45 results about "Real size" patented technology

The invention relates to the technical field of machine vision, in particular to a precision part size automatic detection method and system based on machine vision, precision part images are collected through a high-precision industrial camera, part positioning is carried out, sub-pixel-level topological feature mapping is carried out on interested area images, and precision part size automatic detection is carried out. Comprising the steps of gray histogram equalization, gray morphological processing, edge detection and edge chain code tracking, construction of an edge point topological feature space, execution of sub-pixel subdivision, obtaining of an edge line through contour analysis of a feature distance and a feature angle, and double-constraint geometric reconstruction based on the edge line. The characteristic distance and the characteristic angle are used for rotation matrix conversion and geometric dimension calculation, a relation model of the geometric dimension and the actual dimension of the part is established, precise part dimension measurement is achieved through dynamic error analysis and compensation, the measurement precision is remarkably improved, the risk caused by unreliability of a single characteristic is effectively reduced, and the measurement accuracy is improved. And the measurement stability is improved.

A model and its establishment method and system, a compensation method, equipment and a storage medium, the establishment method comprises the following steps: providing a test layout, including a test pattern, including a plurality of test points; obtaining the initial pattern density of each test point according to the test pattern; taking any test point as a feature point, obtaining the effective pattern density of the feature point according to the initial pattern density of each feature point and the influence of the initial pattern density of the plurality of test points located around the feature point on the feature point, wherein the influence degree of the initial pattern density of the plurality of test points around the feature point on the feature point is set anisotropically; obtaining the actual size deviation corresponding to the feature point; obtaining the estimated size deviation of the feature point; judging whether the estimated size deviation meets the preset condition according to the estimated size deviation and the actual size deviation; when the estimated size deviation meets the preset condition, the establishment of the etching deviation compensation model is completed. The etching deviation compensation model is more accurately established.

The invention discloses a large building concrete column radius construction measurement control method, and relates to the technical field of linear dimension metrology, and the method comprises the specific steps: parameter collection, data collection, model construction, real-time correction and strategy adjustment. Corresponding sensing equipment is arranged in a construction area to collect related associated data and form a real-time parameter set, meanwhile, original historical data of projects of the same type are collected and screened and classified to form a historical parameter error set, so that a linear size error correction model is constructed, influence rules of various associated data are determined, and the accuracy of the linear size error correction model is improved. And substituting the real-time parameter set into the linear size error correction model to obtain a radius measurement error compensation value, and dynamically correcting original radius measurement data, so that historical engineering experience and real-time field information are fully fused in the measurement process, multiple types of influence factors are accurately counteracted, and the measurement accuracy is improved. And the measurement result can truly reflect the actual size state of the concrete column body.

The invention discloses a building model inner path planning method and system based on IFC embedded information, and belongs to the technical field of building model application, and the method comprises the steps: 1, building a building model and embedded node information; a step; 2, exporting a file which contains embedded node information and is in an IFC format; step 3, uploading and primarily processing the IFC file; step 4, IFC file analysis and data extraction; 5, processing and formatting the analysis data; step 6, constructing a reachability network diagram; 7, the user selects a starting point, an ending point and the size of a transported object; step 8, searching a path based on size limitation; step 9, outputting a path result; step 10, carrying out path visualization; according to the building model internal path planning method and system based on IFC embedded information provided by the invention, the problems that data and a BIM model are disjointed in an existing building internal path planning method, and actual size limitation and data maintenance invariability are difficult to effectively consider are solved.

The invention relates to the technical field of high-precision assembly of an LED die bonding process, in particular to a steel mesh and PCB size matching dynamic mapping method and device, and solves the problem of matching failure caused by size change of a steel mesh. The method comprises the following steps of: selecting a PCB (Printed Circuit Board) product of which the stability meets a preset condition after solder paste printing is completed from the current production, and calculating by extracting solder paste images at four corners of the PCB product to obtain the actual size data of the current screen; matching a to-be-matched PCB with the screen plate based on the actual size data of the current screen plate; and when a preset screen data updating condition is met, re-selecting the printed PCB product meeting the preset condition from the current production, extracting the solder paste images at the four corners of the PCB product to update the actual size data of the screen, and continuing to match the subsequent to-be-matched PCB with the screen based on the updated actual size data.

