[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

JP2007041762A5 - - Google Patents

Download PDF

Info

Publication number
JP2007041762A5
JP2007041762A5 JP2005223924A JP2005223924A JP2007041762A5 JP 2007041762 A5 JP2007041762 A5 JP 2007041762A5 JP 2005223924 A JP2005223924 A JP 2005223924A JP 2005223924 A JP2005223924 A JP 2005223924A JP 2007041762 A5 JP2007041762 A5 JP 2007041762A5
Authority
JP
Japan
Prior art keywords
image
target image
position information
target
keyword
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
JP2005223924A
Other languages
Japanese (ja)
Other versions
JP2007041762A (en
Filing date
Publication date
Application filed filed Critical
Priority to JP2005223924A priority Critical patent/JP2007041762A/en
Priority claimed from JP2005223924A external-priority patent/JP2007041762A/en
Publication of JP2007041762A publication Critical patent/JP2007041762A/en
Publication of JP2007041762A5 publication Critical patent/JP2007041762A5/ja
Withdrawn legal-status Critical Current

Links

Claims (3)

撮影位置の情報を備えた一の画像の画像内容を表わすキーワードを抽出するキーワード抽出装置であって、
前記複数の対象画像毎に、前記対象画像の画像特徴量と、前記対象画像の被写体を表わすキーワードと、前記対象画像の被写体を撮影可能な地理的な範囲を表す領域レンジを特定する情報と、前記対象画像の被写体が前記地理的な範囲に固有のものである程度を表す位置情報関連度と、を関連付けて記憶する検索データベースと、
前記一の画像を取得画像として入力する画像入力手段と、
前記取得画像の画像特徴量を抽出する画像特徴量抽出手段と、
前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像の位置情報が、当該対象画像の前記領域レンジ内でない場合には、当該対象画像をキーワードの抽出の対象から除外する対象除外手段と、
前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像データの位置情報が、当該対象画像の前記領域レンジ内である場合には、前記取得画像画像特徴量および前記対象画像の画像特徴量を用いて、前記取得画像前記対象画像の類似の程度を表す信頼性係数を算出する信頼性係数算出手段と、
前記位置情報関連度を用いて、前記算出された信頼性係数の補正を行う信頼性係数補正手段と、
前記補正された信頼性係数が最も高い前記対象画像のキーワードを、前記取得画像のキーワードとして前記データベースから抽出するキーワード抽出手段と、
を備えた画像認識装置。
A keyword extracting device for extracting a keyword representing the image content of one image having information on a photographing position,
For each of the plurality of target images, information specifying an image feature amount of the target image, a keyword representing the subject of the target image, and an area range representing a geographical range in which the subject of the target image can be photographed; A search database that associates and stores position information relevance representing the degree to which the subject of the target image is unique to the geographical range;
Image input means for inputting the one image as an acquired image;
Image feature amount extraction means for extracting an image feature amount of the acquired image;
If the position information relevance of the target image is unique to the geographical range, and the position information of the acquired image is not within the region range of the target image , the target image Exclusion means for excluding keywords from keyword extraction,
When the position information relevance of the target image is specific to the geographical range, and the position information of the acquired image data is within the region range of the target image, using the image feature amount of image feature amount and the object image of the acquired image, and reliability coefficient calculating means for calculating a reliability factor indicating the degree of similarity of the target image and the acquired image,
Reliability coefficient correction means for correcting the calculated reliability coefficient using the position information relevance;
Keyword extracting means for extracting the keyword of the target image having the highest corrected reliability coefficient from the database as a keyword of the acquired image ;
An image recognition apparatus comprising:
撮影位置の情報を備えた一の画像の画像内容を表わすキーワードを抽出するキーワード抽出方法であって、A keyword extraction method for extracting a keyword representing the image content of one image having shooting position information,
(a)前記複数の対象画像毎に、前記対象画像の画像特徴量と、前記対象画像の被写体を表わすキーワードと、前記対象画像の被写体を撮影可能な地理的な範囲を表す領域レンジを特定する情報と、前記対象画像の被写体が前記地理的な範囲に固有のものである程度を表す位置情報関連度と、を関連付けて記憶する検索データベースを用意する工程と、(A) For each of the plurality of target images, an image feature amount of the target image, a keyword representing the subject of the target image, and an area range representing a geographical range in which the subject of the target image can be photographed are specified. Providing a search database that stores information and a degree of relevance of positional information that represents a certain degree that the subject of the target image is unique to the geographical range; and
(b)コンピュータが前記一の画像を取得画像として入力する工程と、(B) a step in which the computer inputs the one image as an acquired image;
