition.
- the result of recognition of the article with respect to the image is stored in the storage unit 220.
- the recognition result stored in the storage unit 220 by the recognition unit 110 is a past display result when viewed from the recognition result of a photographed image obtained by photographing the same fixture. Therefore, the recognition unit 110 and the detection unit 120 can perform processing of each unit based on the past display result.
- the recognition unit 110 may store the recognition result after receiving a predetermined number of detection results from the detection unit 120 in the storage unit 220 as a recognition result for the captured image.
- the recognition unit 110 may store the recognition result in the storage unit 220 when transmitting the recognition result to the detection unit 120.
- FIG. 6 is a flowchart showing an example of the operation flow of the image processing apparatus 200 according to the present embodiment.
- the receiving unit 210 receives a captured image (step S1).
- the recognition unit 110 recognizes an article for the captured image received by the reception unit 210 in step S1 (step S2).
- the detection unit 120 detects, based on the fixture information, a region in which the article is not recognized from the photographed image and has a high possibility that the article is included (recognition failure region) (Step S120). S3). In addition, the detection unit 120 determines an article in which an article is recognized but has a high possibility of erroneous recognition, and detects the article image region 32 of the article (step S4). Note that step S3 and step S4 may be performed simultaneously or in reverse order.
- the recognition unit 110 confirms whether or not a recognition failure and / or erroneous recognition area has been detected. Specifically, the recognizing unit 110 checks whether or not a recognition failure and / or misrecognition region is indicated in the detection result received from the detecting unit 120 (step S5). If the detection result indicates a region of recognition failure and / or misrecognition (YES in step S5), recognition unit 110 performs re-recognition of the article in the region indicated by the detection result of detection unit 120. (Step S6). Then, the detection unit 120 executes Step S3 again to detect a recognition failure region.
- the recognition unit 110 stores the recognition result in the storage unit 220 and ends the process (step S7).
- (effect) According to the image processing apparatus 200 according to the present embodiment, it is possible to detect a region having a high probability of occurrence of recognition failure with higher accuracy. This is because the detection unit 120 detects the region of the article that the recognition unit 110 did not recognize from the captured image based on the fixture information.
- a recognition threshold is set, and a recognition result is output based on the recognition threshold.
- the recognition unit 110 sets the recognition threshold value lower, the incidence of erroneous recognition increases. Therefore, the recognition unit 110 according to the present embodiment sets the recognition threshold value to a value that more effectively suppresses the occurrence rate of erroneous recognition.
- the recognition threshold is set high, there is a high possibility that the recognition failure area increases.
- the detection unit 120 can detect such a recognition failure area based on the fixture information.
- the recognition unit 110 can recognize the article again by changing the recognition parameter only for the detected unrecognized region. Thereby, omission of recognition and misrecognition can be prevented more.
- the image processing apparatus 200 can acquire information indicating the more accurate shelf allocation from the captured image.
- FIG. 7 is a functional block diagram illustrating an example of a functional configuration of the image processing apparatus 300 according to the present embodiment.
- the image processing apparatus 300 according to the present embodiment is configured to further include a display control unit and a correction unit in addition to the image processing apparatus 200 according to the second embodiment described above.
- FIG. 7 is a diagram illustrating an example of a system configuration including the image processing apparatus 300 according to the present embodiment.
- the system includes an image processing device 300, an imaging device 400, and a display operation device 500.
- the image processing apparatus 300 is communicably connected to the imaging apparatus 400.
- the imaging device 400 images the displayed article. Then, the imaging apparatus 400 transmits the captured image (captured image) to the image processing apparatus 300.
- the imaging device 400 is realized by, for example, a non-fixed point camera.
- the display operation device 500 is communicably connected to the image processing device 300.
- the display operation device 500 may be connected to the image processing device 300 via a network, or may be directly connected to the image processing device 300.
- the display operation device 500 includes a display unit 510 and an input unit 520.
- the display operation device 500 is described as being configured separately from the image processing device 300, but the display operation device 500 is formed integrally with the image processing device 300. May be.
- the display operation device 500 is, for example, a touch panel in which the display unit 510 and the input unit 520 are integrally formed.
- the display unit 510 is a display device that displays a GUI (Graphical User Interface) or the like that is operated by the user on the screen based on a signal transmitted from the image processing apparatus 300.
- GUI Graphic User Interface
- the input unit 520 is a device that detects a user instruction.
- the input unit 520 detects the position (coordinates on the screen) of the instruction made on the screen. For example, when the display operation device 500 is a touch panel, the input unit 520 detects an input operation by an object that has touched / closed on the screen. Further, the input unit 520 detects an input operation input by a user operating a mouse or the like, for example.
- the input unit 520 transmits the detection result as an input operation signal to the image processing apparatus 300.
- FIG. 8 is a functional block diagram illustrating an example of a functional configuration of the image processing apparatus 300 according to the present embodiment.
- the image processing apparatus 300 includes a recognition unit 110, a detection unit 120, a reception unit 210, a display control unit 310, and a correction unit 320. Further, the image processing apparatus 300 may further include a storage unit 220.
