# US 12,008,765 B2 · Candidate region estimation device, candidate region estimation method, and program > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,008,765 B2 - Document ID: US12008765B2 - Status: Granted - Kind: B2 - Application number: 17/265,166 - Filed: 2019-08-01 - Published: 2021-09-30 - Granted: 2024-06-11 - Earliest claimed: 2018-08-02 - Assignees: Nippon Telegraph and Telephone Corporation - Applicants: Nippon Telegraph and Telephone Corporation - Inventors: Yukito Watanabe; Shuhei Tarashima; Takashi Hosono; Jun Shimamura; Tetsuya Kinebuchi - Web page: https://patentgrep.com/patent/US12008765B2 ## CPC - G06T 7/12: Edge-based segmentation - G06T 7/11: Region-based segmentation - G06T 7/136: involving thresholding - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract The present invention makes it possible to estimate, with high precision, a candidate region indicating each of multiple target objects included in an image. A parameter determination unit 11 determines parameters to be used when detecting a boundary line of an image 101 based on a ratio between a density of boundary lines included in an image 101 and a density of boundary lines in a region indicated by region information 102 indicating the region including at least one of the multiple target objects included in the image 101. A boundary line detection unit 12 detects the boundary line in the image 101 using the parameter. For each of the multiple target objects included in the image 101, the region estimation unit 13 estimates the candidate region of the target object based on the detected boundary line. ## Claims ### Claim 1 A candidate region estimation apparatus configured to, for each of a plurality of target objects, estimate a candidate region that is a candidate for a region including said each target object, in an image including the plurality of target objects, the candidate region estimation apparatus comprising a processor configured to: receive input of the image and region information indicating a region including at least one of the plurality of target objects included in the image; detect boundary lines in the image using a Canny method and a plurality of types of predetermined parameters; determine a parameter to be used when detecting boundary lines in the image, based on a ratio between a density of boundary lines included in the image and a density of boundary lines in a region indicated by the region information; detect a boundary line in the image using the parameter; and estimate for each of the plurality of target objects included in the image, the candidate region of said each target based on the boundary lines detected by the boundary line detector. ### Claim 2 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein determining the parameter based on a ratio between the density of boundary lines included in the image and the density of boundary lines in a small region that is part of the region indicated by the region information. ### Claim 3 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein estimating the candidate region of said each target object from a region above or a region below the region indicated by the region information. ### Claim 4 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein estimating the candidate region of said each target object from a region to the right or a region to the left of the region indicated by the region information. ### Claim 5 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein selecting a candidate region output as an estimation result from the estimated candidate regions of said each target objects, based on the density of the boundary lines in the region indicated by the candidate region of the target object, and the density of the boundary lines in the region indicated by the region information. ### Claim 6 A candidate region estimation method for estimating, for each of a plurality of target objects, a candidate region that is a candidate for a region including said each target object, in an image including the plurality of target objects, the candidate region estimation method comprising: receiving, input of the image and region information indicating a region including at least one of the plurality of target objects included in the image; detecting boundary lines in the image using a Canny method and a plurality of types of predetermined parameters; determining, a parameter to be used when detecting a boundary line of the image, based on a ratio between a density of boundary lines included in the image and a density of boundary lines in a region indicated by the region information; detecting, boundary lines in the image using the parameter; and for each of the plurality target objects included in the image, estimating the candidate region of said each target object based on the boundary lines detected in the boundary line detection step. ### Claim 7 A system for estimating a candidate region, the system comprises: a processor; and a memory storing computer-executable program instructions that when executed by the processor cause the system to: receive, input of the image and region information indicating a region including at least one of the plurality of target objects included in the image; detect boundary lines in the image using a Canny method and a plurality of types of predetermined parameters; determine, a parameter to be used when detecting a boundary line of the image, based on a ratio between a density of boundary lines included in the image and a density of boundary lines in a region indicated by the region information; detect, boundary lines in the image using the parameter; and for each of the plurality of target objects included in the image, estimate, by a region estimator, the candidate region of said each target object based on the boundary lines detected in the boundary line detection step. ### Claim 8 (depends on 6) The candidate region estimation method according to claim 6, wherein determining the parameter based on a ratio between the density of boundary lines included in the image and the density of boundary lines in a small region that is part of the region indicated by the region information. ### Claim 9 (depends on 6) The candidate region estimation method according to claim 6, wherein estimating the candidate region of said each target object from a region above or a region below the region indicated by the region information. ### Claim 10 (depends on 6) The candidate region estimation method according to claim 6, wherein estimating the candidate region of said each target object from a region to the right or a region to the left of the region indicated by the region information. ### Claim 11 (depends on 6) The candidate region estimation method according to claim 6, wherein selecting a candidate region output as an estimation result from the estimated candidate regions of said each target objects, based on the density of the boundary lines in the region indicated by the candidate region of the target object, and the density of the boundary lines in the region indicated by the region information. ### Claim 12 (depends on 7) The system for estimating the candidate region according to claim 7, wherein determining the parameter based on a ratio between the density of boundary lines included in the image and the density of boundary lines in a small region that is part of the region indicated by the region information. ### Claim 13 (depends on 7) The system for estimating the candidate region according to claim 7, wherein estimating the candidate region of said each target object from a region to the right or a region to the left of the region indicated by the region information. ### Claim 14 (depends on 7) The system for estimating the candidate region according to claim 7, wherein estimating the candidate region of said each target object from a region to the right or a region to the left of the region indicated by the region information. ### Claim 15 (depends on 7) The system for estimating the candidate region according to claim 7, wherein selecting a candidate region output as an estimation result from the estimated candidate regions of said each target objects, based on the density of the boundary lines in the region indicated by the candidate region of the target object, and the density of the boundary lines in the region indicated by the region information. ### Claim 16 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein the density of boundary lines included in the image includes a first ratio of a number of pixels occupied by the boundary lines of the image with respect to a number of pixels inside the image, and wherein the density of boundary lines in the region includes a second ratio of a number of pixels occupied by the boundary lines of the region with respect to a number of pixels inside the region. ### Claim 17 (depends on 6) The candidate region estimation method according to claim 6, wherein the density of boundary lines included in the image includes a first ratio of a number of pixels occupied by the boundary lines of the image with respect to a number of pixels inside the image, and wherein the density of boundary lines in the region includes a second ratio of a number of pixels occupied by the boundary lines of the region with respect to a number of pixels inside the region. ### Claim 18 (depends on 7) The system for estimating the candidate region according to claim 7, wherein the density of boundary lines included in the image includes a first ratio of a number of pixels occupied by the boundary lines of the image with respect to a number of pixels inside the image, and wherein the density of boundary lines in the region includes a second ratio of a number of pixels occupied by the boundary lines of the region with respect to a number of pixels inside the region. ### Claim 19 (depends on 1) The candidate region estimation apparatus according to claim 1, wherein the parameter to be used when detecting boundary lines in the image includes a combination of: a standard deviation of a Gaussian function for convolution associated with the Canny edge detection on the image and the region, a first threshold value based on the density of boundary lines included in the image, and a second threshold value based on the density of boundary lines in a region indicated by the region information. ### Claim 20 (depends on 6) The candidate region estimation method according to claim 6, wherein the parameter to be used when detecting boundary lines in the image includes a combination of: a standard deviation of a Gaussian function for convolution associated with the Canny edge detection on the image and the region, a first threshold value based on the density of boundary lines included in the image, and a second threshold value based on the density of boundary lines in a region indicated by the region information.