WO2019043879A1 - Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme - Google Patents

Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme Download PDF

Info

Publication number
WO2019043879A1
WO2019043879A1 PCT/JP2017/031381 JP2017031381W WO2019043879A1 WO 2019043879 A1 WO2019043879 A1 WO 2019043879A1 JP 2017031381 W JP2017031381 W JP 2017031381W WO 2019043879 A1 WO2019043879 A1 WO 2019043879A1
Authority
WO
WIPO (PCT)
Prior art keywords
product
image
camera
distance
analysis
Prior art date
Application number
PCT/JP2017/031381
Other languages
English (en)
Japanese (ja)
Inventor
俊二 菅谷
Original Assignee
株式会社オプティム
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 株式会社オプティム filed Critical 株式会社オプティム
Priority to PCT/JP2017/031381 priority Critical patent/WO2019043879A1/fr
Publication of WO2019043879A1 publication Critical patent/WO2019043879A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes

Definitions

  • the present invention relates to an image analysis distance information providing system, an image analysis distance information providing method, and a program for analyzing a product image taken by a camera and estimating and providing a distance between the product and the camera.
  • Patent Document 1 an image processing apparatus for processing a thermal image obtained from a two-dimensional infrared sensor and detecting highly accurate information of a person or the like in a detection area.
  • the present invention analyzes an image of a product taken by a camera, compares the size of the specified product with the number of pixels, and estimates and provides the distance between the product and the camera. It is an object of the present invention to provide an image analysis distance information providing system, an image analysis distance information providing method, and a program.
  • the present invention provides the following solutions.
  • the invention specifies the type of the product based on an image acquisition unit that acquires an image of a product taken by a camera, an image analysis unit that analyzes the image, and a result of the analysis.
  • the web content is referred to based on the type specifying means, the pixel number specifying means for specifying the number of pixels of the product shown in the image based on the analysis result, and the type of the specified product.
  • Distance estimation means for estimating the distance between the product and the camera by comparing the size acquisition means for acquiring the size of the product, the number of pixels of the product shown in the image, and the size of the product And a providing means for providing information on the estimated distance.
  • the invention specifies the type of the product based on an image acquisition step of acquiring an image of a product taken by a camera, an image analysis step of analyzing the image, and a result of the analysis.
  • the pixel number identification step for identifying the number of pixels of the product appearing in the image based on the result of the analysis, and the web content based on the type of the identified product
  • a distance obtaining step for obtaining a size of a product
  • a distance estimating step for estimating a distance between the product and the camera by comparing the number of pixels of the product shown in the image with the size of the product And providing the information of the estimated distance.
  • the invention according to the first aspect is characterized in that, on a computer, an image acquisition step of acquiring an image of a product taken by a camera, an image analysis step of analyzing the image, and a type of the product based on a result of the analysis.
  • the web content is referred to based on the type identification step of identifying the number of pixels of the product shown in the image based on the analysis result and the type of the identified product.
  • Estimate the distance between the product and the camera by comparing the size of the product with the size of the product in the image acquisition step of acquiring the size of the product, the number of pixels of the product in the image, and the size of the product
  • a program is provided for causing a distance estimation step and a provision step of providing information on the estimated distance.
  • the distance between the product and the camera can be estimated and provided.
  • FIG. 1 is a schematic view of an image analysis distance information providing system.
  • FIG. 2 is an example of providing estimated distance information.
  • the image analysis distance information providing system analyzes an image of a product taken by a camera, compares the size of the specified product with the number of pixels, and estimates the distance between the product and the camera.
  • FIG. 1 is a schematic view of an image analysis distance information providing system according to a preferred embodiment of the present invention.
  • the image analysis distance information providing system is realized by the control unit reading a predetermined program, an image acquisition unit, an image analysis unit, a type specification unit, a pixel number specification unit, a size acquisition unit, A distance estimation unit and a provision unit are provided.
  • camera information acquisition means, product position estimation means, detection means may be provided. These may be application based, cloud based or otherwise.
  • Each means described above may be realized by a single computer or may be realized by two or more computers (for example, in the case of a server and a terminal).
  • An image acquisition means acquires the image of the product imaged with the camera.
  • the image may be a moving image or a still image.
  • the camera may be a digital camera, a camera of a smartphone, or a camera provided in a drone. Real-time images are preferable for providing distance information in real time.
  • the image analysis means analyzes the image.