- the reception unit 210 receives the captured image transmitted from the imaging device 400 and outputs the received image to the recognition unit 110.
- an article image to be displayed on the display unit 510 is stored in association with information for identifying the article. Yes.
- the article image to be displayed on the display unit 510 may be an article image stored as information necessary for recognizing the article, or may be a sample image showing the article.
- the recognition unit 110 recognizes an article included in the captured image from the received captured image, similarly to the recognition unit 110 according to the second embodiment. Then, the recognition unit 110 outputs the recognition result to the detection unit 120.
- the recognition unit 110 when the recognition unit 110 receives the detection result from the detection unit 120, the recognition unit 110 changes the recognition parameter based on the detection result, and recognizes the article again in the region indicated by the detection result. Then, the recognition unit 110 outputs the recognition result to the detection unit 120.
- the detection unit 120 detects a region where there is a high possibility of recognition failure, as with the detection unit 120 according to the second embodiment. Moreover, the detection unit 120 may further have a function of detecting the article image region 32 of the misrecognized article, similarly to the detection unit 120 according to the second embodiment.
- the detection unit 120 checks whether or not the recognition result has been received from the recognition unit 110 a predetermined number of times. If the recognition result has been received a predetermined number of times, the detection unit 120 outputs the detection result and the received recognition result to the display control unit 310.
- the detection unit 120 detects an area of recognition failure or misrecognition after receiving the first recognition result, and recognizes the detection result, the received recognition result, and the article.
- the captured image is output to the display control unit 310.
- the display control unit 310 receives the detection result, the recognition result, and the captured image from the detection unit 120. Then, the display control unit 310 causes the display unit 510 to display information indicating the area indicated by the detection result. For example, the display control unit 310 generates an image indicating the region indicated by the detection result using the received captured image, or extracts an image indicating the region from the received captured image and displays the image on the display unit 510.
- the display control unit 310 determines an article candidate (referred to as a correction candidate) that may exist in the region based on at least one of the following (1) to (3).
- the display control unit 310 identifies an article placed in a region around the region indicated by the detection result from the received recognition result. Then, the display control unit 310 determines the articles arranged in the surrounding area as correction candidates.
- the display control unit 310 acquires an article image indicating the determined correction candidate article from the storage unit 220. Then, the display control unit 310 causes the display unit 510 to display the acquired article image so that the user can select it.
- the display control unit 310 may calculate the degree of duplication (similarity degree) of each article based on the received recognition result, and determine a correction candidate based on the degree of duplication. For example, the display control unit 310 may calculate the degree of duplication of the types of articles displayed in the stage including the region indicated by the detection result, and may determine the type of articles having a higher degree of duplication as correction candidates. . At this time, the determined correction candidate may be an article included in the recognition result, or an article whose information is stored in the storage unit 220 although it is not included in the recognition result.
- the display control unit 310 has at least one of recognition results, information indicating conditions regarding the display of the items displayed on the fixture, order information, and sales information for the items displayed on the display unit 510 as correction candidates.
- the likelihood may be determined based on either.
- the display control part 310 may determine a correction candidate based on the determined likelihood.
- the information indicating the conditions related to the display of the articles displayed on the fixture is, for example, the condition that articles having the same article name are continuously arranged, and the same type of articles are in close positions. It is a condition that it is displayed.
- the display control unit 310 can narrow down correction candidates to be displayed on the display unit 510.
- the display control unit 310 displays the correction candidates to be displayed on the display unit 510 in descending order of possibility of existing in the area.
- the image processing apparatus 300 can more easily understand an article that is likely to exist in the region among the correction candidates, and can present it to the user.
- the display control unit 310 may determine the order in which there is a high probability of being present in this area, based on the article name and / or the type of the article arranged at a position close to the area. Good.
- the display control unit 310 may determine the order in which there is a high possibility of being present in this area, based on the recognition score included in the recognition result of the article for this area. At this time, the display control unit 310 instructs the recognition unit 110 to recognize the article again for this area, and receives the recognition result of the article for this area. It is preferable that the recognition result includes a plurality of articles that are likely to exist in the region together with the recognition score. Then, the display control unit 310 determines the descending order of the recognition scores of the plurality of articles included in the recognition result as the descending order of possibility of existing in the region.
- the display control unit 310 causes the display unit 510 to display the determined correction candidates so that they can be selected.
- the number of correction candidates displayed on the display unit 510 by the display control unit 310 is not particularly limited.
- the display control unit 310 selects the article image stored in the storage unit 220 in the display unit 510 by the user. It may be displayed as possible. If the article image of the article desired by the user is not stored in the storage unit 220, the display unit 510 may register the article image in the storage unit 220 based on the user instruction transmitted from the input unit 520. Good.
- the display control unit 310 supplies information indicating the image displayed on the display unit 510 and the received recognition result to the correction unit 320.
