# US 12,222,293 B2 · Detection device, label assignment method, and non-transitory computer-readable medium > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,222,293 B2 - Document ID: US12222293B2 - Status: Granted - Kind: B2 - Application number: 17/912,950 - Filed: 2020-03-31 - Published: 2023-05-04 - Granted: 2025-02-11 - Earliest claimed: 2020-03-31 - Assignees: NEC Corporation - Applicants: NEC Corporation - Inventors: Akira Tsuji; Junichi Abe; Taisuke Tanabe - Web page: https://patentgrep.com/patent/US12222293B2 ## CPC - G01N 21/8803: {Visual inspection (measuring projectors G01B9/08)} - G06V 10/44: Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components - G01N 21/8851: {Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges (G01N21/8806 and G01N21/93 - G01N21/95692 take precedence; optical measurement of dimensions G01B11/00; optical scanning G02B26/10; image transformation G06T3/00; computerised image enhancement G06T5/00; image processing per se for flaw detection G06T7/0002)} - G01N 21/9515: {Objects of complex shape, e.g. examined with use of a surface follower device (measuring contours and curvatures G01B11/24)} - G06T 7/0004: {Industrial image inspection} - G06T 7/13: Edge detection - G06V 10/60: relating to illumination properties, e.g. using a reflectance or lighting model - G06V 10/764: using classification, e.g. of video objects - G06V 10/77: Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation - G06V 10/945: {User interactive design; Environments; Toolboxes} - G06V 20/70: Labelling scene content, e.g. deriving syntactic or semantic representations - G01N 2021/8854: {Grading and classifying of flaws} - G01N 2021/8883: {involving the calculation of gauges, generating models} - G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A detection device (20) according to the present disclosure includes an acquisition unit (11) that acquires point cloud data indicating a distance to an object and luminance information obtained from reflected light of a beam emitted when the point cloud data is measured, an edge detection unit (12) that performs edge detection based on the luminance information, a classification unit (21) that classifies, based on distribution of points contained in a crack candidate being a set of points detected as an edge, a plurality of the crack candidates into one of groups, and a label assignment unit (23) that accepts input of a label regarding the group from a user having visually recognized a shape of the crack candidate belonging to the group. ## Claims ### Claim 1 A detection device comprising: at least one memory storing instructions, and at least one processor configured to execute the instructions that cause the at least one processor to: acquire point cloud data indicating a distance from a sensor to an object and luminance information obtained from reflected light of a beam emitted when the point cloud data is measured by the sensor; perform edge detection based on the luminance information; classify, based on distribution of points included in a crack candidate, these points being a set of points detected as an edge, a plurality of the crack candidates into one group of groups of crack candidates; select at least one crack candidate from the plurality of crack candidates belonging to the group based on a level of reliability assigned to the crack candidate; display the selected at least one crack candidate on a display unit; and accept input of a label regarding the group from a user who has visually recognized a shape of the crack candidate belonging to the group, wherein the reliability is determined according to an incident angle of the beam. ### Claim 2 (depends on 1) The detection device according to claim 1, wherein the at least one processor is further configured to execute the instructions to associate a label indicated by the crack candidate input from the user with the group containing the set of points. ### Claim 3 (depends on 1) The detection device according to claim 1, wherein the at least one processor is further configured to execute the instructions to accept, from the user, input of a label indicating whether the shape of the crack candidate is a crack. ### Claim 4 (depends on 1) The detection device according to claim 1, wherein the at least one processor is further configured to execute the instructions to classify the crack candidate into one of the groups based on a first principle component to a third principle component obtained from a principle component analysis performed on the set of points. ### Claim 5 A label assignment method comprising: acquiring point cloud data indicating a distance from a sensor to an object and luminance information obtained from reflected light of a beam emitted when the point cloud data is measured by the sensor; performing edge detection based on the luminance information; classifying, based on distribution of points included in a crack candidate, these points being a set of points detected as an edge, a plurality of the crack candidates into one of groups of crack candidates; selecting at least one crack candidate from the plurality of crack candidates belonging to the group based on a level of reliability assigned to the crack candidate; displaying the selected at least one crack candidate on a display unit; and accepting input of a label regarding the group from a user who has visually recognized a shape of the crack candidate belonging to the group, wherein the reliability is determined according to an incident angle of the beam. ### Claim 6 A non-transitory computer-readable medium storing a program causing a computer to execute: acquiring point cloud data indicating a distance from a sensor to an object and luminance information obtained from reflected light of a beam emitted when the point cloud data is measured by the sensor; performing edge detection based on the luminance information; classifying, based on distribution of points included in a crack candidate, these points being a set of points detected as an edge, a plurality of the crack candidates into one of groups of crack candidates; selecting at least one crack candidate from the plurality of crack candidates belonging to the group based on a level of reliability assigned to the crack candidate; displaying the selected at least one crack candidate on a display unit; and accepting input of a label regarding the group from a user who has visually recognized a shape of the crack candidate belonging to the group, wherein the reliability is determined according to an incident angle of the beam.