38 results about "Surface matching" patented technology

Systems and methods for performing surface matching using three-dimensional (3D) and color information. An example method includes obtaining, by a 3D camera, a first 3D image of a field of view. The 3D image includes (i) a plurality of voxels, or 3D points, of the field of view and (ii) color information of the field of view. A processor normalizes the color information into a common color space, and determines scene histograms from the normalized color information. Each of the scene histograms is determined for a voxel of the plurality of voxels. The method further includes determining a color score from at least the scene histograms and model histograms, the model histograms being indicative of color information of a model image of a 3D object, and determining the presence of a surface of an object in the first 3D image from the color score.

The technology relates to tracking objects in an environment around an autonomous vehicle. A computing system of the autonomous vehicle determines accurate motion characteristics of objects detected in its environment despite various sensor measurement limitations. By correcting motion distortion for fast moving objects and accounting for discrepancies in sensor data gathering, motion characteristics may be determined for the detected objects with enhanced accuracy. Multiple sets of correspondences are determined for clusters from multiple sensor spins, enabling better alignment using a surface matching algorithm even when clusters have fewer data points. Efficiency is also enhanced by selecting hypotheses based on confidence levels. These techniques provide for identifying the types of objects for which a yaw rate can be accurately determined. Object classification can also be improved by accumulating associated clusters corresponding to a detected object. In addition, under- or over-segmentation can be mitigated with such techniques.

This invention discloses a jet polishing method and apparatus for complex curved surfaces, comprising the following steps: Step S1: acquiring processing parameter information; Step S2: performing simulated polishing on the workpiece according to the parameters set in Step S1, and selecting the optimal removal function based on the simulation results; Step S3: performing formal polishing on the workpiece according to the optimal removal function described in Step S2; when switching to polishing a processing surface with different curvatures, changing the polishing nozzle with a different outlet diameter; and matching a new optimal removal function; changing the nozzle with a different outlet diameter can adapt to the operating space of polishing areas of different sizes on complex curved surfaces, improving the overall polishing efficiency; at the same time, based on the relationship between the optimal removal function matched on the previous processing surface and the optimal removal function matched on the switched surface, controlling the polishing nozzle to move to a new starting point before starting polishing; eliminating the problem of seams between adjacent areas.

The invention discloses an intelligent construction method and system for a steel web composite structure based on a space coordinate system, and the method comprises the steps: enabling a first wave plate to be orthogonally projected into a plane wave curve, building a space coordinate system, drawing up a cutting function and a cutting angle, and obtaining a reference cutting point and a target cutting point; the first waved plate is cut, and a contact surface curve space is generated to be in butt joint with the steel pipe in a matched mode; by establishing a space coordinate system and deriving a cutting function, mathematical modeling and a physical cutting process are combined, and high-precision cutting of the waved plate is achieved. Dynamically determining a reference cutting point and a target cutting point through a cutting function according to factors such as amplitude and wavelength of the waved plate, and ensuring that a contact surface curve space meets geometric requirements; a deformation compensation coefficient is introduced, particularly in complex curved surface cutting, curve deformation and adaptive adjustment after cutting are indirectly compensated, the contact surface matching problem is solved through dynamic matching of a coordinate system and cutting parameters, and the cutting precision is improved.

This invention discloses a deterministic shaping device and method for large and complex surfaces based on adaptive surface matching. The device includes a rotary drive assembly, a revolution-rotation transmission housing, and a shaping disk assembly, mounted from top to bottom. The revolution-rotation transmission housing contains a revolution shaft and a rotation shaft. The shaping disk assembly includes a tool disk and a shaping film mounted on the lower side of the tool disk. The output shaft of the rotary drive assembly is connected to the revolution shaft, and the rotation shaft is connected to the eccentric side of the tool disk. The revolution-rotation transmission housing contains a pressure component that applies pressure to the tool disk via the rotation shaft. This invention's deterministic shaping device for large and complex surfaces based on adaptive surface matching has a simple structure and can be easily installed as a modular unit on a CNC machine tool. It achieves shaping of large and complex surfaces through machine tool servo motion, offering good operational flexibility and spatial accessibility.

