Image search apparatus and image search method
AssigneeCanon Kabushiki Kaisha
InventorHirotaka Shiiyama
Life of this patent
Estimating the term…
Abstract
Of the feature amounts of respective regions on an image, a region having a feature amount which is similar to that of a region of interest on a image is specified (S1307). Of the OCR results of the respective regions on the image, a region having an OCR result which matches that of the region of interest on the image is specified (S1308). Regions corresponding to respective regions on the image of the respective regions are specified based on the layout order of the specified regions (S1311). Of the regions obtained by combining neighboring regions on the image, a region corresponding to an inclusion region that includes each specified region is specified (S1314).
Claims13
An image processing apparatus comprising:
- a first image input unit constructed to input a first image;
- a partial region extraction unit constructed to divide the first image into a plurality of partial regions, and to set combined regions each of which includes two or more partial regions;
- a feature amount extraction unit constructed to extract feature amounts of the respective partial regions;
- a comparison unit constructed to compare the respective extracted feature amounts of partial regions of the first image and feature amounts of a second image which are registered in a database;
- and
- a determination unit constructed to determine a similarity between a feature amount of at least one combined region and a feature amount of a similar region of the second image, wherein the similar region is a circumscribing rectangle including two or more partial regions of the second image each having a feature amount similar to any one of the respective extracted feature amounts of partial regions of the first image.
An image processing method comprising:
- a first image input step of inputting a first image into an image processing apparatus;
- a partial region extraction step of dividing the first image into a plurality of partial regions;
- a setting step of setting combined regions each of which includes two or more partial regions;
- a feature amount extraction step of extracting feature amounts of the respective partial regions;
- a comparison step of using the image processing apparatus to compare the respective extracted feature amounts of partial regions of the first image and feature amounts of a second image which are registered in a database;
- and
- a determination step of determining a similarity between a feature amount of at least one combined region and a feature amount of a similar region of the second image, wherein the similar region is a circumscribing rectangle including two or more partial regions of the second image each having a feature amount similar to any one of the respective extracted feature amounts of partial regions of the first image.
An image processing apparatus comprising:
- an input unit constructed to input plural images;
- a partial region extraction unit constructed to divide each of the plural images into a plurality of partial regions and extract a layout order of the respective plurality of partial regions from each of the plural images;
- a feature amount extraction unit constructed to extract feature amounts of the respective partial regions;
- a comparison unit constructed to compare partial regions in a first image and partial regions in a second image, based on the respective feature amounts and the respective layout orders extracted from the first and second images, wherein the first and second images are selected from the plural images;
- a setting unit constructed to set combined regions each of which includes two or more partial regions in the first image;
- and
- a determination unit constructed to determine a similarity between a feature amount of at least one combined region and a feature amount of a similar region of the second image, wherein the similar region is a circumscribing rectangle including two or more partial regions of the second image each having a feature amount similar to any one of the respective extracted feature amounts of partial regions of the first image.
An image processing method comprising:
- an input step of inputting plural images into an image processing apparatus;
- a partial region extraction step of dividing each of the plural images into a plurality of partial regions and extracting a layout order of the respective plurality of partial regions from each of the plural images;
- a feature amount extraction step of extracting feature amounts of the respective partial regions;
- a comparison step of using the image processing apparatus to compare partial regions in a first image and partial regions in a second image, based on the respective feature amounts and the respective layout orders extracted from the first and second images, wherein the first and second images are selected from the plural images;
- a setting step of setting combined regions each of which includes two or more partial regions in the first image;
- and
- a determination step of determining a similarity between a feature amount of at least one combined region and a feature amount of a similar region of the second image, wherein the similar region is a circumscribing rectangle including two or more partial regions of the second image each having a feature amount similar to any one of the respective extracted feature amounts of partial regions of the first image.
Dependencies are inferred from the claim text.