1280 results about "Structured light" patented technology

The invention relates to the technical field of detection, in particular to a complex shell detection method and system based on structured light scanning. The method comprises the following steps: calculating a wrapped phase value and a phase modulation amplitude of each pixel point under each frequency, calculating discrete probability distribution containing candidate stripe series and a corresponding probability for each pixel point based on a multi-frequency wrapped phase difference, generating an initial pixel confidence map, and constructing an energy function; the penalty weight of the smooth item is determined according to the phase modulation amplitude and the wrapped phase gradient local variance, obtaining an initial fringe order distribution diagram, updating discrete probability distribution by using an initial pixel confidence map, re-minimizing an energy function to obtain a corrected fringe order distribution diagram, and reconstructing a final three-dimensional shape. And iteratively optimizing a projector principal point and a radial distortion coefficient in system calibration parameters. According to the scheme, the geometric details of the surface can be reserved, and the processing effect and the measurement precision of the complex curved surface and the high-curvature area are improved.

The invention discloses a straight gear shape-free measurement method based on single-group line structured light. The method comprises the following steps: constructing an optical platform coordinate system, carrying out line structured light coordinate system calibration through a spherical calibration body, and calibrating the rotation center of a precision turntable; installing a measured straight gear, establishing a straight gear coordinate system, and adjusting the position of the linear structure optical measuring head, so that the measuring field of view of the linear structure optical measuring head covers a root circle to an addendum circle of the straight gear; the straight gear is installed on a precision rotary table, the straight gear is driven to rotate at a constant speed, a linear structure optical probe scans and collects straight gear contour point cloud data and circular grating angle information, and tooth surface three-dimensional point cloud data is obtained; based on the three-dimensional point cloud data of the tooth surface, extracting an addendum circle, a dedendum circle and an involute tooth profile, and calculating spur gear parameters; performing reverse three-dimensional shape modification tooth surface identification on the shape modification straight gear to generate a three-dimensional deviation diagram; and outputting a measurement result and printing a report. The problems of shape-free rapid measurement, modification tooth surface identification and tooth surface three-dimensional deviation characterization of the straight gear with unknown basic parameters are solved.

The invention relates to the technical field of workpiece detection, and discloses an intelligent visual detection method and system for the surface of a workpiece, and the method comprises the steps: projecting a polarization structure light pattern to the surface of the workpiece, and synchronously collecting an image of the polarization structure light pattern after deformation on the surface of the workpiece; separating a structured light component and a polarized light component from the image, reconstructing a surface three-dimensional shape according to the structured light component, and generating surface polarization degree distribution according to the polarized light component; performing multi-scale gradient scanning on the surface three-dimensional morphology, extracting edge change features, performing texture consistency analysis on the surface polarization degree distribution, and identifying an abnormal texture mode; identifying a mirror reflection area based on the surface polarization degree distribution, performing height compensation on a corresponding area in the surface three-dimensional morphology, generating a corrected surface three-dimensional morphology, and combining edge change features and an abnormal texture mode; according to the invention, the accuracy and robustness of defect identification can be improved, and the reliability and efficiency of the system in industrial field application are enhanced.

The embodiment of the invention provides an aircraft skin damage identification method and device, equipment, a storage medium and a product. The method comprises the following steps: acquiring an initial point cloud of an aircraft skin surface through structured light equipment, and performing data preprocessing on the initial point cloud to obtain a preliminary correction point cloud; calculating reflection intensity distribution of each area on the surface of the aircraft skin based on the preliminary correction point cloud, and determining a compensated point cloud; extracting geometric features from the compensated point cloud, and performing denoising processing on the compensated point cloud to obtain a denoised point cloud; constructing a three-dimensional grid model representing a three-dimensional structure of the aircraft skin surface based on the de-noised point cloud, and determining a damage candidate area of the aircraft skin surface; obtaining a point cloud subset associated with the damage candidate area and carrying out damage type classification on the point cloud subset to obtain a classification result of damage types; according to a classification result, generating a damage labeling graph; and based on the damage marking graph, calibrating the damage size parameter to generate a final damage report.