The invention discloses a non-standard workpiece machine tool positioning method and device based on multi-feature cooperative matching, and relates to the technical field of machine tool machining and positioning, and the method comprises the steps: recognizing a plurality of workpiece feature elements of a non-standard workpiece; acquiring a clamp assembly and a machining assembly of the machine tool; identifying clamp interaction features of the clamp assembly and tool machining features of the machining assembly; constructing a reference surface matching model; the multiple workpiece feature elements, the clamp interaction features and the tool machining features are input into a reference surface matching model, all reference surfaces of the non-standard workpiece are recognized, and a positioning reference surface is output; and the machine tool performs positioning operation on the non-standard workpiece by using the positioning reference surface. The technical problem that due to the diversity of the structure, geometry and surface characteristics, it is difficult to accurately determine the positioning datum plane, and then the machining precision and efficiency are affected is solved, and the effect of improving the machining precision and machining efficiency of the non-standard workpiece through matching of geometry compatibility and mechanical stability is achieved.

The disclosed method encompasses a computer-implemented medical method for surface matching registration of an object in image guided navigation. In particular, the method provides an identification of an optimal matching solution from a plurality of similar matching solutions comprising the lowest distance matching solution based on at least one matching parameter. This method allows finding a better optimal matching solution.

The invention provides a checking method and device for a conical surface matching structure and electronic equipment, and is applied to the technical field of automobile chassis. According to the technical scheme, firstly, size data, mainly including various characteristic parameters of a taper pin structure and a taper hole structure, of a conical surface matching structure are obtained; checking is carried out based on the size data of the structure, the compressive strength and the settling volume, and insufficient strength or stress concentration caused by size deviation is avoided. And then the pressure stress, the tensile stress and the settling volume borne by the taper hole structure are checked based on the size data, it is ensured that the conical surface matching structure meets the stress requirement, and the situation that due to the unreasonable conical surface matching structure, the conical surface matching structure is affected by external loads and the like, connection of the conical surface matching structure fails is avoided.

The technology relates to tracking objects in an environment around an autonomous vehicle. A computing system of the autonomous vehicle determines accurate motion characteristics of objects detected in its environment despite various sensor measurement limitations. By correcting motion distortion for fast moving objects and accounting for discrepancies in sensor data gathering, motion characteristics may be determined for the detected objects with enhanced accuracy. Multiple sets of correspondences are determined for clusters from multiple sensor spins, enabling better alignment using a surface matching algorithm even when clusters have fewer data points. Efficiency is also enhanced by selecting hypotheses based on confidence levels. These techniques provide for identifying the types of objects for which a yaw rate can be accurately determined. Object classification can also be improved by accumulating associated clusters corresponding to a detected object. In addition, under-or over-segmentation can be mitigated with such techniques.

A computer-implemented medical method for surface matching registration of a patient (O) in image guided navigation comprises the steps of obtaining (81) registration data (Dr) of a surface of a patient (O), wherein the registration data (Dr) comprises a plurality of registered points (P), determining (S2), for each of the plurality of points (P) a matching weight (W) based on at least one quality criteria value (Q1, Q2, Q3), wherein the matching weight (W) reflects a quality of the respective point (P) and performing (S3) a matching algorithm between image data (Di) of at least part of the patient (O) and the registration data (Dr), using the matching weight (W) of each of the plurality of points (P), thereby determining a matching solution (Ms). Thus, for each point a quality can be determined that is considered in the matching algorithm. In other words, the matching weight (W) leads to a more accurate intraoperatively acquired point cloud and consequently to an improved matching accuracy.

