# US 10,482,348 B1 · System and method for tracking coherently structured feature dynamically defined with migratory medium > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,482,348 B1 - Document ID: US10482348B1 - Status: Granted - Kind: B1 - Application number: 15/900,275 - Filed: 2018-02-20 - Granted: 2019-11-19 - Earliest claimed: 2012-01-22 - Assignees: SR2 Group, LLC - Applicants: Sr2 Group, LLC - Inventors: Jeffrey Mark Sieracki; John J. Dellomo - Web page: https://patentgrep.com/patent/US10482348B1 ## CPC - G06T 7/246: using feature-based methods, e.g. the tracking of corners or segments - G06T 7/20: Analysis of motion (motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51) - G06V 10/462: {Salient features, e.g. scale invariant feature transforms [SIFT]} - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A system and method are provided for discriminating and tracking a coherently structured feature dynamically defined in evolving manner within a migratory medium. A data set is captured for a plurality of physical points defined within a multi-dimensional physical space, in terms of a plurality of scalar parameter values. At least one pre-selected target feature type is established, as is a scalar field predefined by at least one of the scalar parameters. A sparse set of key points is selectively generated a within the scalar field. Each key point is associated with one of the physical points and descriptive information adaptively determined therefor from the data content within a neighborhood of the physical point coinciding therewith. At least one predetermined feature-based operation is executed responsive to the descriptive information of the key points as a surrogate for execution generally on the data set. ## Claims ### Claim 1 A method for discriminating and tracking a coherently structured flow feature dynamically evolving over time within a migratory medium, comprising: capturing a data set for a medium undergoing migratory flow with respect to a plurality of physical points within a physical space, said data set including data content captured for each physical point in terms of a plurality of scalar parameter values; selectively generating a sparse set of key points within a scalar field defined by at least one of said scalar parameters, said key points corresponding to at least one coherently structured flow feature of a predetermined feature type occurring and substantially recurring over time with respect to other portions of the migratory medium; adaptively determining descriptive information for each said key point according to a proximity-based weighting of said data content relative to said key point within a neighborhood of a corresponding physical point, said neighborhood being defined responsive to said predetermined feature type; and, executing over a time period at least one predetermined feature-based operation on said data content of said key points within said sparse set as a surrogate for execution on said data set entirely to generate evolving flow feature data substantially representative of said data set; wherein said feature-based operation includes discriminating and tracking said coherently structured flow feature occurring with evolving shape within different data sets. ### Claim 2 A system for discriminating and tracking a coherently structured flow feature dynamically evolving over time within a migratory medium, comprising: computer memory receiving a data set for a medium undergoing migratory flow with respect to a plurality of physical points within a multi-dimensional physical space, said data set including data content captured for each physical point in terms of a plurality of scalar parameter values; a decomposition processor coupled to said computer memory selectively generating a sparse set of key points within a scalar field predefined by at least one of said scalar parameters, said key points corresponding to at least one coherently structured flow feature of a pre-selected feature type occurring and substantially recurring over time with respect to other portions of the migratory medium; a data reduction processor coupled to said decomposition processor, said data reduction processor adaptively determining descriptive information for each said key point according to a proximity-based weighting of said data content within a neighborhood of said physical point, said neighborhood being defined responsive to said pre-defined feature type; and, a feature processor coupled to said data reduction processor, said feature processor executing over a time period at least one predetermined feature-based operation on said data content of said key points within said sparse set as a surrogate for execution on said data set entirely to generate evolving flow feature data substantially representative of said data set; wherein said feature-based operation includes discriminating and tracking said coherently structured flow feature occurring with evolving shape within different data sets. ### Claim 3 (depends on 1) The method as recited in claim 1, wherein the different data sets are captured for different points in time within the flow of the migratory medium. ### Claim 4 (depends on 1) The method as recited in claim 1, wherein the data set includes simulation data captured for the medium undergoing migratory flow at the plurality of physical points within the multi-dimensional physical space. ### Claim 5 (depends on 1) The method as recited in claim 1, wherein the physical space of said data set is defined by at least three dimensions; and, said key points are identified according to a process selected from the group consisting of: a scale invariant feature transformation (SIFT) of the data set, and a Speeded Up Robust Features (SURF) of the data set. ### Claim 6 (depends on 1) The method as recited in claim 1, further comprising data reduction of said sparse set to reduce in number said key points independently indexed therein, said data reduction including: identifying a match between at least a pair of said key points according to a similarity criteria determined according to the descriptive information respectively thereof; and, combining the matched key points in index within said sparse set for collective indexing during execution of said predetermined feature-based operation. ### Claim 7 (depends on 6) The method as recited in claim 6, wherein said data reduction includes generating a descriptive geometric parameter of one said key point relative to at least one other said key point for identifying a match between said key points. ### Claim 8 (depends on 7) The method as recited in claim 7, wherein a match is identified between said key points of different data sets. ### Claim 9 (depends on 7) The method as recited in claim 7, wherein said feature-based operation includes volume tracking according to a predetermined volume defined by a set of volume vertices within a space described in the physical space by said sparse set of key points, said volume tracking comprising: identifying a match amongst said key points disposed within said predetermined volume to generate a constellation of key points; and, predictively