The invention belongs to the technical field of structured light three-dimensional measurement, and particularly relates to a super-density stripe structured light three-dimensional surface scanning method and system, and the method comprises the steps: obtaining camera parameters; projecting a super-density stripe structure light pattern formed by alternating wide bright and dark stripes to an object; acquiring an image by using a camera, and extracting a central point pixel coordinate of the stripe-structured light pattern by using a sub-pixel precision positioning method; calibrating a light plane equation of each stripe structure light pattern under a camera coordinate system by means of a plane target; reversely projecting the pixel coordinate of the central point of the stripe-structured light pattern into a ray and intersecting the ray with a corresponding light plane to obtain the three-dimensional coordinate of the ray and form an initial three-dimensional point cloud; and finally, carrying out gray scale estimation and filling on missing data caused by signal saturation or difficult positioning based on a radiation brightness physical model to obtain complete and accurate three-dimensional surface shape data. According to the invention, high-precision, high-density and high-integrity three-dimensional surface scanning is realized, and the reliability and applicability of a measurement result are remarkably improved.

Disclosed is a three-dimensional reconstruction method based on kinematic calibration of line structured light point set individualization, and the method includes the following steps: rigidly assembling an area-array camera and a line structured light emitter, and mounting the assembly on a motion platform; calibrating structural parameters of the motion platform of a measuring device by using a reference radius of a standard sphere and line structured light scanned data; registering a spatial position relationship of a line structured light point set by taking a calibrated spherical center position of the standard sphere and the reference radius of the standard sphere as constraints; controlling the five-axis motion platform to move and acquiring scanned data of a measured member under different poses; and completing high-quality three-dimensional model reconstruction of the measured member subjected to multi-pose scanning by using the scanned data and a real-time point cloud registration matrix.

The invention provides a box girder defect recognition method and system based on structured light and deep reinforcement learning, and relates to the technical field of box girder defect recognizing.The method comprises the steps that equipment integration and joint calibration are conducted on a to-be-detected box girder component, structured light projection is conducted according to a preset track, and a to-be-detected box girder component is obtained; the method comprises the following steps: triggering an event camera to image by utilizing controlled stroboscopic scanning of structured light, obtaining an event stream and a structured light depth map, carrying out representation conversion and multi-modal alignment, obtaining multi-modal mapping alignment data to carry out primary defect identification, obtaining an initial defect identification result, and triggering a deep reinforcement learning active perceptron to carry out perception strategy analysis. And obtaining a perception strategy analysis result, and carrying out deepening defect identification on the to-be-detected box girder component to obtain a defect identification result. The technical problems that in the prior art, box girder defect detection is low in detection efficiency, not high in accuracy and poor in detection comprehensiveness are solved. The technical effect of improving the efficiency, accuracy and comprehensiveness of box girder defect detection is achieved.

A surface evaluation system that includes one or more vision systems that generate target surface data during evaluation of a surface, the one or more vision systems comprising two or more of: at least one light, a camera, a structured light camera, a laser scanner and a 3D scanner.

The invention belongs to the technical field of computer vision and optical measurement, and discloses a three-dimensional reconstruction method and device based on multi-frequency fringe synchronous optimization and a medium, and the method comprises the following steps: based on a monocular structured light system, projecting a plurality of groups of sine fringe images with different frequencies to the surface of an object to be measured through a projector, and carrying out the stepping of a plurality of phases, acquiring a corresponding image sequence by using a camera; using a structured light phase shift method to obtain a relative phase value of a pixel point; obtaining an initial absolute phase value and modulation intensity based on a structured light phase shift method and a multi-frequency heterodyne method; and carrying out joint optimization on the absolute phase and modulation intensity parameters of the multi-frequency fringes through a Gaussian-Newton iteration method. According to the method, the anti-noise capability and the reconstruction precision are improved, a parameter adjustment and threshold recovery mechanism is introduced in the optimization process, the convergence stability and abnormal value elimination are ensured, and the method is suitable for high-precision structured light application scenes such as industrial vision measurement and complex surface detection.