  • Machine learning may improve the accuracy of image analysis. For example, machine learning is performed using past images as teacher data.
  • the type identification means identifies the type of product based on the analysis result. As dimensions such as length, width and height differ depending on the type of product, it is necessary to specify the type of product.
  • the product type is, for example, a monitor of FlexScanEV3237-GY of EIZO, a Tokyo Tower, a chair of Lagoon Hard DBR of Nitori, etc.
  • the pixel number specifying means specifies the number of pixels of the product shown in the image based on the analysis result. For example, it is specified that the number of pixels of a product appearing in an image is 120 ⁇ 75 pixels.
  • the angle of view and magnification of the camera may be used to specify the number of pixels of the product appearing in the image. If the angle of view and the magnification are used, the accuracy in specifying the number of pixels is enhanced.
  • the mounting position of the camera, the mounting angle, and the resolution may be used to specify the number of product pixels in the image. The mounting position, mounting angle, and resolution further increase the accuracy of the pixel count.
  • the size obtaining means obtains the size of the product by referring to the Web content based on the type of the specified product. For example, if the specified product type is a monitor of FlexScanEV3237-GY of EIZO, the FlexScanEV3237-GY size is acquired with reference to the EIZO homepage. On the manufacturer's website, dimensions such as length, width, and height are described. In addition, even if it is not the manufacturer's website, even if it refers to the Web content in which dimensions such as length, width, height, etc. are described, such as SNS or second-hand product sales site, the size of the product is acquired Good.
  • the distance estimating means compares the number of pixels of the product shown in the image with the size of the product to estimate the distance between the product and the camera. For example, it is assumed that the number of pixels of the product shown in the image is 120 ⁇ 75 pixels, and the size of the product is 120 cm ⁇ 75 cm. Here, if it is known that a product of 120 centimeters is imaged at 120 pixels when it is 10 meters away from the camera, it can be inferred that the distance between the product and the camera is 10 meters . By using the ratio, it is possible to estimate the distance between the product and the camera, even with other values.
  • the providing means provides the information of the estimated distance. For example, as shown in FIG. 2, information on the distance estimated by being superimposed on the image may be provided. It can be seen that the distance from the camera to the front monitor is 3 mails, and the distance from the camera to the back tower is 220 meters. In addition, only information of the estimated distance without displaying in the image may be provided in association with the type of product. For example, 5 meters to the EIZO FlexScan EV 3237-GY monitor.
  • the camera information acquisition means acquires the camera position and the camera orientation of the camera.
  • a camera position can be acquired from GPS information etc. of a smart phone.
  • the camera position can be obtained from the GPS information or the like provided in the drone.
  • the camera orientation can be acquired from an orientation sensor such as a geomagnetic sensor provided in a smartphone or a drone.
  • the product position estimation means estimates the product position of the product from the camera position and the distance estimated as the camera orientation. Knowing the camera position, the camera orientation, and the distance between the camera and the product, the product position can be inferred based on the camera position.
  • the providing means may provide information of the estimated distance and the estimated product position. Providing information on the distance from the camera to the product and the position of the product increases the value and utilization of the information.
  • the image acquisition unit may acquire a plurality of time-series images. It can be made in time series by the time when the image was acquired. In addition, since time information is often attached as meta information to an image, it can be time-series.
  • the detection means detects the traveling direction and the traveling speed of the product from the plurality of time-series images acquired. If there are a plurality of time-series images, it is possible to detect the traveling direction and traveling speed of the product from the difference.
  • the providing means may provide the information of the estimated distance, the traveling direction of the product, and the traveling speed of the product.
  • the image analysis distance information providing method analyzes an image of a product taken by a camera, compares the size of the specified product with the number of pixels, and estimates the distance between the product and the camera. Is a method to provide.
  • the image analysis distance information providing method includes an image acquisition step, an image analysis step, a type identification step, a pixel number identification step, a size acquisition step, a distance estimation step, and a provision step.
  • a camera information acquisition step, a product position estimation step, and a detection step may be similarly provided.
  • the image acquisition step acquires an image of a product captured by a camera.
  • the image may be a moving image or a still image.
  • the camera may be a digital camera, a camera of a smartphone, or a camera provided in a drone. Real-time images are preferable for providing distance information in real time.
  • the image analysis step analyzes the image.
  • Machine learning may improve the accuracy of image analysis. For example, machine learning is performed using past images as teacher data.