- the correction unit 320 receives an input operation signal indicating the input operation detected by the input unit 520. Further, the correction unit 320 receives information indicating the image displayed on the display unit 510 by the display control unit 310 and the recognition result from the display control unit 310. Then, the correction unit 320 corrects the recognition result based on the selection result for the article candidate displayed on the display unit 510 indicated by the received input operation signal.
- the correction unit 320 adds information on the correction candidates selected by the user to the recognition result.
- the correction unit 320 displays information on the article that has been determined to be erroneously recognized by the detection unit 120 from the recognition result. delete. Then, the correction unit 320 adds information related to the correction candidate selected by the user to the recognition result. Thereby, the recognition unit 110 can output a new recognition result (corrected recognition result).
- the correction unit 320 stores the corrected recognition result in the storage unit 220 as the recognition result of the article with respect to the photographed image in which the article is recognized.
- the recognition result stored in the storage unit 220 by the correction unit 320 is a past display result when viewed from the recognition result of a photographed image obtained by photographing the same fixture. Therefore, the recognition unit 110, the detection unit 120, and the display control unit 310 can perform processing of each unit based on the past display result.
- FIG. 9 to 11 are diagrams showing examples of display images displayed on the display unit 510 by the display control unit 310 of the image processing apparatus 300 according to the present embodiment.
- the display control unit 310 causes the display unit 510 to display information indicating an area of recognition failure and / or misrecognition in the captured image.
- the display control unit 310 indicates a region in which a substantially rectangular thick frame (the article image region 32 described above) is not displayed as a recognition failure and / or erroneous recognition region.
- the input unit 520 uses the information indicating the position selected by the user as an input operation signal as an input operation signal. It transmits to the processing apparatus 300.
- the correction unit 320 identifies the selected region from the input operation signal transmitted from the input unit 520, and transmits information indicating the region to the display control unit 310.
- the display control unit 310 causes the display unit 510 to display an image of a recognition failure and / or misrecognition region selected by the user based on the information received from the correction unit 320. For example, when the user selects the “chocolate C” portion in FIG. 9, the display control unit 310 causes the display unit 510 to display an image of the “chocolate C” portion as shown on the left side of FIG. 10. Since the left diagram in FIG. 10 shows a region to be corrected, it is also called a correction target region.
- the display control part 310 displays the correction candidate with respect to this correction object area
- the correction unit 320 corrects the recognition result based on the selection result.
- the display control unit 310 may cause the display unit 510 to display an image including only a recognition failure and / or misrecognition region as illustrated in FIG. 11 instead of the screen illustrated in FIG. 9.
- the display control unit 310 may display the screen of FIG. 9 and the screen of FIG. 10 or the screen of FIG. 11 and the screen of FIG. 10 on one screen.
- FIG. 12 is a flowchart showing an example of the operation flow of the image processing apparatus 300 according to the present embodiment.
- steps S11 to S14 are the same processes as steps S1 to S4 in the operation of the image processing apparatus 200 according to the second embodiment, the description thereof is omitted.
- the detection unit 120 checks whether or not the recognition result has been received from the recognition unit 110 a predetermined number of times (step S15). When the recognition result has not been received a predetermined number of times (NO in step S15), the recognition unit 110 performs re-recognition of the article in the region indicated by the detection result by the detection unit 120 (step S16). Then, the detection unit 120 executes Step S3 again to detect a recognition failure region.
- step S15 When the recognition result is received a predetermined number of times (YES in step S15), the display control unit 310 controls the display unit 510 to display the correction candidates on the screen. Then, display unit 510 displays the correction candidates on the screen (step S17).
- the correction unit 320 corrects the recognition result based on the selection result (step S18). And the correction part 320 stores a recognition result in the memory
- the image processing apparatus 300 according to the present embodiment can obtain the same effects as those of the image processing apparatuses according to the first and second embodiments described above.
- the display control unit 310 displays on the screen a candidate for an article that may exist in the region detected by the detection unit 120 so as to be selectable. Therefore, the image processing apparatus 300 according to the present embodiment can reduce the complexity of the correction work by the user. Thereby, the image processing apparatus 300 can acquire information indicating the more accurate shelf allocation from the captured image without increasing the burden on the user.
- the display control unit 310 is built in the image processing apparatus 300, but the display control unit 310 is realized as a display control apparatus separate from the image processing apparatus 300. It may be a thing.
- This display control device is an area in which an article is not recognized in an image obtained by photographing the displayed article, and an area in which the article may be displayed is an area in which the article is not recognized. Is displayed on the screen of the display unit 510. Thereby, the display control apparatus can present a highly probable region where recognition failure has occurred to the user.
- Example of hardware configuration a configuration example of hardware capable of realizing the image processing apparatuses (100, 200, 300) according to the above-described embodiments will be described.
- the above-described image processing apparatuses (100, 200, 300) may be realized as a dedicated apparatus, but may be realized using a computer (information processing apparatus).
- FIG. 13 is a diagram illustrating a hardware configuration of a computer (information processing apparatus) capable of realizing each embodiment of the present invention.
