# US 11,113,839 B2 · Method, apparatus, and system for feature point detection > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,113,839 B2 - Document ID: US11113839B2 - Status: Granted - Kind: B2 - Application number: 16/286,265 - Filed: 2019-02-26 - Published: 2020-08-27 - Granted: 2021-09-07 - Earliest claimed: 2019-02-26 - Assignees: HERE Global B.V. - Applicants: Here Global B.V. - Inventors: Anish Mittal; Zhanwei Chen - Web page: https://patentgrep.com/patent/US11113839B2 ## CPC - G06V 40/168: {Feature extraction; Face representation} - G06T 7/73: using feature-based methods - G06V 10/454: {Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]} - G06V 10/82: using neural networks - G06T 2207/20021: Dividing image into blocks, subimages or windows - G06T 2207/20084: Artificial neural networks [ANN] - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract An approach is provided for feature point detection and representation. The approach, for example, involves processing (e.g., using a neural network or equivalent) image data associated with a grid cell of an image to determine a feature point corresponding to a position of a feature detected in the image data. The approach also involves encoding the position of the feature with respect to a coordinate system referenced to the grid cell. The output comprises one or more parameters indicating the encoded position, one or more attributes of the feature, or a combination thereof. ## Claims ### Claim 1 A method comprising: processing image data associated with a grid cell of an image to determine at least one feature point corresponding to a feature detected in the image; parametrically encoding a position of the at least one feature point with respect to a coordinate system referenced to the grid cell; and providing an output to represent the at least one feature point, wherein the output includes one or more parameters indicating an encoded position of the at least one feature point, one or more attributes of the feature, or a combination thereof. ### Claim 2 (depends on 1) The method of claim 1, wherein the output includes a respective output channel for the encoded location, the one or more attributes, or a combination thereof. ### Claim 3 (depends on 2) The method of claim 2, wherein the output is a tensor with dimensions based on an image size of the image, a cell size of the grid cell, and a total number of the respective output channels. ### Claim 4 (depends on 3) The method of claim 3, wherein the grid size is based on a density of feature points in the image. ### Claim 5 (depends on 3) The method of claim 3, wherein the total number of the respective output channels is further based on a number of feature points per grid cell. ### Claim 6 (depends on 5) The method of claim 5, wherein the number of feature points per grid cell is based on the grid size and the density of feature points. ### Claim 7 (depends on 1) The method of claim 1, wherein the processing of the image data in the grid cell is performed using a node of a neural network. ### Claim 8 (depends on 7) The method of claim 7, wherein a neighboring node of the neural network processes the image data to detect the feature and generate a redundant output to represent the feature point. ### Claim 9 (depends on 1) The method of claim 1, wherein the coordinate system includes an x-axis corresponding to a horizontal width of the grid cell and a y-axis corresponding to a vertical height of the grid cell. ### Claim 10 (depends on 1) The method of claim 1, wherein the one or more attributes include a domain of the feature detected in the image. ### Claim 11 An apparatus comprising: at least one processor; and at least one memory including computer program code for one or more programs, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following, segment an input image into a plurality of grid cells; process image data associated with a grid cell of an image to determine at least one feature point corresponding to a feature detected in the image; parametrically encode a position of the at least one feature point with respect to a coordinate system referenced to the grid cell; and provide an output to represent the at least one feature point, wherein the output includes one or more parameters indicating an encoded position of the at least one feature point, one or more attributes of the feature, or a combination thereof. ### Claim 12 (depends on 11) The apparatus of claim 11, wherein the output is a tensor with dimensions based on an image size of the image, a grid size of the grid cell, and a total number of output channels. ### Claim 13 (depends on 11) The apparatus of claim 11, wherein the processing of the image data in the grid cell is performed using a node of a neural network. ### Claim 14 (depends on 13) The apparatus of claim 13, wherein a neighboring node of the neural network processes the image data to detect the feature and generate a redundant output to represent the feature point. ### Claim 15 (depends on 11) The apparatus of claim 11, wherein the coordinate system includes an x-axis corresponding to a horizontal width of the grid cell and a y-axis corresponding to a vertical height of the grid cell. ### Claim 16 A non-transitory computer-readable storage medium, carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to perform: processing image data associated with a grid cell of an image to determine at least one feature point corresponding to a feature detected in the image; parametrically encoding a position of the at least one feature point with respect to a coordinate system referenced to the grid cell; and providing an output to represent the at least one feature point, wherein the output includes one or more parameters indicating an encoded position of the at least one feature point, one or more attributes of the feature, or a combination thereof. ### Claim 17 (depends on 16) The non-transitory computer-readable storage medium of claim 16, wherein the output includes a respective output channel for the encoded location, the one or more attributes, or a combination thereof. ### Claim 18 (depends on 17) The non-transitory computer-readable storage medium of claim 17, wherein the output is a tensor with a dimension based on an image size of the image, a cell size of the grid cell, and a total number of the respective output channels. ### Claim 19 (depends on 18) The non-transitory computer-readable storage medium of claim 18, wherein the grid size is based on a density of feature points in the image. ### Claim 20 (depends on 18) The non-transitory computer-readable storage medium of claim 18, wherein the total number of the respective output channels is further based on a number of feature points per grid cell.