The utility model discloses the technical field of machining measurement device, more specifically discloses a kind of for the size measurement tool of incomplete circle, comprising: tool main body, detachably set on part;Compensation component, set on tool main body, can be synchronous with the arc to be processed Machining;Wherein, the straight line of the center of the arc to be processed of compensation component and arc to be processed can pass through arc to be processed;Measuring component is used to measure the diameter between compensation component and arc to be processed;By using the utility model, arc can be measured in real time during processing, cancel the use of radius gauge and three-dimensional coordinate instrument;Greatly reduce the difficulty of real-time measurement arc actual size in part processing process, improve the arc size measurement stability and accuracy, can effectively improve the product quality and stability of part.

The invention discloses a product quality detection method capable of adaptively adjusting a feature fusion layer. The method specifically comprises the following steps: collecting a data set containing targets of various sizes; selecting two target detection networks with the same architecture as a basic model; wherein a layer of check is added to one model for enhancing detection of a small target; pre-training the two models respectively; putting each image into the two models for detection, and respectively obtaining results with relatively high detection precision and selection accuracy; and recording the corresponding model, determining the size information of each data in the data set, and obtaining a feature size interval for detecting the two models according to the precision comparison. According to the method, by introducing a target size identification and self-adaptive adjustment strategy, the number of feature fusion layers can be dynamically adjusted according to the actual size of the target in the input picture, and the optimal detection effect can be achieved under the targets of different sizes.

A method for determining estimated real dimensions of an elevator car by: acquiring a virtual 2-dimensional map of the car based on data provided by a lidar sensor, the map representing reflection surfaces from which laser light emitted by the sensor is reflected; defining a rectangular starting area at a position located within the real dimensions of the car; successively extending the area by one pixel in each of four extension directions orthogonal to edges of the area, and when any pixel in one of the extension directions corresponds to one of the reflection surfaces in the map, stopping extending the area in this extension direction and successively extending the area by one pixel in each of the remaining extension directions the area is extended to a maximum in all the extension directions; and determining the fully extended rectangular area as the estimated real dimensions of the car.

The invention relates to a visual dimension conversion method, which is characterized by comprising the following steps of: 1) arranging a plurality of longitudinal and transverse scales on a mold table in a vision field of a camera, and arranging scales on the scales; (2) placing an object to be measured on a mold table, and then taking a picture by a camera; and 3) taking a picture, and comparing the to-be-measured object in the picture with the nearest scale to obtain a real size. According to the method, a picture object is compared with the nearest scale, the scale is enabled to distort along with the lens, so that the correct size of the object in the image can be obtained even if the object distorts, a distortion correction algorithm is not needed through the method, the method can be applied to a large image scene, a complex algorithm is abandoned, and the method is easy to implement. And the obtained numerical value is more reliable.

The invention provides a method for restoring internet map vector tile data to construct a real geographic position three-dimensional scene, computer equipment and a computer program product, is suitable for the technical field of model modeling, and restores real geographic coordinates by reversely calculating a matrix coordinate range of tiles, so that the real geographic position three-dimensional scene is constructed. Offset errors generated by encryption of a map service provider are eliminated; splicing the fractured vector elements in a cross-tile manner, and recovering the original topological integrity of the fractured vector elements; real parameters are reversely mapped based on semantic tags, and high-precision three-dimensional modeling is driven; and based on the target geographic coordinate system, generating and outputting the three-dimensional model of the real size, and ensuring that the three-dimensional scene is strictly aligned with the real world position. Therefore, the problem of geographic information distortion caused by slicing, offset and simplification of the Internet vector tile data is solved. Finally, a high-fidelity three-dimensional geographic scene is automatically constructed at low cost, and professional application is supported.