The invention relates to a defect detection device and method, computer equipment and a computer readable storage medium. The device comprises an imaging subsystem, a structured light illumination subsystem and a data fusion processing unit. The structured light illumination subsystem comprises a coupling device, and the coupling device is used for coupling an illumination light path of the structured light illumination subsystem into an imaging light path of the imaging subsystem; the imaging subsystem is used for synchronously acquiring a light field image of the tested sample under the condition that the structured light illumination subsystem projects the structured light to the tested sample, and the light field image comprises light field depth data and a structured light image; and the data fusion processing unit is used for fusing the light field depth data and the structured light image to reconstruct the three-dimensional morphology of the detected sample, and generating a defect detection result of the detected sample based on the three-dimensional morphology. By adopting the device, the reconstruction precision of the three-dimensional morphology can be improved, and tiny pit defects on the surface of a transparent / weak texture sample can be identified.

The utility model discloses a silicon optical wafer test device and an alignment coupling mechanism and a test system thereof. The alignment coupling mechanism comprises an optical fiber coupling module, a piezoelectric displacement table and a tail end connecting structure which are connected in sequence. The optical fiber coupling module comprises three translation assemblies and three rotating assemblies which are sequentially connected in series; the three translation assemblies are used for driving the tail end connecting structure to do translation motion in the first direction, the second direction and the third direction correspondingly. The three rotating assemblies are used for driving the tail end connecting structure to rotate with the first rotating shaft, the second rotating shaft and the third rotating shaft as the center correspondingly. The first direction, the second direction and the third direction are vertical in pairs; the first rotating shaft, the second rotating shaft and the third rotating shaft are perpendicular pairwise. Each assembly only needs to achieve motion adjustment on one degree of freedom, the function is single, precise control over motion adjustment is facilitated, all the assemblies only need to be sequentially connected in series, spliced and connected, implementation is simple and easy, and equipment cost can be reduced.

The invention provides a structured light projector calibration method and system and a medium, and the method comprises the steps: projecting a multi-frequency stripe pattern to a target region based on an MEMS micromirror array module, obtaining a distorted image sequence, and carrying out the phase decoding, and obtaining a phase diagram corresponding to each frequency; resolving the phase diagram based on a phase shift method and an optical path triangulation principle to obtain a preliminary three-dimensional point coordinate; constructing a neural network model and a calibration sample set, obtaining a real three-dimensional coordinate by using a standard calibration plate or a known curved surface, and calculating the deviation between the initial three-dimensional point coordinate and the real three-dimensional coordinate to obtain a training sample set; and training the neural network model according to a training sample set based on an Euclidean distance loss function, outputting a three-dimensional offset compensation amount according to the neural network model, correcting the initial three-dimensional point coordinates, and realizing an end-to-end error self-compensation mechanism through the neural network model, so that deviation compensation is effectively performed on the three-dimensional point coordinates, and the accuracy of the three-dimensional point coordinates is improved. And the three-dimensional reconstruction precision is improved.

The invention provides a roadbed slope collapse early warning method and system based on linear structured light scanning, and is suitable for the technical field of slope safety monitoring. The method comprises the steps of unmanned aerial vehicle-scanning equipment cooperative deployment, fixed path automatic scanning, original point cloud data space registration, data feature matching and early warning response. The unmanned aerial vehicle carries line structure light scanning equipment, fixed path automatic scanning and real-time single-frame image three-dimensional reconstruction are combined, space-time continuous point cloud data of the roadbed slope are obtained, efficient dynamic monitoring of large-range slope surface deformation is achieved, and the problems that traditional manual inspection is low in efficiency and insufficient in coverage are solved; through the space-time synchronization registration and abnormal point intelligent identification technology, real slump displacement and environment interference are effectively distinguished by utilizing displacement direction analysis and threshold screening, the false alarm rate is greatly reduced, and the anti-interference capability in a complex scene is enhanced; the slope sudden change characteristics of slope crest displacement are matched through regularized piecewise linear modeling, so that the early warning response time is advanced.