adjusting said predetermined volume for application to said key points generated for a subsequent time slice to account for the evolution of said coherently structured flow feature between time slices. ### Claim 10 (depends on 7) The method as recited in claim 7, wherein said feature-based operation includes adaptive isosurface identification, comprising: predetermining a volume by a set of volume vertices within a space described in the physical space by said sparse set of key points; computing an isovalue for a preselected one of said scalar parameters occurring in said data content for said key points disposed within the predetermined volume; and, defining an isosurface based on said isovalue. ### Claim 11 (depends on 10) The method as recited in claim 10, wherein said feature-based operation includes isosurface tracking, comprising: identifying a match amongst said key points disposed within said isosurface to generate a constellation of key points; and, predictively adjusting the predetermined volume for application to said key points generated for a subsequent time slice to account for the evolution of said coherently structured flow feature between time slices. ### Claim 12 (depends on 1) The method as recited in claim 1, wherein the descriptive information for each key point includes a descriptor and primary orientation, the descriptor being defined with respect to a scale measure and an orientation measure determined from said data content within said neighborhood of said key point, the primary orientation being defined based on the orientation measure. ### Claim 13 (depends on 1) The method as recited in claim 1, wherein said scalar field is defined according to at least one non-vector parameter selected from the group consisting of: pressure component P along one dimension of the physical space, a Q-criterion, and a stream-wise velocity component W. ### Claim 14 (depends on 2) The system as recited in claim 2, wherein the different data sets are captured for different points in time within the flow of the migratory medium. ### Claim 15 (depends on 2) The system as recited in claim 2, wherein the data set includes simulation data captured for the medium undergoing migratory flow at the plurality of physical points within the multi-dimensional physical space. ### Claim 16 (depends on 2) The system as recited in claim 2, wherein the physical space of said data set is defined by at least three dimensions; and, said key points are identified according to a process selected from the group consisting of: a scale invariant feature transformation (SIFT) of the data set, and a Speeded Up Robust Features (SURF) of the data set. ### Claim 17 (depends on 2) The system as recited in claim 2, wherein said data reduction processor reduces in number the key points independently indexed in said sparse set, said data reduction processor executing to: identify a match between at least a pair of said key points according to a similarity criteria determined according to the respective descriptive information thereof; and, combine the matched key points in index within said sparse set for collective indexing during execution of said predetermined feature-based operation. ### Claim 18 (depends on 17) The system as recited in claim 17, wherein said data reduction processor further executes to generate a descriptive geometric parameter of one said key point relative to at least one other said key point to identify a match between said key points. ### Claim 19 (depends on 18) The system as recited in claim 18, wherein said feature processor executes volume tracking according to a predetermined volume defined by a set of volume vertices within a space described in the physical space by said sparse set of key points, said volume tracking comprising: identifying a match amongst said key points disposed within said predetermined volume to generate a constellation of key points; and, predictively adjusting said predetermined volume for application to said key points generated for a subsequent time slice to account for the evolution of said coherently structured flow feature between time slices. ### Claim 20 (depends on 18) The system as recited in claim 18, wherein said feature processor executes adaptive isosurface identification and isosurface tracking, comprising: establishing a predetermined volume defined by a set of volume vertices within a space described in the physical space by said sparse set of key points; computing an isovalue for a preselected one of said scalar parameters occurring in said data content for said key points disposed within said predetermined volume; defining an isosurface based on said isovalue; identifying a match amongst said key points disposed within said isosurface to generate a constellation of key points; and, predictively adjusting the predetermined volume for application to said key points generated for a subsequent time slice to account for the evolution of said coherently structured flow feature between time slices. ### Claim 21 (depends on 2) The system as recited in claim 2, wherein said descriptive information for each said key point includes a descriptor and primary orientation, the descriptor being defined with respect to a scale measure and an orientation measure determined from said data content within said neighborhood of said key point, the primary orientation being defined based on said orientation measure. ### Claim 22 (depends on 2) The system as recited in claim 2, wherein said scalar field is defined according to at least one non-vector parameter selected from the group consisting of: pressure component P along one dimension of the physical space, a Q-criterion, and a stream-wise velocity component W. ### Claim 23 A method for discriminating and tracking a predetermined flow feature dynamically evolving over time within a migratory medium, comprising: capturing a data set for a medium undergoing dynamically evolving flow with respect to a plurality of physical points within a physical space, said data set including data content captured for each physical point in terms of a plurality of scalar parameter values; selectively generating a sparse set of key points within a scalar field defined by a plurality of the scalar parameters, said key points corresponding to the predetermined flow feature recurring in evolving form over time with respect to other portions of the migratory medium; adaptively determining descriptive information for each said key point according to a proximity-based weighting of said data content relative to said key point within a neighborhood of a corresponding physical point, said neighborhood being defined according to a feature type of the predetermined flow feature; and, executing at least one predetermined feature-based operation on said data content of said key points within said sparse set over a time period as a surrogate for execution on said data set as a whole, said feature-based operation including discriminating and tracking of said predetermined flow feature evolving in shape over different data sets responsive to identification of a match between said key points of different data sets.