- the hardware of the information processing apparatus (computer) 10 shown in FIG. 13 includes the following members.
- CPU Central Processing Unit
- I / F Communication interface
- I / F Communication interface
- I / F Communication interface
- I / F Communication interface
- I / F Communication interface
- ROM Read Only Memory
- -RAM Random Access Memory
- the input / output user interface 13 is a man-machine interface such as a keyboard which is an example of an input device and a display as an output device.
- the communication interface 12 is a general communication means for the devices according to the above-described embodiments (FIGS. 1, 2, and 8) to communicate with an external device via the communication network 20.
- the CPU 11 controls the overall operation of the information processing apparatus 10 that implements the image processing apparatuses (100, 200, 300) according to the embodiments.
- a program (computer program) that can realize the processing described in each of the above-described embodiments is supplied to the information processing apparatus 10 illustrated in FIG. This is realized by reading out and executing.
- the program is, for example, the various processes described in the flowcharts (FIGS. 6 and 12) referred to in the description of the above embodiments, or the block diagrams shown in FIGS. It may be a program capable of realizing each unit (each block) shown in the apparatus.
- the program supplied to the information processing apparatus 10 may be stored in a readable / writable temporary storage memory (15) or a non-volatile storage device (17) such as a hard disk drive. That is, in the storage device 17, the program group 17 ⁇ / b> A is a program that can realize the functions of the respective units shown in the image processing devices (100, 200, 300) in the above-described embodiments, for example.
- the various kinds of stored information 17B are, for example, a captured image, information for recognizing an article, a recognition result, a detection result, fixture information, sales information, ordering information, an article image, and the like in each of the above-described embodiments.
- the structural unit of each program module is not limited to the division of each block shown in the block diagrams (FIG. 1, FIG. 2, and FIG. 8). May be selected as appropriate during mounting.
- a method for supplying a program into the apparatus can employ a general procedure as follows.
- -CD Compact Disk
- a method of installing in the apparatus via various computer-readable recording media (19) such as ROM and flash memory A method of downloading from the outside via a communication line (20) such as the Internet.
- each embodiment of the present invention can be considered to be configured by a code (program group 17A) constituting the computer program or a storage medium (19) in which the code is stored. .
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Operations Research (AREA)
- Human Resources & Organizations (AREA)
- Mechanical Engineering (AREA)
- Game Theory and Decision Science (AREA)
- Image Analysis (AREA)
- Cash Registers Or Receiving Machines (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Display Racks (AREA)
- Multimedia (AREA)
Abstract
Description
本発明の第1の実施の形態について、図面を参照して説明する。本実施の形態では、本発明の課題を解決する基本の構成について説明する。図1は、本実施の形態に係る画像処理装置100の機能構成の一例を示す機能ブロック図である。図1に示す通り、本実施の形態に係る画像処理装置100は、認識部110と、検出部120とを備えている。また、図面中の矢印の方向は、一例を示すものであり、ブロック間の信号の向きを限定するものではない。以降に参照する、他のブロック図においても同様に、図面中の矢印の方向は、一例を示すものであり、ブロック間の信号の向きを限定するものではない。