(c)コンピュータが前記取得画像の画像特徴量を抽出する工程と、(C) a step in which a computer extracts an image feature amount of the acquired image;
(d)コンピュータが前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像の位置情報が、当該対象画像の前記領域レンジ内でない場合には、当該対象画像をキーワードの抽出の対象から除外する工程と、(D) When the computer indicates that the position information relevance of the target image is unique to the geographical range, and the position information of the acquired image is not within the region range of the target image Excludes the target image from keyword extraction, and
(e)コンピュータが前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像データの位置情報が、当該対象画像の前記領域レンジ内である場合には、前記取得画像の画像特徴量および前記対象画像の画像特徴量を用いて、前記取得画像と前記対象画像の類似の程度を表す信頼性係数を算出する工程と、(E) the computer represents that the position information relevance of the target image is unique to the geographical range, and the position information of the acquired image data is within the region range of the target image In this case, using the image feature amount of the acquired image and the image feature amount of the target image, calculating a reliability coefficient representing the degree of similarity between the acquired image and the target image;
(f)コンピュータが前記位置情報関連度を用いて、前記算出された信頼性係数の補正を行う工程と、(F) a step in which the computer corrects the calculated reliability coefficient using the position information relevance;
(g)コンピュータが前記補正された信頼性係数が最も高い前記対象画像のキーワードを、前記取得画像のキーワードとして前記データベースから抽出する工程と、(G) extracting a keyword of the target image having the highest corrected reliability coefficient from the database as a keyword of the acquired image;
を備えた画像認識装置。  An image recognition apparatus comprising:
撮影位置の情報を備えた一の画像の画像内容を表わすキーワードの抽出をコンピュータに実行させるためのコンピュータプログラムであって、A computer program for causing a computer to extract a keyword representing the image content of one image having information on a photographing position,
前記複数の対象画像毎に、前記対象画像の画像特徴量と、前記対象画像の被写体を表わすキーワードと、前記対象画像の被写体を撮影可能な地理的な範囲を表す領域レンジを特定する情報と、前記対象画像の被写体が前記地理的な範囲に固有のものである程度を表す位置情報関連度と、を関連付けて記憶する検索データベースを用意する第1の機能と、  For each of the plurality of target images, information specifying an image feature amount of the target image, a keyword representing the subject of the target image, and an area range representing a geographical range in which the subject of the target image can be photographed; A first function that prepares a search database that stores in association with a position information relevance level that represents a certain degree that the subject of the target image is unique to the geographical range;
(b)前記一の画像を取得画像として入力する第2の機能と、(B) a second function of inputting the one image as an acquired image;
(c)前記取得画像の画像特徴量を抽出する第3の機能と、(C) a third function of extracting an image feature amount of the acquired image;
(d)前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像の位置情報が、当該対象画像の前記領域レンジ内でない場合には、当該対象画像をキーワードの抽出の対象から除外する第4の機能と、(D) When the position information relevance of the target image is unique to the geographical range, and the position information of the acquired image is not within the region range of the target image, A fourth function for excluding the target image from keyword extraction;
(e)前記対象画像の位置情報関連度が前記地理的な範囲に固有なものであることを表し、かつ、前記取得画像データの位置情報が、当該対象画像の前記領域レンジ内である場合には、前記取得画像の画像特徴量および前記対象画像の画像特徴量を用いて、前記取得画像と前記対象画像の類似の程度を表す信頼性係数を算出する第5の機能と、(E) When the position information relevance of the target image is unique to the geographical range, and the position information of the acquired image data is within the region range of the target image A fifth function for calculating a reliability coefficient representing a degree of similarity between the acquired image and the target image using the image feature amount of the acquired image and the image feature amount of the target image;
(f)前記位置情報関連度を用いて、前記算出された信頼性係数の補正を行う第6の機能と、(F) a sixth function for correcting the calculated reliability coefficient using the position information relevance;
(g)前記補正された信頼性係数が最も高い前記対象画像のキーワードを、前記取得画像のキーワードとして前記データベースから抽出する第7の機能と、(G) a seventh function of extracting the keyword of the target image having the highest corrected reliability coefficient from the database as a keyword of the acquired image;
を前記コンピュータに実現させる画像認識装置。  An image recognition apparatus for causing the computer to realize the above.
JP2005223924A 2005-08-02 2005-08-02 Image recognition device Withdrawn JP2007041762A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2005223924A JP2007041762A (en) 2005-08-02 2005-08-02 Image recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2005223924A JP2007041762A (en) 2005-08-02 2005-08-02 Image recognition device

