# US 10,503,886 B2 · Biometric authentication based on gait pattern or writing motion with an inertial measurement unit > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,503,886 B2 - Document ID: US10503886B2 - Status: Granted - Kind: B2 - Application number: 15/709,474 - Filed: 2017-09-20 - Published: 2018-03-29 - Granted: 2019-12-10 - Earliest claimed: 2016-09-27 - Assignees: Hong Kong Baptist University - Applicants: Hong Kong Baptist University - Inventors: Jeffrey Tai Kin Cheung - Web page: https://patentgrep.com/patent/US10503886B2 ## CPC - G06F 21/32: using biometric data, e.g. fingerprints, iris scans or voiceprints - G06V 40/25: {Recognition of walking or running movements, e.g. gait recognition} - G06F 21/6218: {to a system of files or objects, e.g. local or distributed file system or database} - G06F 2218/16: by matching signal segments - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract The present invention relates to use an Inertial Measurement Unit (IMU) to record the acceleration trajectory of a person's gait or pen-less handwriting motion or any predesignated gestures, and to convert the data to a unique biometric pattern. The pattern is unique for each case and can be used as biometric security authentication. ## Claims ### Claim 1 A method for biometric authentication based on a repeated movement, comprising: obtaining, by at least one first inertia measurement sensor located on each individual of a subject group, time dependent measurements of magnitude of acceleration vectors, xori(t), yori(t), and zori(t), in three Cartesian directions of said each individual of the subject group performing said repeated movement; computing, by a computer processor, an authentication function, S(t), for each individual of said subject group based on the acceleration vectors, wherein S(t)=[Ax(t)2+Ay(t)2+Az(t)2]1/2 and Ax(t)=xori(t)−xoff, Ay(t)=yori(t)−yoff, Az(t)=zori(t)−zoff, and xoff, yoff, and zoff are zero offset values of each acceleration vector at the start of the respective measurements and where the magnitude of the authentication function, S(t), as it proceeds through a trajectory characteristic of an individual during each cycle of the repeated movement, is converted into a single point in a multi-dimensional space; analyzing S(t), by the computer processor, over at least two cycles of said repeated movement using multivariate analysis to create an analyzed, reference S(t); storing on a computer readable storage medium the analyzed, reference S(t) for each individual to create stored, analyzed reference values S(t) of said subject group as biometric identifiers in a database of biometric identifiers; obtaining, by at least one second inertia measurement sensor located on a test subject, time dependent measurements of magnitude of acceleration vectors, xori(t), yori(t), and zori(t), in three Cartesian directions of said test subject performing said repeated movement; computing, by a computer processor, a test subject authentication function, S′(t), of said test subject based on the acceleration vectors; comparing the stored, analyzed reference values of S(t) for the subject group with the test subject authentication function S′(t); matching the stored, analyzed reference values of S(t) of said test subject with the test subject authentication function S′(t) as the biometric identifiers of the subject group in the database of biometric identifiers using a Partia Least Squares Discriminant Analysis; and authenticating said test subject if the test subject authentication function S′(t) is within a predetermined tolerance level of one or more stored, analyzed reference values of S(t) and wherein authentication is determined by a motion component of the movement with least phase shift sensitivity tolerance. ### Claim 2 (depends on 1) The method according to claim 1 wherein said repeated movement is hand writing. ### Claim 3 (depends on 1) The method according to claim 1 wherein said repeated movement is walking. ### Claim 4 (depends on 1) The method according to claim 1 wherein said repeated movement is running. ### Claim 5 A non-transitory computer-readable medium comprising computer instructions for causing a computer processor to obtain, by at least one first inertia measurement sensor located on each individual of a subject group, time dependent measurements of magnitude of acceleration vectors, xori(t), yori(t), and zori(t), in three Cartesian directions of said each individual of the subject group performing said repeated movement; compute an authentication function, S(t), for each individual of said subject group based on the acceleration vectors, wherein S(t)=[Ax(t)2+Ay(t)2+Az(t)2]1/2 and Ax(t)=xori(t)−xoff, Ay(t)=yori(t)−yoff, Az(t)=zori(t)−zoff, and xoff, yoff, and zoff are zero offset values of each acceleration vector at the start of the respective