次に、上述した第1の実施の形態を基本とする、本発明の第2の実施の形態について、図面を参照して説明する。図2は、本実施の形態に係る画像処理装置200の機能構成の一例を示す機能ブロック図である。なお、説明の便宜上、前述した第1の実施の形態で説明した図面に含まれる部材と同じ機能を有する部材については、同じ符号を付す。
(a)什器を撮影した撮影画像の撮影時刻より過去の時刻に該什器を撮影した撮影画像に対する、認識部110による認識結果(過去の陳列結果と呼ぶ)、
(b)什器において推奨される棚割りの情報、
(c)什器の段数、
(d)什器の各段に配置可能な物品数、
(e)什器に陳列された物品の陳列に関する条件を示す情報。
(A)認識部110による認識結果(認識部110が既に認識した物品の情報が含まれる認識結果)、
(B)過去の陳列結果、
(C)什器において推奨される棚割りの情報。
次に、図6を参照して、画像処理装置200の動作の流れについて説明する。図6は、本実施の形態に係る画像処理装置200の動作の流れの一例を示すフローチャートである。
本実施の形態に係る画像処理装置200によれば、認識漏れが発生している蓋然性の高い領域を、より高精度に検出することができる。なぜならば、検出部120が、撮影画像から認識部110が認識しなかった物品の領域を、什器情報に基づいて、検出するからである。
次に、本発明の第3の実施の形態について説明する。第3の実施の形態も、第2の実施の形態と同様に、上述した第1の実施の形態を基本とする。図7は、本実施の形態に係る画像処理装置300の機能構成の一例を示す機能ブロック図である。なお、説明の便宜上、前述した第1および第2の実施の形態で説明した図面に含まれる部材と同じ機能を有する部材については、同じ符号を付し、その説明を省略する。本実施の形態に係る画像処理装置300は、上述した第2の実施の形態に係る画像処理装置200に、更に、表示制御部と、修正部とを備える構成である。
(1)過去の陳列結果、
(2)什器において推奨される棚割りの情報、
(3)認識部110が既に認識した物品(受信した認識結果)。
次に、図12を参照して、画像処理装置300の動作の流れについて説明する。図12は、本実施の形態に係る画像処理装置300の動作の流れの一例を示すフローチャートである。
本実施の形態に係る画像処理装置300は、上述した第1および第2の実施の形態に係る画像処理装置と同様の効果を得ることができる。
ここで、上述した各実施の形態に係る画像処理装置(100、200、300)を実現可能なハードウェアの構成例について説明する。上述した画像処理装置(100、200、300)は、専用の装置として実現してもよいが、コンピュータ(情報処理装置)を用いて実現してもよい。
・CPU(Central Processing Unit)11、
・通信インタフェース(I/F)12、入出力ユーザインタフェース13、
・ROM(Read Only Memory)14、
・RAM(Random Access Memory)15、
・記憶装置17、及び
・コンピュータ読み取り可能な記憶媒体19のドライブ装置18。
また、これらはバス16を介して接続されている。入出力ユーザインタフェース13は、入力デバイスの一例であるキーボードや、出力デバイスとしてのディスプレイ等のマンマシンインタフェースである。通信インタフェース12は、上述した各実施の形態に係る装置(図1、図2および図8)が、外部装置と、通信ネットワーク20を介して通信するための一般的な通信手段である。係るハードウェア構成において、CPU11は、各実施の形態に係る画像処理装置(100、200、300)を実現する情報処理装置10について、全体の動作を司る。
・CD(Compact Disk)-ROM、フラッシュメモリ等のコンピュータ読み取り可能な各種の記録媒体(19)を介して当該装置内にインストールする方法、
・インターネット等の通信回線(20)を介して外部よりダウンロードする方法。
そして、このような場合において、本発明の各実施の形態は、係るコンピュータプログラムを構成するコード(プログラム群17A)或いは係るコードが格納された記憶媒体(19)によって構成されると捉えることができる。
110 認識部
120 検出部
200 画像処理装置
210 受信部
220 記憶部
300 画像処理装置
310 表示制御部(表示制御装置)
320 修正部
400 撮像装置
500 表示操作装置
510 表示部
520 入力部
31 物品
32 物品画像領域
Claims (12)
- 陳列された物品を撮影した撮影画像から、前記物品を認識する認識手段と、
前記物品が陳列された什器に関連する什器情報に基づき、前記撮影画像に含まれる物品であって、前記認識手段によって認識されなかった物品の領域を検出する検出手段と、を備えることを特徴とする画像処理装置。 - 前記什器情報は、(a)前記什器を撮影した撮影画像の撮影時刻より過去の時刻に該什器を撮影した撮影画像に対する前記認識手段による認識結果、(b)前記什器における推奨される棚割りの情報、(c)前記什器の段数または列数、(d)前記什器の各段または各列に配置可能な物品数、および、(e)前記什器に陳列された物品の陳列に関する条件を示す情報、の少なくとも何れかである、ことを特徴とする請求項1に記載の画像処理装置。
- 前記認識手段は、前記物品の認識の際に使用する認識パラメータを、(a)前記什器を撮影した撮影画像に対する認識結果、(b)前記撮影画像の撮影時刻より過去の時刻に該什器を撮影した撮影画像に対する前記認識手段による認識結果、(c)前記什器における推奨される棚割りの情報、の少なくとも何れかに基づいて変化させ、前記検出された領域に含まれる物品を認識する、ことを特徴とする、請求項1または2に記載の画像処理装置。
- 前記検出手段は、更に、認識手段によって認識された前記物品の物品間距離、認識された前記物品と前記什器の外縁との距離、認識された前記物品の大きさ、および、物品が認識されなかった領域に含まれる特徴点の数の少なくとも何れかに基づいて、前記認識手段によって認識されなかった物品の領域を検出する、ことを特徴とする請求項1から3の何れか1項に記載の画像処理装置。
- 前記検出手段は、更に、(a)前記什器を撮影した撮影画像の撮影時刻より過去の時刻に該什器を撮影した撮影画像に対する前記認識手段による認識結果、(b)前記什器における推奨される棚割りの情報、(c)前記什器に陳列された物品の陳列に関する条件を示す情報、(d)物品の発注情報、および、(e)物品の売上情報、の少なくとも何れかに基づいて、前記認識手段によって認識された物品のうち、誤認識の可能性が高い物品を判定し、誤認識の可能性が高いと判定した前記物品の前記撮影画像上における領域を検出する、ことを特徴とする請求項1から4の何れか1項に記載の画像処理装置。
- 前記検出手段によって、検出された領域を示す情報を画面に表示させる表示制御手段を更に備える、ことを特徴とする請求項1から5の何れか1項に記載の画像処理装置。
- 前記表示制御手段は、前記領域に存在する可能性がある物品の候補を選択可能に前記画面に表示させる、ことを特徴とする、請求項1から6の何れか1項に記載の画像処理装置。
- 前記画面に表示された物品の候補に対する選択結果に基づいて、前記認識結果を修正する修正手段、を更に備えることを特徴とする、請求項7に記載の画像処理装置。
- 前記表示制御手段は、前記領域に存在する可能性がある物品の候補を、(a)前記什器を撮影した撮影画像の撮影時刻より過去の時刻に該什器を撮影した撮影画像に対する前記認識手段による認識結果、(b)前記什器における推奨される棚割りの情報、(c)認識手段による認識結果、(d)前記什器に陳列された物品の陳列に関する条件を示す情報、(e)物品の発注情報、および、(f)物品の売上情報、の少なくとも何れかに基づいて決定する、ことを特徴とする請求項7または8に記載の画像処理装置。