Publications (2)

Publication Number Publication Date
JP2007041762A JP2007041762A (en) 2007-02-15
JP2007041762A5 true JP2007041762A5 (en) 2008-09-11

Family

ID=37799695

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2005223924A Withdrawn JP2007041762A (en) 2005-08-02 2005-08-02 Image recognition device

Country Status (1)

Country Link
JP (1) JP2007041762A (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5544250B2 (en) * 2010-09-02 2014-07-09 エヌ・ティ・ティ・コムウェア株式会社 Display image search method
JP6099146B2 (en) * 2013-08-05 2017-03-22 Kddi株式会社 Image identification apparatus and program
JP6963352B2 (en) * 2018-09-27 2021-11-05 株式会社オプティム Object condition judgment system, object condition judgment method, and program
CN111026896B (en) * 2019-11-15 2023-09-01 浙江大华技术股份有限公司 Feature value storage and processing method, device and storage device

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000163436A (en) * 1998-11-27 2000-06-16 Toshiba Corp System and method for information retrieval
JP2002297648A (en) * 2001-03-29 2002-10-11 Minolta Co Ltd Device and program for information retrieval, and recording medium
JP2003036267A (en) * 2001-07-23 2003-02-07 Canon Inc Information retrieval device, and information retrieval method
JP2005100274A (en) * 2003-09-26 2005-04-14 Mazda Motor Corp Information providing system, information retrieval device and information providing method

Similar Documents

Publication Publication Date Title
US12019675B2 (en) Recognizing text in image data
US9076069B2 (en) Registering metadata apparatus
Schönberger et al. A vote-and-verify strategy for fast spatial verification in image retrieval
US8971641B2 (en) Spatial image index and associated updating functionality
JP4337064B2 (en) Information processing apparatus, information processing method, and program
WO2016199662A1 (en) Image information processing system
EP4057163B1 (en) Facilitating use of images as search queries
US9576218B2 (en) Selecting features from image data
US20160085774A1 (en) Context based image search
JP5544250B2 (en) Display image search method
JP6211407B2 (en) Image search system, image search device, search server device, image search method, and image search program
US8660302B2 (en) Apparatus and method for tracking target
CN111160288A (en) Gesture key point detection method and device, computer equipment and storage medium
JP2017531854A5 (en)
TWI528197B (en) Photo grouping system, photo grouping method, and computer-readable storage medium
WO2012063544A1 (en) Image processing device, image processing method, and recording medium
WO2016058520A1 (en) Method and apparatus for recognizing name of face picture
CN106682187B (en) Method and device for establishing image base
CN109086690B (en) Image feature extraction method, target identification method and corresponding device
CN108665459A (en) A kind of image fuzzy detection method, computing device and readable storage medium storing program for executing
Tsai et al. No more ambiguity in 360deg room layout via bi-layout estimation
JP2007041762A5 (en)
US20180189602A1 (en) Method of and system for determining and selecting media representing event diversity
JP2012226429A (en) Image retrieval and recognition system
JP5188290B2 (en) Annotation apparatus, annotation method and program