measurements and where the magnitude of the authentication function, S(t), as it proceeds through a trajectory characteristic of an individual during each cycle of the repeated movement, is converted into a single point in a multi-dimensional space; analyze S(t) over at least two cycles of said repeated movement using multivariate analysis to create an analyzed, reference S(t); store the analyzed, reference S(t) for each individual to create stored, analyzed reference values S(t) of said subject group as biometric identifiers in a database of biometric identifiers; obtain, by at least one second inertia measurement sensor located on a test subject, time dependent measurements of magnitude of acceleration vectors, xori(t), yori(t), and zori(t), in three Cartesian directions of said test subject performing said repeated movement; compute a test subject authentication function, S′(t), of said test subject based on the acceleration vectors; compare the stored, analyzed reference values of S(t) for the subject group with the test subject authentication function S′(t); match the stored, analyzed reference values of S(t) of said test subject with the test subject authentication function S′(t) as the biometric identifiers of the subject group in the database of biometric identifiers using a Partia Least Squares Discriminant Analysis; and authenticate said test subject if the test subject authentication function S′(t) is within a predetermined tolerance level of one or more stored, analyzed reference values of S(t) and wherein authentication is determined by a motion component of the movement with least phase shift sensitivity tolerance. ### Claim 6 (depends on 5) The non-transitory computer-readable medium according to claim 5, wherein the repeated movement is hand writing. ### Claim 7 (depends on 5) The non-transitory computer-readable medium according to claim 5, wherein the repeated movement is walking. ### Claim 8 (depends on 5) The non-transitory computer-readable medium according to claim 5, wherein the repeated movement is running. ### Claim 9 An apparatus for biometric authentication based on a repeated movement, comprising: at least one first inertia measurement sensor located on each individual of a subject group configured to obtain time dependent measurements of magnitude of acceleration vector, xori(t), yori(t), and zori(t), in three Cartesian directions of said each individual of said subject group performing said repeated movement; a first computer processor configured to compute an authentication function, S(t), of said subject group based on the acceleration vectors, wherein S(t)=[Ax(t)2+Ay(t)2+Az(t)2]1/2 and Ax(t)=xori(t)−xoff, Ay(t)=yori(t)−yoff, Az(t)=zori(t)−zoff, and xoff, yoff, and zoff are zero offset values of each acceleration vector at the start of the measurement and where the magnitude of the authentication function, S(t), as it proceeds through a trajectory characteristic of an individual during each cycle of the repeated movement, is converted into a single point in a multi-dimensional space; analyzing S(t), by the first computer processor, over at least two cycles of said repeated movement using multivariate analysis to create an analyzed, reference S(t); a computer readable storage medium for storing the analyzed reference S(t) for each individual of said subject group as biometric identifiers in a database of biometric identifiers; at least one second inertia measurement sensor located on a test subject configured to obtain time dependent measurements of magnitude of acceleration vector along xori(t), yori(t), and zori(t) directions and phase relationship of said test subject performing said repeated movement; a second computer processor configured to: compute a test subject authentication function S′(t) of said test subject based on the acceleration vectors; compare the stored, analyzed reference values of S(t) for the subject group with the test subject authentication function S′(t); match the stored, analyzed reference S(t) of said test subject with the test subject authentication function S′(t) as the biometric identifiers of the subject group in the database of biometric identifiers using a Partia Least Squares Discriminant Analysis; and authenticate said test subject if the test subject authentication function S′(t) is within a predetermined tolerance level of one or more stored, analyzed reference values of S(t) and wherein authentication is determined by a motion component of the movement with least phase shift sensitivity tolerance. ### Claim 10 (depends on 9) The apparatus according to claim 9 wherein said repeated movement is hand writing. ### Claim 11 (depends on 9) The apparatus according to claim 9 wherein said repeated movement is walking. ### Claim 12 (depends on 9) The apparatus according to claim 9 wherein said repeated movement is running.