The embodiments of the present application disclose a design surface data processing method, device, equipment and computer-readable medium, and relate to the field of computer technology. The method includes: detecting a design surface target area selected by a user on a design surface to be extended; based on the target point position input by the user, generating an offset curve that matches the boundary of the design surface target area; wherein the offset curve includes multiple offset control points; based on the direction of the boundary of the design surface target area in the design surface, adding multiple offset control points corresponding to the offset curve to the control point matrix to obtain an updated control point matrix; based on the updated control point matrix, generating an updated design surface, the updated design surface includes an extended surface that matches the design surface target area. Through this method, the computational complexity of the computer is reduced, the generation efficiency of the extended surface is improved, and thus the efficiency of the design surface data processing is improved.

The invention provides a segment detection method and system based on a two-dimensional slide rail and binocular structured light, and the method comprises five steps: the five steps automatically complete full-coverage scanning through the two-dimensional slide rail, improve the measurement efficiency, effectively solve the problem of weak texture surface matching in combination with a binocular light sensing unit and coding structured light, and improve the detection precision. Through automatic splicing of system calibration, accumulated errors are eliminated, a complete three-dimensional model is obtained, and objective and accurate three-dimensional data support is provided.

Disclosed are a spine registration guide plate, a manufacturing method, and a model construction method and apparatus. The present invention relates to the technical field of computer-aided modeling of surgical instruments. The model construction method comprises: importing a three-dimensional spine model and an original spine surface matching model; adjusting a pose of the original spine surface matching model, enabling the original spine surface matching model to be in a relative pose relationship matched with a surface contour of the three-dimensional spine model, and recording a final pose of the original spine surface matching model at the moment; performing Boolean operation on the three-dimensional spine model and the original spine surface matching model to obtain a final spine surface matching model; importing a visual navigation combination model, and adjusting the visual navigation combination model to the final pose; combining the visual navigation combination model with the final spine surface matching model to generate a spine registration guide plate model. The present invention addresses the high-quality requirements of traditional Boolean operations for three-dimensional models, reduces Boolean operation error rates, and reduces the computational load for generating the guide plate model.

This application belongs to the field of nuclear industry robot technology and aims to solve the problem of extrinsic parameter calibration of monocular cameras and lidar on large equipment. This application discloses a single-line lidar and camera calibration method based on point, line, and surface matching. Based on the multi-plane intersection features in the real environment, a visual target is pasted on the intersecting plane. Single-line laser point and line features, and line and surface features of the monocular camera based on the visual target are extracted. Then, the extracted laser point and line segment features are matched with spatial planes and lines to construct error equations for the distance from a spatial point to a line and the distance from a line segment to a spatial plane. Finally, the extrinsic parameters of the lidar and monocular camera are optimized and solved using a nonlinear optimization method. This application can achieve extrinsic parameter calibration of monocular cameras and lidar on large equipment, and can obtain higher accuracy and more stable calibration results.

A method comprising obtaining a first 3D image of a field of view by a 3D camera wherein the first 3D image comprises a plurality of 3D points of the field of view and colour information of the field of view. A processor determines a plurality of voxels of the 3D image wherein each voxel comprises one or more of the 3D points. The processor compiles scene histograms through normalising the colour information to a common colour space which are determined for a voxel of the plurality of voxels of the first 3D image. The processor further determines a colour score by comparing the scene histograms to model histograms having colour information of a model image of a 3D object. The presence of an object surface in the first 3D image is determined by the colour score. The method may be suitable for performing three-dimensional surface matching. Determining the colour score may comprise comparing corresponding voxels of the first 3D image and voxels of the model image by a weighted sum; 3D transformation may be performed on the first 3D image or the model image. A colour histogram may be determined for each voxel of the first 3D image. The colour score may be compared to a threshold score when determining the presence of an object surface. A second 3D image may be obtained and processed. A system and computer-readable instructions to be performed on a processor are also claimed.