  • the type identification step identifies the type of product based on the analysis result. As dimensions such as length, width and height differ depending on the type of product, it is necessary to specify the type of product.
  • the product type is, for example, a monitor of FlexScanEV3237-GY of EIZO, a Tokyo Tower, a chair of Lagoon Hard DBR of Nitori, etc.
  • the pixel number specifying step specifies the pixel number of the product appearing in the image based on the analysis result. For example, it is specified that the number of pixels of a product appearing in an image is 120 ⁇ 75 pixels. Also, the angle of view and magnification of the camera may be used to specify the number of pixels of the product appearing in the image. If the angle of view and the magnification are used, the accuracy in specifying the number of pixels is enhanced. In addition, the mounting position of the camera, the mounting angle, and the resolution may be used to specify the number of product pixels in the image. The mounting position, mounting angle, and resolution further increase the accuracy of the pixel count.
  • the size obtaining step refers to Web content and obtains the size of the product based on the type of the specified product. For example, if the specified product type is a monitor of FlexScanEV3237-GY of EIZO, the FlexScanEV3237-GY size is acquired with reference to the EIZO homepage. On the manufacturer's website, dimensions such as length, width, and height are described. In addition, even if it is not the manufacturer's website, even if it refers to the Web content in which dimensions such as length, width, height, etc. are described, such as SNS or second-hand product sales site, the size of the product is acquired Good.
  • the distance estimation step compares the number of pixels of the product shown in the image with the size of the product to infer the distance between the product and the camera. For example, it is assumed that the number of pixels of the product shown in the image is 120 ⁇ 75 pixels, and the size of the product is 120 cm ⁇ 75 cm. Here, if it is known that a product of 120 centimeters is imaged at 120 pixels when it is 10 meters away from the camera, it can be inferred that the distance between the product and the camera is 10 meters . By using the ratio, it is possible to estimate the distance between the product and the camera, even with other values.
  • the providing step provides the information of the estimated distance. For example, as shown in FIG. 2, information on the distance estimated by being superimposed on the image may be provided. It can be seen that the distance from the camera to the front monitor is 3 mails, and the distance from the camera to the back tower is 220 meters. In addition, only information of the estimated distance without displaying in the image may be provided in association with the type of product. For example, 5 meters to the EIZO FlexScan EV 3237-GY monitor.
  • the camera information acquisition step acquires the camera position and the camera orientation of the camera.
  • a camera position can be acquired from GPS information etc. of a smart phone.
  • the camera position can be obtained from the GPS information or the like provided in the drone.
  • the camera orientation can be acquired from an orientation sensor such as a geomagnetic sensor provided in a smartphone or a drone.
  • the product position estimation step infers the product position of the product from the camera position and the distance estimated by the camera orientation. Knowing the camera position, the camera orientation, and the distance between the camera and the product, the product position can be inferred based on the camera position.
  • the providing step may provide information of the estimated distance and the estimated product position. Providing information on the distance from the camera to the product and the position of the product increases the value and utilization of the information.
  • the image acquisition step may acquire a plurality of time-series images. It can be made in time series by the time when the image was acquired. In addition, since time information is often attached as meta information to an image, it can be time-series.
  • the detection step detects the traveling direction and the traveling speed of the product from the plurality of time-series images acquired. If there are a plurality of time-series images, it is possible to detect the traveling direction and traveling speed of the product from the difference.
  • the providing step may also provide information on the estimated distance, the direction of travel of the product, and the speed of travel of the product. By providing information on the estimated distance, the direction of product travel, and the speed of product travel, the value and utilization of the information increases.
  • the above-described means and functions are realized by a computer (including a CPU, an information processing device, and various terminals) reading and executing a predetermined program.
  • the program may be, for example, an application installed on a computer, or a SaaS (software as a service) provided from a computer via a network, for example, a flexible disk, a CD It may be provided in the form of being recorded in a computer readable recording medium such as a CD-ROM or the like, a DVD (DVD-ROM, DVD-RAM or the like).
  • the computer reads the program from the recording medium, transfers the program to the internal storage device or the external storage device, stores it, and executes it.
  • the program may be recorded in advance in a storage device (recording medium) such as, for example, a magnetic disk, an optical disk, or a magneto-optical disk, and may be provided from the storage device to the computer via a communication line.
  • nearest neighbor method naive Bayes method
  • decision tree naive Bayes method
  • support vector machine e.g., support vector machine
  • reinforcement learning e.g., reinforcement learning, etc.
  • deep learning may be used in which feature quantities for learning are generated by using a neural network.