The invention discloses a method for performing visual inspection on surface forming quality characteristics of a welding seam by utilizing three-dimensional reconstruction, which comprises the following steps of: S1, performing uniform-speed scanning on the welding seam along the direction of a welding bead by using linear structured light emitted by a linear laser, and performing image acquisition on linear structured light stripes formed on the surface of the welding seam by using an industrial camera; s2, converting the acquired RGB image into a gray level image frame by frame, and carrying out image distortion correction and image denoising processing; s3, carrying out ROI (Region of Interest) positioning on the formed line structured light stripe image on the surface of the welding seam by adopting a pixel point gray level distribution calculation method; s4, carrying out image segmentation on the ROI region of the weld line structured light stripe image; and S5, carrying out sub-pixel-level center line extraction on the ROI of the weld line structured light stripe image by adopting a gray extreme value weighted centroid method. Automatic detection of welding seam forming quality characteristics such as laser welding and electric arc welding can be achieved, and the operation efficiency of a production line and the welding seam quality detection precision can be greatly improved.

The invention discloses a full-automatic glass cover plate detection system and method based on multi-sensor fusion, and belongs to the technical field of industrial automatic detection. The posture of a glass cover plate is obtained and calibrated, and a standardized input material is formed; obtaining a three-dimensional point cloud model through structured light projection and phase reconstruction, and extracting a thickness feature matrix; synchronously acquiring a visible light image and an infrared thermal image, and extracting texture features and a thermal anomaly image; mapping the multi-source data to a unified coordinate system, and constructing a micro-defect fusion map; extracting local feature vectors by using a sliding window method, and inputting the local feature vectors into a convolutional neural network to identify defect types, levels and space coordinates; high-density point cloud scanning and curvature analysis are performed on the defect area, and microscopic contour features are extracted and compared to confirm authenticity; if the defect is effective, converting the coordinate into a control parameter, and pre-marking on the surface of the glass cover plate; according to the method, high-precision, multi-dimensional and traceable defect detection is realized, and the method is suitable for an online full-automatic quality inspection scene of the intelligent terminal glass cover plate.

Disclosed in the present invention is a three-dimensional reconstruction method based on kinematic calibration for the monomerization of a line-structured light point set. The method comprises: rigidly combining an area-array camera with a linear-line-structured light emitter, and mounting same on a motion platform; using a standard radius value of a standard sphere and line-structured light scanning data to calibrate structural parameters of the motion platform of a measurement apparatus; with the calibrated spherical center position of the standard sphere and the standard radius of the standard sphere as constraints, calibrating a spatial position relationship of a line-scanned point set; controlling the motion of the five-axis motion platform, and acquiring scanning data of a measured piece in different poses; and using the scanning data and a real-time point cloud calibration matrix to complete high-quality three-dimensional model reconstruction of the measured piece from its multi-pose scanning. The system comprises: an apparatus fixing module, a calibration module, a point cloud calibration module, a scanning module, and a point cloud reconstruction module. By means of using the present invention, the reconstruction accuracy of a line-structured light three-dimensional point cloud model can be improved. The present invention can be widely applied to the field of three-dimensional reconstruction.

The invention relates to a structured light illumination confocal super-resolution measurement system and method, and relates to the technical field of optical measurement. The system comprises a digital micromirror device (DMD), a sample platform, an electric displacement platform, a microscopic imaging assembly, a complementary metal oxide semiconductor (CMOS) imaging assembly and computer equipment, and the DMD is used for adjusting the point light source array and the structured light field. According to the DMD-based confocal super-resolution measurement method, a structured light field and a signal-noise separation technology are combined, and high-precision and high-resolution three-dimensional surface reconstruction can be realized. By using the flexibility of the DMD and the advantages of the super-resolution technology, the spatial resolution and speed of confocal imaging are effectively improved, so that the measurement process is more accurate. In addition, low-rank decomposition and self-supervised deep learning methods are fused, the image reconstruction process is optimized, noise interference is suppressed, and the overall performance of the system is further improved.

