# US 2021/0241007 A1 · Automatic detection and positioning of structure faces > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0241007 A1 - Document ID: US20210241007A1 - Status: Published application - Kind: A1 - Application number: 17/234,411 - Filed: 2021-04-19 - Published: 2021-08-05 - Earliest claimed: 2018-12-05 - Assignees: HERE Global B.V. - Applicants: HERE Global B.V. - Inventors: David Doria; Engin Anil - Web page: https://patentgrep.com/patent/US20210241007A1 ## CPC - G06V 20/58: Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads - G01S 17/89: for mapping or imaging - G01S 17/931: of land vehicles - G05D 1/024: {in combination with a laser (lasers per se H01S)} - G05D 1/0248: {in combination with a laser (lasers per se H01S)} - G06T 7/75: {involving models} - G06V 20/588: {Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road} - G06T 2207/10028: Range image; Depth image; 3D point clouds - G06T 2207/30256: Lane; Road marking - G06T 2207/30261: Obstacle - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G01S: RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES - G05D: SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract An apparatus and method automatically detects and positions structure faces. After receiving data points describing a geographical area, neighborhoods are defined based on the data points and classified as linear, planar, or volumetric. Neighborhoods are merged into at least one cluster based on local surface normals. At least one bounding frame is fit to the at least one cluster and modified based on a field of interest. ## Claims ### Claim 1 A computer implemented method for automatic detection and positioning of object faces, the method comprising: receiving, by a processor, a plurality of data points describing a geographical area; defining, by the processor, a plurality of neighborhoods based on the plurality of data points; determining, by the processor, a classification for each of the plurality of neighborhoods; merging, by the processor, at least one neighborhood of the plurality of neighborhoods into one or more clusters based on local surface normals; fitting, by the processor, one or more bounding frames to the one or more clusters; and modifying, by the processor, the one or more bounding frames based on a field of interest. ### Claim 2 (depends on 1) The method of claim 1, wherein the plurality of data points are produced by a light detection and ranging (LIDAR) device. ### Claim 3 (depends on 1) The method of claim 1, further comprising: calculating, by the processor, a probability that the plurality of neighborhoods is linear. ### Claim 4 (depends on 1) The method of claim 1, further comprising: calculating, by the processor, a probability that the plurality of neighborhoods is planar. ### Claim 5 (depends on 1) The method of claim 1, further comprising: calculating, by the processor, a probability that the plurality of neighborhoods is volumetric. ### Claim 6 (depends on 1) The method of claim 1, further comprising: calculating, by the processor, one or more eigenvector values for the at least one of the plurality of neighborhoods. ### Claim 7 (depends on 6) The method of claim 6, further comprising: comparing the one or more eigenvector values to a threshold, wherein the classification is based on the comparison of the one or more eigenvector values to the threshold. ### Claim 8 (depends on 6) The method of claim 6, wherein the one or more eigenvector values are calculated using a covariance matrix . ### Claim 9 (depends on 1) The method of claim 1, wherein at least two of the plurality neighborhoods classified as planar are merged into the one or more clusters. ### Claim 10 (depends on 7) The method of claim 7, wherein the merging occurs when the at least two of the plurality neighborhoods are within a threshold distance and have parallel local surface normals. ### Claim 11 (depends on 1) The method of claim 1, wherein the plurality of neighborhoods are sized in a range that correlates with a size of the object faces, a density of the plurality of data points, or a combination thereof. ### Claim 12 (depends on 1) The method of claim 1 further comprising: comparing, by the processor, the one or more clusters to a size threshold, wherein the one or more bounding frames are fit to the one or more clusters that are larger than the size threshold. ### Claim 13 (depends on 1) The method of claim 1, wherein the fitting comprises: fitting a plane to the data points merged in the one or more clusters; projecting the data points onto the plane; and determining the one or more bounding frames based on the projected data points. ### Claim 14 (depends on 1) The method of claim 1, wherein the one or more bounding frames comprise at least two bounding frames that are merged when the at least two bounding frames comprise parallel surface normals relative to each other and are within a threshold distance from each other. ### Claim 15 (depends on 1) The method of claim 1, wherein the modifying comprises removing a portion of the one or more bounding frames that is within the field of interest. ### Claim 16 (depends on 1) The method of claim 1, wherein the field of interest comprises a road lane model. ### Claim 17 (depends on 1) The method of claim 1, further comprising: generating, by the processor, a localization model using the modified one or more bounding frames and the plurality of data points; and providing, by the processor, the localization model to an autonomous vehicle for localizing the autonomous vehicle. ### Claim 18 (depends on 1) The method of claim 1, wherein the object faces are part of a tunnel entrance, a building facade, a sign, or a combination thereof. ### Claim 19 A non-transitory computer-readable medium including instructions that when executed are operable to cause a processor to: receive a plurality of point cloud data points describing a geographical area; define a plurality of neighborhoods based on the plurality of point cloud data points; determine a classification for each of the plurality of neighborhoods; merge at least two neighborhoods of the plurality of neighborhoods into one or more clusters based on local surface normals and a distance threshold; fit one or more bounding frames to the one or more clusters that exceed a size threshold; and modify the one or more bounding frames based on a field of interest. ### Claim 20 An apparatus for automatic detection and positioning of object faces, the apparatus comprising: an interface module configured to receive a plurality of data points describing a geographical area, wherein the plurality of data points are used to define a plurality of neighborhoods; a calculation module configured to classify the plurality of neighborhoods; a cluster module configured to merge at least two neighborhoods of the plurality of neighborhoods into one or more clusters based on local surface normals; and an edit module configured to fit one or more bounding frames to the one or more clusters that exceed a size threshold and modify the one or more bounding frames based on a field of interest.