Abstract

[Problème] Analyser une image d'un produit capturée par une caméra; comparer la taille d'un produit spécifié et le nombre de pixels; et estimer et indiquer la distance entre le produit et la caméra. [Solution] L'invention concerne un système de fourniture d'informations de distance par analyse d'image, qui comprend : un moyen d'acquisition d'image qui obtient une image d'un produit capturée par une caméra; un moyen d'analyse d'image qui analyse l'image; un moyen de spécification de type qui spécifie le type de produit sur la base des résultats de l'analyse; un moyen de spécification de nombre de pixels qui spécifie le nombre de pixels pour le produit apparaissant dans l'image, sur la base des résultats de l'analyse; un moyen d'acquisition de taille qui explore un contenu web et obtient la taille du produit sur la base du type de produit identifié; un moyen d'estimation de distance qui compare la taille du produit et le nombre de pixels pour le produit apparaissant dans l'image, et estime la distance entre le produit et la caméra; et un moyen de fourniture qui fournit des informations sur la distance estimée.
PCT/JP2017/031381 2017-08-31 2017-08-31 Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme WO2019043879A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2017/031381 WO2019043879A1 (fr) 2017-08-31 2017-08-31 Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2017/031381 WO2019043879A1 (fr) 2017-08-31 2017-08-31 Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme

Publications (1)

Publication Number Publication Date
WO2019043879A1 true WO2019043879A1 (fr) 2019-03-07

Family

ID=65525195

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2017/031381 WO2019043879A1 (fr) 2017-08-31 2017-08-31 Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme

Country Status (1)

Country Link
WO (1) WO2019043879A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06255423A (ja) * 1993-03-04 1994-09-13 Sharp Corp 車載用監視カメラ装置
JP2002366937A (ja) * 2001-06-08 2002-12-20 Fuji Heavy Ind Ltd 車外監視装置
JP2009031870A (ja) * 2007-07-24 2009-02-12 Seiko Epson Corp 被写体距離推定のための画像処理
JP2014167676A (ja) * 2013-02-28 2014-09-11 Fujifilm Corp 車間距離算出装置およびその動作制御方法
JP2016014549A (ja) * 2014-07-01 2016-01-28 Ｉｈｉ運搬機械株式会社 車両位置算出装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06255423A (ja) * 1993-03-04 1994-09-13 Sharp Corp 車載用監視カメラ装置
JP2002366937A (ja) * 2001-06-08 2002-12-20 Fuji Heavy Ind Ltd 車外監視装置
JP2009031870A (ja) * 2007-07-24 2009-02-12 Seiko Epson Corp 被写体距離推定のための画像処理
JP2014167676A (ja) * 2013-02-28 2014-09-11 Fujifilm Corp 車間距離算出装置およびその動作制御方法
JP2016014549A (ja) * 2014-07-01 2016-01-28 Ｉｈｉ運搬機械株式会社 車両位置算出装置

Similar Documents

Publication Publication Date Title
US10643073B2 (en) System, method, program for display on wearable terminal
US9305217B2 (en) Object tracking system using robot and object tracking method using a robot
WO2014041912A1 (fr) Système de traitement d'image, procédé de traitement d'image et programme
JP2011248548A (ja) コンテンツ決定プログラムおよびコンテンツ決定装置
US9729792B2 (en) Dynamic image selection
CN104919794A (zh) 用于从主从式相机跟踪系统提取元数据的方法和系统
JP6615800B2 (ja) 情報処理装置、情報処理方法およびプログラム
US9064178B2 (en) Edge detection apparatus, program and method for edge detection
US20190370543A1 (en) Land use determination system, land use determination method and program
US11669977B2 (en) Processing images to localize novel objects
EP2798576A2 (fr) Procédé et système pour la composition d'une vidéo
KR102043366B1 (ko) 최대 공통 뷰를 이용한 공간참조 영상들간의 궤적 유사도 측정방법
JP6638723B2 (ja) 画像解析装置、画像解析方法、及び、画像解析プログラム
US20180114339A1 (en) Information processing device and method, and program
JP2017027197A (ja) 監視プログラム、監視装置、及び監視方法
KR102226372B1 (ko) 다수 카메라와 라이다 센서 융합을 통한 객체 추적 시스템 및 방법
US20180314893A1 (en) Information processing device, video image monitoring system, information processing method, and recording medium
JPWO2014027500A1 (ja) 特徴抽出方法、プログラム及びシステム
WO2019043879A1 (fr) Système de fourniture d'informations de distance par analyse d'image, procédé de fourniture d'informations de distance par analyse d'image, et programme
JP2014048966A (ja) 物体検出システムおよびプログラム
JP2020144758A (ja) 移動物体検出装置、移動物体検出方法およびコンピュータプログラム
WO2019043877A1 (fr) Système de fourniture d'informations de distance d'analyse d'image, procédé de fourniture d'informations de distance d'analyse d'image et programme
JP2015082295A (ja) 人物検出装置及びプログラム
KR101480955B1 (ko) 동영상을 이용한 부유 구조물의 변위 측정 시스템 및 방법
JP6780639B2 (ja) 画像解析装置、画像解析方法、及び、画像解析プログラム

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: 17923280

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17923280

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP