# US 2017/0270673 A1 · Border Detection > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2017/0270673 A1 - Document ID: US20170270673A1 - Status: Published application - Kind: A1 - Application number: 15/310,054 - Filed: 2015-05-13 - Published: 2017-09-21 - Earliest claimed: 2014-05-13 - Assignees: String Labs Limited - Applicants: String Labs Limited - Inventors: Johan ØVERBYE - Web page: https://patentgrep.com/patent/US20170270673A1 ## CPC - G06T 7/13: Edge detection - G06T 7/136: involving thresholding - G06T 7/143: involving probabilistic approaches, e.g. Markov random field [MRF] modelling - G06T 7/174: involving the use of two or more images - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A method for assessing the presence of a border in an image, the method comprising: for each of two pixel positions estimating the strength of the second-order rate of change of a convolution of said image at the pixel position with respect to a plurality of directions and determining a binary condition of the pixel position in dependence on whether the strength exceeds a predetermined threshold; and assessing the presence of a border by comparing the two binary conditions. ## Claims ### Claim 1 A method for assessing the presence of a border in an image, the method comprising: for each of two pixel positions estimating the strength of the second-order rate of change of a convolution of said image at the pixel position with respect to a plurality of directions and determining a binary condition of the pixel position in dependence on whether the strength exceeds a predetermined threshold; and assessing the presence of a border by comparing the two binary conditions, wherein the strength of the second-order rate of change is estimated as the highest absolute value of the second-order derivatives of said convolved image at the pixel position with respect to the plurality of directions. ### Claim 2 (depends on 1) A method as claimed in claim 1, wherein the second-order derivative with respect to a given direction is estimated by observing an attribute of the convolved image at the pixel position and two other positions along the direction and on either side of the pixel position. ### Claim 3 (depends on 1) A method as claimed in claim 1, wherein the second-order derivative is estimated with respect to four directions. ### Claim 4 (depends on 1) A method as claimed in claim 1, wherein the directions are regularly rotationally offset. ### Claim 5 (depends on 1) A method as claimed in claim 1, wherein the image is defined by an array of pixels, each pixel having a set of discrete attributes. ### Claim 6 (depends on 1) A method as claimed in claim 1, wherein the said two pixels are adjacent pixels. ### Claim 7 (depends on 1) A method as claimed in claim 1, comprising assessing a border to be present between two pixel positions if the said binary conditions of those pixel positions differ. ### Claim 8 (depends on 1) A method as claimed in claim 1, comprising assessing a border not to be present between two directly adjacent pixel positions if the said binary conditions of those pixel positions are the same. ### Claim 9 (depends on 1) A method as claimed in claim 1, where the convolution of the image at a pixel position is dependent on data defining the image as represented in a plurality of pixels neighbouring that pixel position. ### Claim 10 (depends on 1) A method as claimed in claim 1, where the convolution of the image is an isotropic Gaussian convolution. ### Claim 11 A device for assessing the presence of a border in an image, the device being configured to perform the steps of: for each of two pixel positions estimating the strength of the second-order rate of change of a convolution of said image at the pixel position with respect to a plurality of directions and determining a binary condition of the pixel position in dependence on whether the strength exceeds a predetermined threshold; and assessing the presence of a border by comparing the two binary conditions, wherein the strength of the second-order rate of change is estimated as the highest absolute value of the second-order derivatives of said convolved image at the pixel position with respect to the plurality of directions. ### Claim 12 (depends on 11) A device as claimed in claim 11, wherein the device comprises a camera and the image is derived from the camera. ### Claim 13 (depends on 11) A device as claimed in claim 11, wherein the second-order derivative is estimated with respect to four directions. ### Claim 14 (depends on 11) A device as claimed in claim 11, wherein the device is further configured to perform the step of assessing a border to be present between two pixel positions if the said binary conditions of those pixel positions differ. ### Claim 15 (depends on 11) A device as claimed in claim 11, wherein the second-order derivative with respect to a given direction is estimated by observing an attribute of the convolved image at the pixel position and two other positions along the direction and on either side of the pixel position. ### Claim 16 (depends on 11) A device as claimed in claim 11, wherein the directions are regularly rotationally offset. ### Claim 17 (depends on 11) A device as claimed in claim 11, wherein the device is further configured to perform the step of assessing a border not to be present between two directly adjacent pixel positions if the said binary conditions of those pixel positions are the same. ### Claim 18 (depends on 11) A device as claimed in claim 11, where the convolution of the image at a pixel position is dependent on data defining the image as represented in a plurality of pixels neighbouring that pixel position. ### Claim 19 (depends on 11) A device as claimed in claim 11, where the convolution of the image is an isotropic Gaussian convolution. ### Claim 20 A non-transitory computer readable storage medium having stored thereon computer readable instructions that, when executed at a computer system, cause the computer system to perform a method for assessing the presence of a border in an image, the method comprising: for each of two pixel positions estimating the strength of the second-order rate of change of a convolution of said image at the pixel position with respect to a plurality of directions and determining a binary condition of the pixel position in dependence on whether the strength exceeds a predetermined threshold; and assessing the presence of a border by comparing the two binary conditions, wherein the strength of the second-order rate of change is estimated as the highest absolute value of the second-order derivatives of said convolved image at the pixel position with respect to the plurality of directions.