- 陳列された物品を撮影した撮影画像のうち、前記物品が認識されなかった領域であって、前記物品が陳列されている可能性がある領域を、前記物品が認識されていない領域として画面に表示させる表示制御装置。
- 陳列された物品を撮影した撮影画像から、前記物品を認識し、
前記物品が陳列された什器に関連する什器情報に基づき、前記撮影画像に含まれる物品であって、認識されなかった物品の領域を検出する、ことを特徴とする画像処理方法。 - 陳列された物品を撮影した撮影画像から、前記物品を認識する処理と、
前記物品が陳列された什器に関連する什器情報に基づき、前記撮影画像に含まれる物品であって、認識されなかった物品の領域を検出する処理と、をコンピュータに実行させるプログラムを記憶する、コンピュータ読み取り可能な記録媒体。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/521,277 US10438079B2 (en) | 2014-10-23 | 2015-10-09 | Image processing apparatus, image processing method and recording medium |
CN201580057604.2A CN107077686B (zh) | 2014-10-23 | 2015-10-09 | 图像处理装置和图像处理方法 |
EP15852272.2A EP3211580B1 (en) | 2014-10-23 | 2015-10-09 | Image processing apparatus, display control apparatus, image processing method and recording medium |
JP2016555071A JP6620755B2 (ja) | 2014-10-23 | 2015-10-09 | 画像処理装置、表示制御装置、画像処理方法、および、プログラム |
US16/570,809 US11403843B2 (en) | 2014-10-23 | 2019-09-13 | Image processing apparatus, display control apparatus, image processing method and recording medium |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014216117 | 2014-10-23 | ||
JP2014-216117 | 2014-10-23 |
Related Child Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/521,277 A-371-Of-International US10438079B2 (en) | 2014-10-23 | 2015-10-09 | Image processing apparatus, image processing method and recording medium |
US16/570,809 Continuation US11403843B2 (en) | 2014-10-23 | 2019-09-13 | Image processing apparatus, display control apparatus, image processing method and recording medium |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016063484A1 true WO2016063484A1 (ja) | 2016-04-28 |
Family
ID=55760545
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/005151 WO2016063484A1 (ja) | 2014-10-23 | 2015-10-09 | 画像処理装置、表示制御装置、画像処理方法、および、記録媒体 |
Country Status (5)
Country | Link |
---|---|
US (2) | US10438079B2 (ja) |
EP (1) | EP3211580B1 (ja) |
JP (4) | JP6620755B2 (ja) |
CN (1) | CN107077686B (ja) |
WO (1) | WO2016063484A1 (ja) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018010652A1 (zh) * | 2016-07-12 | 2018-01-18 | 腾讯科技(深圳)有限公司 | 图像识别回调通知的方法、服务器和计算机可读存储介质 |
WO2019188443A1 (ja) * | 2018-03-30 | 2019-10-03 | 日本電気株式会社 | 情報処理装置、情報処理システム、制御方法、及びプログラム |
JP2020135822A (ja) * | 2019-02-26 | 2020-08-31 | 富士ゼロックス株式会社 | 情報処理装置及びプログラム |
JP2021015368A (ja) * | 2019-07-10 | 2021-02-12 | 富士通株式会社 | 物品推定プログラム、物品推定装置、及び物品推定方法 |
WO2021054266A1 (ja) * | 2019-09-17 | 2021-03-25 | 日本電気株式会社 | 画像処理装置、画像処理方法、及びプログラム |
US10973720B2 (en) | 2017-04-07 | 2021-04-13 | Hill-Rom Services, Inc. | Image based monitoring of the state of a device |
WO2021240902A1 (ja) * | 2020-05-27 | 2021-12-02 | パナソニックIpマネジメント株式会社 | 発注管理装置及び発注管理方法 |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10664971B2 (en) | 2017-12-19 | 2020-05-26 | Wipro Limited | Method and device for detecting region of interest based on images |
CN111339791B (zh) * | 2018-12-18 | 2021-05-25 | 北京京东乾石科技有限公司 | 商品识别方法和通道、控制装置、商品结算通道 |
JP7234741B2 (ja) * | 2019-03-28 | 2023-03-08 | 株式会社三洋物産 | 遊技機 |
JP7234740B2 (ja) * | 2019-03-28 | 2023-03-08 | 株式会社三洋物産 | 遊技機 |
WO2023141696A1 (en) * | 2022-01-31 | 2023-08-03 | Unefi Inc. | Data gathering for store fixtures and the products therein |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007021007A (ja) * | 2005-07-20 | 2007-02-01 | Casio Comput Co Ltd | 商品情報表示装置及びプログラム |
JP2009265998A (ja) * | 2008-04-25 | 2009-11-12 | Lambda Systems Inc | 画像処理システム、画像処理方法及び画像処理プログラム |
JP2012193873A (ja) * | 2011-03-15 | 2012-10-11 | Nikon Corp | 収納装置 |