The invention discloses a target object measurement method based on monocular vision, and relates to the technical field of target object measurement, and the method comprises the steps: carrying out the Gaussian filtering and Canny edge extraction of an image, and obtaining an angular point sequence through high-precision polygon approximation; and then traversing each angular point, calculating an included angle of adjacent vectors, identifying angular point pairs conforming to right-angle characteristics, further calculating a pixel side length and selecting a minimum value, and finally converting the pixel side length into an actual size through a calibration proportion. According to the method, the internal independent square structures can be separated and positioned from the visually connected complex contours by detecting the continuous right-angle characteristics from the angle of basic geometric attributes of the graph; according to the mechanism, the false recognition rate of overlapped and adhered targets is reduced, so that reliable and fine graph analysis capability is provided in a scene containing dense and staggered graphs.

The invention relates to a method for optimizing dimension chain models of sub-parts of a product assembly. The method comprises the steps that the dimension chain models of all the sub-parts are constructed based on three-dimensional geometric data, assembly process parameters and measurement dimension data of the sub-parts; determining simulation size data of the size chain model of the product assembly obtained by assembling the sub-parts based on the size chain model of the sub-parts; measuring actual size data of a product assembly obtained by assembling the sub-parts; and correcting model parameters of the dimension chain model of the sub-part based on first deviation data between the simulated dimension data and the actual dimension data of the product assembly. According to the method and the device, the difference between the simulation size data and the actual size data of the product assembly is reduced in an accelerated manner in a pilot production stage or a product development and debugging stage, the actual size data of the product assembly can meet the dimensional tolerance of the size chain model after less times of tests, the dimensional accuracy of the product assembly is effectively improved, and the product assembly is more accurate. And the problem of size out-of-tolerance in the sub-part assembling process is reduced.

The application discloses a kind of digital processing method of space special-shaped curtain wall decorative panel, panel and construction method, including based on modeling graphics, along the contour edge of decorative panel, draw sorting guide line;Determine the geometric center of each panel, and make relative position mark;The position of the geometric center of each described panel is unchanged relative to the sorting guide line, the sorting guide line is converted into straight line, while several described geometric center is flattened into dot matrix graphics;After flattening dot matrix graphics and straight line state sorting guide line are projected into the same plane in world coordinate system, and according to actual size, etc. Enlargement, get flattened plane graphics;Each described panel is placed in flattened plane graphics according to original position distribution one-to-one correspondence, and get processing graphics. Numerous panel individuals can be block linked, the relative position of each panel and arrangement mode are determined, to be more beneficial to processing and arranging, so as to improve panel utilization and reduce material cost.

The invention relates to a coating optimization method and device for a surface coating of a high-voltage electrode in GIS equipment. The method comprises the steps that a GIS model is manufactured according to the real size scaling of the GIS equipment; continuously pressurizing and electrifying the GIS model on which the coating samples with different thicknesses are placed until the coating samples are broken down; drawing a first insulation life curve corresponding to the GIS model; constructing a GIS electric field simulation calculation model; drawing a second insulation life curve of the GIS equipment; and coating optimization is carried out according to the second insulation life curve and the actual life demand. According to the method, through scaling model experiment and electric field intensity conversion, accurate prediction of the insulation life of the high-voltage electrode coating in the GIS equipment is realized; and meanwhile, on the basis of the Weibull distribution theory and MATLAB fitting, a correlation formula of the electrode coating insulation life and the surface electric field intensity under different thicknesses is obtained, so that the optimal laying scheme of the high-voltage electrode is optimized, and a guarantee is provided for the insulation stability of the GIS.

The invention discloses a characterization method for the three-dimensional average size of inclusions in steel based on two-dimensional data statistics. The characterization method comprises the following steps: obtaining the size data of the inclusions of a steel sample on a two-dimensional detection plane; according to the inclusion size data, calculating to obtain a statistical mean value and a statistical variance of the inclusion size on the two-dimensional detection plane; and on the basis of the statistical mean value and the statistical variance, the average size of the inclusions in the steel in the three-dimensional space is calculated through a preset theoretical model for reflecting mathematical correlation between the two-dimensional section statistical magnitude and the three-dimensional space size. On the basis of conventional two-dimensional plane inclusion measurement, statistics of the size variance of the inclusion in the detection plane is added, accurate derivation from a two-dimensional measurement result to a three-dimensional real size is achieved in combination with a theoretical formula, and the real size condition of the inclusion in the steel can be represented more accurately.