The invention discloses a landslide risk prediction method based on a landslide disaster assessment model, and belongs to the technical field of image processing. The method comprises the following steps: firstly, extracting a remote sensing image of a to-be-monitored area, obtaining a channel earth surface matching value and a channel soil matching value, and constructing R, G and B channel soil earth surface comparison maps; obtaining a plurality of adjacent moment comparison diagrams, obtaining soil surface characteristic values according to the change of the soil surface comparison value, and constructing a characteristic diagram; then constructing a cumulative landslide weight matrix according to the landslide weight coefficients of the suspected variable pixel points, and constructing a cumulative soil surface feature map according to the soil surface feature values; and finally, processing the accumulated landslide weight matrix and the accumulated soil surface feature map corresponding to the three channels by adopting a landslide disaster assessment model to obtain a landslide risk score. The accuracy of landslide risk prediction is improved.

The invention discloses a random response surface selection and distribution point optimization method and related equipment. The method comprises the following steps: generating a to-be-selected and distribution point set according to a power distribution network; selecting a plurality of matching points from the to-be-selected matching point set as the input of a Hermite chaotic polynomial, and establishing a constructor function of a Hermite coefficient matrix related to the matching point combination; taking the collocation combination as an optimization variable, substituting the collocation combination into a constructor function to obtain # imgabs0 #, establishing a constraint condition that a first row of elements of # imgabs1 # are non-negative values, and establishing an optimization target that a minimum value of the first row of elements of # imgabs2 # is maximum; and based on the constraint condition and the optimization target, performing optimization by adopting an ant colony algorithm to obtain an optimal collocation combination. According to the method, the minimum value of the first row of elements of the # imgabs3 is maximized, so that sensitivity distribution is more reasonable, the first row of elements of the # imgabs4 are strictly ensured to be non-negative through constraints, and physical significance deficiency caused by negative sensitivity coefficients is avoided.

The invention particularly discloses a complementary surface parallel matching method and system based on three-dimensional topology and a medium, and belongs to the technical field of precise surface matching. According to the method, a matched surface is fixed to a multi-axis adjusting mechanism, and three-dimensional shape data of the matched surface are collected by combining an optical scanning module and a double-laser displacement sensor; and constructing a topological map. And in the data processing platform, through feature point extraction and block height difference calculation, in combination with a concave-convex complementary relationship, a multi-axis adjusting mechanism is driven to carry out dynamic adjustment, and high-precision surface alignment under the condition of no preset mark is realized. The method has good adaptability to irregular concave-convex surfaces, the matching precision is smaller than or equal to 0.1 micrometer, the concave-convex area matching rate is larger than or equal to 95%, and the method is suitable for the technical field of high-precision assembly assembling.

A 3D navigation system for surgical and procedural instrument position tracking, according to one embodiment of the present invention, comprises: a camera, which photographs an instrument for a surgery and a procedure of a patient so as to obtain a camera image; a skin surface extraction unit for generating, from an MRI image of the patient, a 3D point cloud having a first surface point corresponding to the skin surface of the patient; a landmark matching unit for calculating the position correspondence relationship between the MRI image and the camera image on the basis of landmark matching, and then matching same so as to generate a first matching image; a surface point acquisition unit for extracting a second skin surface point on the basis of a physical marker of the patient in the camera image; a surface matching unit for calculating, on the basis of surface matching, the position correspondence relationship between the 3D point cloud having the surface point and the camera image from which the second skin surface point has been extracted, and then matching same so as to generate a second matching image; and a coordinate conversion unit for finely adjusting the first matching image by matching, through a conversion matrix calculation method, the first matching image and the second matching image, which are different from each other.

The utility model relates to the technical field of telemetering cover plate detection tools, and discloses a telemetering cover plate bevel edge included angle detection device, which comprises a base, the upper end of which is provided with a supporting curved surface matched with the arc-shaped bottom surface of a telemetering cover plate; the bevel edge included angle detection assembly comprises a go gauge and a no-go gauge, and the detection bevel included angle of the go gauge is smaller than the detection bevel included angle of the no-go gauge; the screw hole detection assembly comprises four detection holes which are symmetrically formed in the base; according to the utility model, through collaborative design of the bevel edge included angle detection assembly and the supporting curved surface, efficient and accurate detection of the bevel edge included angle of the telemetering cover plate is realized, the assembly problem caused by angle out-of-tolerance in a traditional detection method is avoided, and the detection accuracy and the batch detection efficiency are significantly improved; according to the screw hole detection assembly, detection holes which are symmetrically distributed and a round rod are inserted, and integrated detection of the position precision of the counterbore is achieved through the aperture difference design and the axis common center characteristic.