WO2014136559A1 (ja) * | 2013-03-04 | 2014-09-12 | 日本電気株式会社 | 物品管理システム、情報処理装置およびその制御方法と制御プログラム |
JP2014170431A (ja) * | 2013-03-04 | 2014-09-18 | Nec Corp | 情報処理システム、情報処理装置およびその制御方法と制御プログラム |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7246078B2 (en) * | 2000-12-04 | 2007-07-17 | Ncr Corporation | System and methods for graphically representing purchase profiles and sales guidance to a customer service representative |
US8249885B2 (en) * | 2001-08-08 | 2012-08-21 | Gary Charles Berkowitz | Knowledge-based e-catalog procurement system and method |
US7309009B2 (en) * | 2005-06-21 | 2007-12-18 | International Business Machines Corporation | Retail store fly-around product locator |
US20080043013A1 (en) * | 2006-06-19 | 2008-02-21 | Kimberly-Clark Worldwide, Inc | System for designing shopping environments |
WO2008100537A2 (en) * | 2007-02-12 | 2008-08-21 | Sorensen Associates Inc. | Still image shopping event monitoring and analysis system and method |
US9135491B2 (en) * | 2007-08-31 | 2015-09-15 | Accenture Global Services Limited | Digital point-of-sale analyzer |
US8630924B2 (en) * | 2007-08-31 | 2014-01-14 | Accenture Global Services Limited | Detection of stock out conditions based on image processing |
CN102763123B (zh) * | 2009-12-02 | 2015-03-25 | 高通股份有限公司 | 通过使查询图像和模型图像中的检测到的关键点成群集而进行特征匹配 |
US9749823B2 (en) * | 2009-12-11 | 2017-08-29 | Mentis Services France | Providing city services using mobile devices and a sensor network |
US8509526B2 (en) | 2010-04-13 | 2013-08-13 | International Business Machines Corporation | Detection of objects in digital images |
JP2012126532A (ja) | 2010-12-16 | 2012-07-05 | Japan Tobacco Inc | 商品管理装置、商品管理方法およびプログラム |
FR2973540B1 (fr) | 2011-04-01 | 2013-03-29 | CVDM Solutions | Procede d'extraction automatisee d'un planogramme a partir d'images de lineaire |
US9785898B2 (en) | 2011-06-20 | 2017-10-10 | Hi-Tech Solutions Ltd. | System and method for identifying retail products and determining retail product arrangements |
CN103150540A (zh) * | 2011-12-06 | 2013-06-12 | 原熙 | 一种通过移动通讯设备的摄像装置获取物品信息的系统 |
JP6043094B2 (ja) | 2012-05-30 | 2016-12-14 | 辛東主 | 商品陳列情報集計システム |
CN102930264B (zh) * | 2012-09-29 | 2015-10-28 | 李炳华 | 基于图像识别技术的商品陈列信息采集分析系统及方法 |
US20140195302A1 (en) * | 2013-01-10 | 2014-07-10 | Oracle International Corporation | Guided walkthrough provider |
US9886678B2 (en) * | 2013-09-25 | 2018-02-06 | Sap Se | Graphic representations of planograms |
US10592854B2 (en) * | 2015-12-18 | 2020-03-17 | Ricoh Co., Ltd. | Planogram matching |
-
2015
- 2015-10-09 CN CN201580057604.2A patent/CN107077686B/zh active Active
- 2015-10-09 WO PCT/JP2015/005151 patent/WO2016063484A1/ja active Application Filing
- 2015-10-09 JP JP2016555071A patent/JP6620755B2/ja active Active
- 2015-10-09 EP EP15852272.2A patent/EP3211580B1/en active Active
- 2015-10-09 US US15/521,277 patent/US10438079B2/en active Active
-
2019
- 2019-09-13 US US16/570,809 patent/US11403843B2/en active Active
- 2019-11-20 JP JP2019209693A patent/JP6962356B2/ja active Active
-
2021
- 2021-08-24 JP JP2021136373A patent/JP7279750B2/ja active Active
-
2023
- 2023-04-28 JP JP2023074071A patent/JP2023109773A/ja active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007021007A (ja) * | 2005-07-20 | 2007-02-01 | Casio Comput Co Ltd | 商品情報表示装置及びプログラム |
JP2009265998A (ja) * | 2008-04-25 | 2009-11-12 | Lambda Systems Inc | 画像処理システム、画像処理方法及び画像処理プログラム |
JP2012193873A (ja) * | 2011-03-15 | 2012-10-11 | Nikon Corp | 収納装置 |