The invention discloses a PCB soldering paste printing three-dimensional defect detection system based on multispectral imaging, and relates to the field of machine vision detection, and the system comprises a multispectral image acquisition module, a four-channel industrial camera based on RGB and near infrared, and a tunable LED light source array; acquiring reflection characteristic data under different penetration depths through a wavelength switching mechanism; the motion control trigger module is used for realizing positioning based on an XYZ three-axis objective table driven by a servo motor in cooperation with feedback of an encoder; a pulse width modulation signal is adopted to coordinate the moving speed of a camera shutter and a platform; according to the method, RGB and NIR four-channel imaging is combined with Beer-Lambert law modeling, so that double verification of material component quantitative analysis and three-dimensional shape reconstruction is realized, and limitation of monocular vision is avoided; structured light projection and binocular stereo matching technologies are adopted, cross-frame data alignment is realized in cooperation with an ICP algorithm, and detail defects can be better detected.

The invention relates to the technical field of machine vision detection, in particular to a flaw detection method and system for large mold production based on machine vision. The method comprises the following steps: positioning the surface and a key curvature area of a spherical acoustic focusing lens mold, and establishing a three-dimensional coordinate reference frame of the spherical acoustic focusing lens mold; the method comprises the following steps: acquiring an image of a spherical area of a whole spherical acoustic focusing lens mold, preprocessing the image, and acquiring point cloud data of the surface of the spherical acoustic focusing lens mold; and based on the preprocessed image, utilizing a convolutional neural network model to identify the surface defect of the spherical acoustic focusing lens mold, and combining point cloud data to detect an area with geometric anomaly. According to the method, a unified three-dimensional coordinate reference frame is established, and high-resolution image data and point cloud data generated by high-precision structured light / laser scanning are precisely registered in space.

The invention discloses an on-site mixed loading explosive truck hole searching and accurate hole aligning method which comprises the following steps: acquiring a design coordinate set of each blast hole in a to-be-loaded area, acquiring a tail end coordinate of the tail end of a mechanical arm under a geographic coordinate system, and establishing a mechanical arm base coordinate system by taking a base coordinate as an original point; the joint angle and the expansion and contraction amount of the mechanical arm are obtained through iterative inverse kinematics with constraints; controlling the mechanical arm to carry out coarse alignment iteration until the coarse deviation is smaller than a first threshold value; obtaining a fine positioning result of the hole center and the hole axis direction; calculating the fine deviation of the tail end hole aligning assembly relative to the hole center and the hole shaft; after a hole alignment completion signal is output, triggering insertion of a charging conduit and checking of a charging interlocking condition; and when the interlocking condition is met, the charging operation is executed. The GNSS RTK, the laser radar and the 3D structured light camera are fused, quick positioning and alignment of the blast hole are achieved, and the charging efficiency and safety are greatly improved.

The invention discloses a regenerated wafer scanning method, and relates to the technical field of semiconductor wafer scanning, and the method comprises the steps: obtaining the identification information of a target wafer transmission box when the target wafer transmission box is detected; detecting the in-place state of the wafer through a light curtain sensor, and generating a slot mapping table; controlling a preset micro-rotation mechanism to carry out micro-rotation on the wafer at each slot position within a set angle range, and collecting an annular structure light image; constructing polar coordinates and calculating a first attitude angle of each wafer; driving a micro-rotation mechanism to adjust the wafer to a preset reference angle according to the first attitude angle to obtain a standard attitude wafer; the mechanical arms are driven to be sequentially conveyed to a scanning station; respectively controlling a rotary positioning module to rotate in a first angle range and a second angle range, and acquiring a coded image and a spectral image; the wafer code is analyzed, confidence verification is carried out, and a wafer process type label is identified; and warehousing and recording the wafer codes, the process labels and the transmission box identification information. According to the invention, the storage management efficiency of the regenerated wafer can be effectively improved.