The present technology relates to tracking objects in an environment around an autonomous vehicle. Despite various sensor measurement limitations, a computing system (110) of an autonomous vehicle determines accurate motion characteristics of objects detected in its environment. By correcting for motion distortion of fast-moving objects (808) and accounting for differences in sensor data collection, the motion characteristics of the detected objects can be determined with enhanced accuracy (908). Multiple sets of correspondences are determined for clusters from multiple sensor spins, enabling better alignment using a surface matching algorithm even when the clusters have fewer data points (904). Efficiency can also be improved by selecting hypotheses based on confidence levels. These techniques provide for identifying types of objects for which yaw rates can be accurately determined. Object classification can also be improved by accumulating associated clusters corresponding to detected objects (186). In addition, such techniques can be used to alleviate under-segmentation or over-segmentation.

The invention relates to the technical field of automobile manufacturing, in particular to an identification universality judgment method and device based on profile matching, electronic equipment and a storage medium, and the method comprises the steps: obtaining a solid three-dimensional model of an identification to be borrowed, and extracting an adhesive layer reference surface from the solid three-dimensional model; and on the basis of a preset space matching rule, carrying out pose alignment on the adhesive layer reference surface and the theoretical bonding surface of the target component in a three-dimensional space. According to the method, a traditional low-efficiency process which depends on artificial experience, real object trial assembly and repeated adjustment is converted into a standardized digital process which can be completed in a computer environment within a few minutes, so that the judgment efficiency and the automation level are remarkably improved, the objectivity and the accuracy of a judgment result are ensured, and the judgment efficiency is improved. Therefore, the repeated development cost and the subsequent quality risk of the mold caused by misjudgment are effectively avoided.

The invention relates to the technical field of performance optimization, in particular to a performance optimization method and system for a unit type modular roofing, and adopts the technical scheme that connection effect analysis comprehensively considers multiple factors of connection surface matching and damage, and waterproof performance evaluation is combined with damage influence and connection influence coefficients; environmental use anomaly analysis covers multiple environmental factors and considers the amplification effect of connection effect anomaly, waterproof anomaly and reflection deviation degree can be quantified when the waterproof coating thickness is finally determined, a repair scheme can be customized according to the initial waterproof condition of each area, excessive or insufficient repair is avoided, and the repair efficiency is improved. The waterproof performance of the roof is comprehensively improved; and the cost is optimized.

This invention provides a drone VR landscape construction system based on spatial computing, comprising: a digital twin reference construction module for constructing a digital twin as an absolute spatial reference; a drone edge SLAM module for generating a sparse point cloud containing three-dimensional coordinates; a point-to-surface alignment correction module for continuously aligning the sparse point cloud to the surface of the digital twin using a point-to-surface matching algorithm to eliminate cumulative drift during drone operation; a virtual reality interaction module for displaying the reconstructed landscape scene from a first-person perspective and capturing the operator's gaze hotspots in real time; a dynamic re-collection feedback module for converting gaze hotspots into regions of interest and generating re-collection tasks to drive the drone to dynamically re-collect and verify the regions of interest; and an output module for generating a spatially computed, drift-free, and quality-labeled 3D VR landscape model. This invention aims to achieve drift-free, real-time human-machine collaborative high-fidelity VR landscape construction.

The application discloses a game resource production method and device, a storage medium and a terminal, relates to the technical field of data processing, and mainly aims to solve the problem that the rain and snow game resource produced by direct covering cannot achieve the display effect of real rain and snow covering and the accuracy of simulating a real scene in game resource production is reduced. The method comprises the following steps: creating a local area surface matching at least one preset inclination angle and / or horizontal angle of a first target element; generating a second target element on the local area surface; and baking the second target element to the first target element according to the normal direction of the local area surface, so as to obtain a game resource. The method is mainly used for game resource production.