The present invention discloses a rudder simulation device for sailboat simulation training, comprising a rudder, a clamping device a, a rudder angle recognition device, a rudder rudder plate, a clamping device b, a resistance feedback adjustment device connecting plate, a stern plate, and a rudder resistance feedback adjustment device. The rudder is composed of a rudder of real size, ensuring the authenticity of the simulation; the clamping device a and the clamping device b clamp and fix the entire device to the stern plate to avoid damaging the original sailboat structure and maintain the integrity of the sailboat; the rudder angle recognition device comprises a synchronous wheel, a synchronous belt and an encoder, which is used to output the rudder rotation angle to a computer; the resistance feedback adjustment device connecting plate is used to connect the rudder resistance feedback adjustment device to the rudder. The force feedback adjustment device and the clamping device a and the clamping device b are provided with a structure with adjustable length, which can be adjusted and fixed according to different rudders; the rudder resistance feedback adjustment device includes a resistance generating device and a resistance adjustment device, wherein the resistance adjustment device mainly includes two parts: a transmission device and a reducer device; the present invention solves the technical problems in the prior art that the angle information of the rudder cannot be read by the computer to form effective feedback, and the motion and visual simulation system cannot make real-time feedback changes according to the user's sailboat operation. The present invention can be quickly installed and disassembled, is simple to operate, maintains the integrity of the original sailboat equipment, meets the requirements of portability and installation, and is highly economical.

The application discloses a method and device for inverse form-finding of cable elements, and a storage medium, which projects a building space curved surface to an xoy plane to obtain a building outer contour projection line; an initial net shell structure model is designed according to the building outer contour projection line and combined with building modeling needs; real boundary conditions are set for the net shell structure model, real loads are applied, all beam elements in the net shell structure model are replaced by cable elements, initial stiffness is set, and an initial cable structure model is obtained; a finite element analysis software is used to obtain deformation forms and deformation data of the cable structure model under real loads and real boundary conditions; the obtained deformation data is scaled and inverted in combination with actual building sizes to obtain an optimal structure form matched with an actual building form. The form-finding calculation process of the application does not need multiple iterative calculations, effectively saves time cost, greatly improves structure form-finding efficiency, is simple to operate, low in design difficulty and high in design efficiency.

A method for determining object's size, including: providing a calibration card having a picture printed thereon; providing a first real size of the picture; setting a camera of a computing device to a fixed focus; capturing a first image of the calibration card when the picture is in focus; measuring a first virtual size of the at least one dimension in the first image in pixels; calculating a pixel density of the first image, by dividing the first virtual size by the first real size; capturing a second image of an object, when the object is in focus; measuring a second virtual size of at least one distance between two virtual points located on the second image; calculating a second real size of at least one distance between two real points of the object corresponding to the two virtual points, by dividing the second virtual size by the pixel density.

To provide a structure management device capable of knowing the shape of an abnormal part on the surface of a concrete structure in a correct size.SOLUTION: A structure management device 1 includes a photographing device 5 that photographs a predetermined portion of a photographing target 3, a distance measurement device 7 that measures a distance between the predetermined portion of the photographing target 3 and the photographing device 5, and a dimension calculation unit 9 that calculates an actual size of an abnormal portion 21 appearing on a surface 17 of the photographing target 3 by using an image 15 obtained by photographing of the photographing device 5 and a value of the distance measured by the distance measurement device 7.SELECTED DRAWING: Figure 1