The embodiment of the invention discloses an MEMS structured light high-precision three-dimensional measurement method, device and equipment and a medium. The method comprises the following steps: determining first orthogonal phase distribution information of a to-be-measured target according to a surface stripe image; determining a first space linear equation of the fitting projection light of the preset projection phase; coordinates of the first space linear equation are converted to obtain a second space linear equation, a projection device coordinate system is established according to the first space linear equation, and a mapping model of the deviation angle and the projection phase is determined based on the corresponding relation between the second space linear equation and the preset projection phase; mapping the phase to obtain a first offset angle; and determining the to-be-measured target point cloud based on the first deviation angle to obtain a three-dimensional coordinate. According to the invention, the method does not depend on a high-precision calibration plate, builds the incidence relation between the phase and the angle and the ray mapping model, accurately represents the projection characteristics of the projection device, and solves a technical problem that the precision is not high when three-dimensional measurement is carried out based on an MEMS structured light system.

The invention discloses a method for generating a staggered superposition double-layer Airy light beam array capable of realizing double focusing, and belongs to the technical field of structured light regulation and control and special light beam generation. Aiming at the problems that the focusing characteristics of light beam arrays in transmission in different superposition modes are difficult to fully reveal in the prior art and the generation of an optical bottle depends on complex phase modulation or multiple optical elements, the method comprises the following steps: constructing a first layer of Airy light beam array and setting regulation and control parameters, constructing a second layer of Airy light beam array with staggered azimuth angles and setting regulation and control parameters, and constructing a second layer of Airy light beam array with staggered azimuth angles; a longitudinal double-focusing structure is realized through staggered superposition of double-layer Airy light beam arrays, and a vortex phase is introduced into double-focusing array light beams to form an optical bottle structure. According to the invention, the structure is simple, the longitudinal double-focusing and optical bottle structure can be realized only through the initial light field structure design, the focusing intensity and the regulation freedom degree are high, and the application of the light beam in the related fields of microscopic imaging, particle capture and the like is expanded.

The invention discloses a global path planning method for a power converter valve inspection operation robot, and the method comprises the steps: obtaining the spatial point cloud data of a converter valve region through a three-dimensional structured light camera, and constructing a two-dimensional environment map containing obstacle expansion information through projection and rasterization; generating a snakelike or spiral full-coverage global theoretical path according to the scanning step length of the camera; in the inspection execution process, the feasibility of the path section is judged through discrete sampling and continuous collision detection; local obstacle avoidance adopts an improved A * algorithm, forward and backward bidirectional search is simultaneously executed in a limited local search area, and a dynamic heuristic function of a coverage potential item and a safety score item is introduced. And finally, splicing the local path and the theoretical path to form a global feasible inspection path. The method effectively solves the problem that the coverage rate and the obstacle avoidance efficiency are difficult to consider in a complex converter valve hall environment in a traditional method, and is particularly suitable for automatic inspection of equipment-intensive scenes such as an electric power converter valve hall.

The invention relates to the technical field of underwater target detection, and discloses an underwater target detection device and method based on binocular fusion line structured light, the device comprises a binocular structured light system and an information processing unit which are installed at the bow of an underwater robot, and the binocular structured light system is in communication connection with the information processing unit; the binocular structured light system comprises vertically arranged array laser and an underwater binocular camera, and the array laser is used for projecting line structured light stripes to an underwater target and forming a structured light image on the surface of the target; the underwater binocular camera is used for capturing a structured light image and sending the structured light image to the information processing unit; and the information processing unit is used for performing edge detection on the received structured light image to obtain line structured light projection stripes of the target surface, and further obtaining three-dimensional point cloud coordinates of the target surface. According to the invention, an underwater three-dimensional space can be fully covered, and three-dimensional perception and accurate identification of underwater suspended and semi-buried targets can be realized.