Systems and methods for performing surface matching using three-dimensional (3D) and color information. An example method includes obtaining a first 3D image of a field of view by a 3D camera. The 3D image includes (i) a plurality of voxels or 3D points of the field of view and (ii) color information of the field of view. A processor normalizes the color information into a common color space and determines scene histograms from the normalized color information. Each of the scene histograms is determined for a voxel of the plurality of voxels. The method further includes determining a color score from at least the scene histograms and the model histograms, the model histograms indicating color information of a model image of a 3D object, and determining the presence of a surface of an object in the first 3D image from the color score.

Methods and systems for determining a corrective stage in orthodontic treatment. Methods may include presenting a menu for a user to enter a pre-determined tolerance or a maximum number of iterations for performing a surface matching algorithm. Methods may include matching one or more teeth of a previously segmented teeth model to one or more corresponding teeth of a current teeth image. The matching may include executing the surface matching algorithm. The methods may include iteratively finding closes sample points, determining whether distances between closest sample points exceed a predetermined tolerance, and adjusting the distances. A current teeth model may be generated based on the adjusted previously segmented teeth model. A current teeth model using the corrective stage may be generated.

The invention discloses an error-controllable curved surface matching generation method based on a Trimmed curve and a lofting method thereof, and relates to the technical field of three-dimensional modeling, two curved surfaces S1 (u1, v1) and S2 (u2, v2) are given, matching is carried out by modifying a cutting curve, namely the Trimmed curve, on the curved surfaces S1 (u1, v1) and the curved surfaces S2 (u2, v2), the matched curved surfaces completely coincide with the curved surfaces S1 (u1, v1) and are spliced with the cutting curve on the curved surfaces S2 (u2, v2), and the three-dimensional modeling is carried out. G0, G1 and G2 are continuous at the joint. According to the method, the high-quality matching curved surface can be generated within the preset splicing error threshold value, it is ensured that the generated matching curved surface has high fairness and quality, compared with industrial standard software CATIA software, the matching curved surface generated through the method can completely coincide with the reference curved surface while the quality of the matching curved surface generated through the method is kept the same, and the matching effect is good. And subsequent adjustment and modification are easier. The method can also be applied to lofting, and a lofting curved surface with controllable errors is obtained.

The invention relates to the technical field of tunnel quality evaluation, in particular to a multi-view three-dimensional reconstruction and streamline curved surface matching tunnel clearance detection method. Comprising the steps of high-precision calibration plate arrangement, image acquisition, three-dimensional reconstruction, design contour construction, normal vector constraint registration and curved surface matching. The method ensures that the photo overlapping rate is greater than or equal to 80%, and eliminates the manual shooting jitter error; the three-dimensional reconstruction precision is improved; rapid matching and deviation statistics of a designed curved surface and an actually measured curved surface are realized; the system supports one-key operation from image acquisition, three-dimensional reconstruction to back break report generation, and is compatible with open source and commercial software such as Visual SFM, PhotoScan and the like.

The application discloses a multi-target tracking method based on point-face matching association, comprising the following steps: acquiring a video to be detected and performing serialization, constructing a detector according to a feature extraction network, sequentially performing target detection on video frames in the video frame sequence according to the detector, obtaining a tth video frame and a t+1th video frame from the video frame sequence, dividing targets in the t+1th video frame into high-confidence targets and low-confidence targets, performing first association matching on the high-confidence targets according to a target frame similarity perception algorithm, adding matched targets into target tracks of the tth video frame, initializing target tracks of the targets which fail to match, respectively, performing second association matching on the low-confidence targets, adding matched targets into target tracks of the tth video frame, discarding the targets which fail to match, and sequentially traversing all the video frames until all the target tracks are obtained. The detection confidence of the occluded target is improved, and the accuracy of multi-target track association is improved.