WO2014136559A1 (ja) * | 2013-03-04 | 2014-09-12 | 日本電気株式会社 | 物品管理システム、情報処理装置およびその制御方法と制御プログラム |
JP2014170431A (ja) * | 2013-03-04 | 2014-09-18 | Nec Corp | 情報処理システム、情報処理装置およびその制御方法と制御プログラム |
Non-Patent Citations (1)
Title |
---|
See also references of EP3211580A4 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018010652A1 (zh) * | 2016-07-12 | 2018-01-18 | 腾讯科技(深圳)有限公司 | 图像识别回调通知的方法、服务器和计算机可读存储介质 |
US10973720B2 (en) | 2017-04-07 | 2021-04-13 | Hill-Rom Services, Inc. | Image based monitoring of the state of a device |
WO2019188443A1 (ja) * | 2018-03-30 | 2019-10-03 | 日本電気株式会社 | 情報処理装置、情報処理システム、制御方法、及びプログラム |
JP2019179488A (ja) * | 2018-03-30 | 2019-10-17 | 日本電気株式会社 | 情報処理装置、情報処理システム、制御方法、及びプログラム |
JP2020135822A (ja) * | 2019-02-26 | 2020-08-31 | 富士ゼロックス株式会社 | 情報処理装置及びプログラム |
JP7302200B2 (ja) | 2019-02-26 | 2023-07-04 | 富士フイルムビジネスイノベーション株式会社 | 情報処理装置及びプログラム |
JP2021015368A (ja) * | 2019-07-10 | 2021-02-12 | 富士通株式会社 | 物品推定プログラム、物品推定装置、及び物品推定方法 |
JP7346951B2 (ja) | 2019-07-10 | 2023-09-20 | 富士通株式会社 | 物品推定プログラム、物品推定装置、及び物品推定方法 |
WO2021054266A1 (ja) * | 2019-09-17 | 2021-03-25 | 日本電気株式会社 | 画像処理装置、画像処理方法、及びプログラム |
WO2021240902A1 (ja) * | 2020-05-27 | 2021-12-02 | パナソニックIpマネジメント株式会社 | 発注管理装置及び発注管理方法 |
Also Published As
Publication number | Publication date |
---|---|
US11403843B2 (en) | 2022-08-02 |
US10438079B2 (en) | 2019-10-08 |
JP2023109773A (ja) | 2023-08-08 |
EP3211580B1 (en) | 2021-03-03 |
EP3211580A4 (en) | 2018-04-18 |
JP6620755B2 (ja) | 2019-12-18 |
JP2020061158A (ja) | 2020-04-16 |
JP6962356B2 (ja) | 2021-11-05 |
JP7279750B2 (ja) | 2023-05-23 |
JPWO2016063484A1 (ja) | 2017-09-07 |
JP2022002100A (ja) | 2022-01-06 |
CN107077686A (zh) | 2017-08-18 |
US20170308764A1 (en) | 2017-10-26 |
EP3211580A1 (en) | 2017-08-30 |
US20200082193A1 (en) | 2020-03-12 |
CN107077686B (zh) | 2020-11-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6620755B2 (ja) | 画像処理装置、表示制御装置、画像処理方法、および、プログラム | |
US11900316B2 (en) | Information processing apparatus, control method, and program | |
US11210513B2 (en) | Detection method and detection device | |
US10311295B2 (en) | Heuristic finger detection method based on depth image | |
US10318152B2 (en) | Modifying key size on a touch screen based on fingertip location | |
JP5670471B2 (ja) | 動画像認識装置及び動画像認識方法 | |
US9256792B2 (en) | Image processing apparatus, image processing method, and program | |
US20170309113A1 (en) | Vending machine recognition apparatus, vending machine recognition method, and recording medium | |
JP2019012546A (ja) | 情報処理装置、画像処理方法とプログラム、ならびに、pos端末装置 | |
US20150253861A1 (en) | Detecting device and detecting method | |
US8860695B2 (en) | Optical touch system and electronic apparatus including the same | |
JP5876129B2 (ja) | 動画像認識装置及び動画像認識方法 | |
JP6623565B2 (ja) | 棚割情報生成装置、棚割情報生成システム、棚割情報生成方法、撮像装置、およびプログラム | |
JPWO2018168515A1 (ja) | 画像処理装置、画像処理方法及び記録媒体 | |
CN104281381B (zh) | 控制配备有触摸屏的用户界面的装置和方法 | |
US9075465B2 (en) | Method of identifying touch event on touch panel by shape of signal group and computer readable medium thereof | |
US20160139682A1 (en) | Displaying corrected logogram input | |
JP2018200629A (ja) | 画像処理装置、画像処理方法、及び画像処理プログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15852272 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016555071 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15521277 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
REEP | Request for entry into the european phase |
Ref document number: 2015852272 Country of ref document: EP |