The invention discloses a printed pattern edge sharpening processing control and hardware imaging optimization system, and relates to the technical field of printed pattern imaging processing, and the system comprises a fragile feature pre-analysis module which recognizes the boundary of a fine line and a high saturation color block, and generates a feature map; the ambient light acquisition module obtains illumination parameters and positions a strong light reflection area; the structured light compensation imaging module projects adaptive structured light to optimize imaging; the texture layering module divides and marks image layers; and the layered sharpening module carries out differentiated sharpening on each image layer according to the ambient light parameters and outputs a high-quality image. According to the method, the fragile characteristics of the printing are accurately identified, and imaging optimization is carried out in combination with ambient light changes, so that edge sharpening and imaging quality collaborative improvement are realized; based on feature analysis and light parameter acquisition, an imaging strategy adaptive to a scene is constructed, layered differentiation sharpening is performed, accurate reduction of patterns is guaranteed, and high-precision printing detection and display requirements are met.

A method of overhaul is provided. During this overhaul method, a substrate is scanned using structured light to provide substrate scan data. The substrate is from a component previously installed within an engine. The substrate scan data is compared to substrate reference data to provide additive manufacturing data. Material is deposited with the substrate using an additive manufacturing device based on the substrate scan data to provide a first object. The first object is scanned using the structured light to provide first object scan data. The first object scan data is compared to first object reference data to provide machining data. The first object is machined using the machining data.

The invention discloses a visual inspection sorting method and system for surface defects of intelligent electric meter shell injection molded parts, and relates to the technical field of visual inspection, and the method comprises the following steps: selecting an intelligent electric meter shell as a reference part, dividing the surface of the shell into a plurality of functional partitions, collecting structured light form and brightness information, and obtaining a reference appearance map; and positioning a to-be-detected electric meter shell to a position consistent with the reference part, obtaining the online three-dimensional form and brightness response of each functional zone, mapping the online three-dimensional form and brightness response to the reference appearance map, marking a stable mirror surface with high brightness, and interpolating along an observation track to generate an online appearance map. According to the invention, through functional partition sampling layout of a window sealing area, a wiring terminal area, a stud and reinforcing rib area, a nameplate and two-dimensional code area and a general decoration side wall area, and in cooperation with double-light-source collection of diffused light and oblique incident light combined with structured light, a reference appearance map containing a highlight tolerance mask is constructed; stable characterization of mirror reflection and a transparent window is realized, and false detection caused by high brightness of the mirror surface is remarkably reduced.

The invention discloses a surface defect detection system based on a microcrystalline glass panel finished product, and relates to the technical field of surface defect detection. The defect full-dimensional dynamic state map construction module is used for constructing and updating a defect full-dimensional dynamic state map containing defect three-dimensional coordinate information, multi-modal feature confidence information, virtual-entity parameter deviation information and feature definition-registration precision linkage threshold information in real time; the structured light three-dimensional contour acquisition module is used for acquiring three-dimensional contour data of the surface of the microcrystalline glass panel; the multi-modal feature fusion module is used for collecting two-dimensional image data of multi-light-source imaging and extracting multi-modal feature information; the digital twin virtual debugging module is used for constructing a virtual detection environment based on the defect full-dimension dynamic state graph; the problem of feature confusion caused by light source combinational logic solidification in an existing multi-light-source switching system is effectively solved.

The invention discloses a grasp point pose calculation method based on a geometric clustering algorithm without deep learning and a GPU (Graphics Processing Unit) and application of the grasp point pose calculation method. The grabbing point pose calculation method comprises the steps that 100, three-dimensional point cloud data of a to-be-grabbed object are obtained through an industrial-grade structured light three-dimensional camera, a geometric clustering algorithm is adopted for screening the three-dimensional point cloud data, and a unique target point cloud cluster conforming to the target form and size is extracted; step 200, performing rough matching on the target point cloud cluster and the template point cloud through the FPFH feature vector and RANSAC to obtain a rough matching result; step 300, performing fine matching on the target point cloud cluster and the template point cloud through the rough matching result and the GICP to obtain a fine registration result; step 400, judging whether a precise registration result meets a preset requirement or not according to the interior point root-mean-square error and the overlap ratio, if so, entering the next step, and otherwise, terminating the process; and 500, according to the fine registration result and coordinate transformation, the grabbing point pose of the to-be-grabbed object is obtained through calculation.