The invention provides a construction site size measurement method and device and electronic equipment, and the method comprises the steps: obtaining a three-dimensional model of a to-be-measured construction site, planning an inspection route of a measurement device according to the three-dimensional model, controlling the measurement device to move to a target measurement point along the inspection route, and carrying out the posture correction of the measurement device based on an optical mark. Scanning a to-be-measured target based on the measuring device to obtain local point cloud data, converting the local point cloud data to a global coordinate system to obtain global point cloud data, determining actual size parameters of the to-be-measured target according to the global point cloud data, and determining theoretical size parameters of the to-be-measured target according to the three-dimensional model. If the difference value between the actual size parameter and the theoretical size parameter is smaller than a preset difference value threshold value, the actual size parameter serves as the size measurement result of the to-be-measured target, and the technical problems that a traditional manual measurement mode is low in efficiency and poor in precision and has safety risks are solved through the method.

The invention discloses a space surveying and mapping method and device and electronic equipment. The method comprises the steps of obtaining measurement data of a to-be-designed space; based on the entity data in the wall body, designing first object information corresponding to the entity data in the wall body and second object information required in the to-be-designed space; determining space information required by each object based on the space range data of the house, the first object information and the second object information; and based on the spatial information, visually presenting the first object information and the second object information in a manner of stacking a plurality of small cubes to obtain a spatial surveying and mapping model. According to the method, the first object information and the second object information in the to-be-designed space are designed through the measurement data, and the space surveying and mapping model is visually presented in a small cube stacking manner, so that a user can intuitively check a decoration effect picture of house design on the real size of a house; unified purchase of furniture and household appliances in real sizes can be met through the space surveying and mapping model.

This application relates to a method for manufacturing high-density buried via multilayer boards suitable for AI server storage modules, comprising the following steps: drilling, high-temperature baking, flash plating, dry film deposition of the vias, resin plug application, ceramic brushing, and browning film application are sequentially performed on a sub-board that has undergone one-time lamination; during the above expansion and contraction control processing steps and before the second lamination, layer misalignment detection strips pre-set in the non-functional areas of the sub-board are measured to obtain real-time expansion and contraction data characterizing the actual dimensional expansion and contraction state of the sub-board after processing; the real-time expansion and contraction data are input into a second lamination alignment system, which dynamically adjusts the grasping compensation coefficient of the composite alignment target set on the sub-board based on the real-time expansion and contraction data, and performs subsequent processing on at least two sub-boards according to the compensated alignment reference to form a multilayer board. This manufacturing method reduces expansion and contraction variables through source process optimization, significantly improving alignment accuracy and yield.

The embodiment of the present application provides a kind of face size determination method and device, wherein the method comprises: the first test face image of the target to be measured is collected, wherein the first test face image includes the first test face of the target to be measured and circular reference object;First pixel value of the first test face is obtained, and the second pixel value of the circular reference object is obtained;In the case where the circularity of the circular reference object is greater than the preset circularity threshold, the horizontal tilt angle and the vertical tilt angle of the circular reference object are obtained according to the standard major axis of the circular reference object;According to the first pixel value, the second pixel value, the horizontal tilt angle, the vertical tilt angle and the actual size of the circular reference object, the first size of the first test face is determined.The actual size of the circular reference object, the tilt angle and the pixel value in the first test face image are used to determine the first size of the first test face in the embodiment of the present application, which simplifies the measurement operation steps in the prior art and improves the accuracy of measuring the size of the face.

The invention relates to target detection, in particular to a small target measurement method and system based on vision and deep learning, and the method comprises the steps: collecting a front view, a top view and a side view of a to-be-detected product with a grid background; intercepting a grid image from the image, extracting a grid contour by adopting a maximum between-class variance method, and calculating an actual size proportion corresponding to each pixel according to the side length of a grid; the method comprises the following steps: extracting a product contour from an image of a to-be-detected product by using a pre-trained SAM model, respectively obtaining a minimum enclosing rectangle of each image, and calculating according to an actual size proportion corresponding to each pixel to obtain initial length, width and height; and constructing a depth-of-field error correction network based on the multi-layer perceptron model, taking the initial length, width and height of the product to be detected as input, and outputting the corrected length, width and height. Measurement errors caused by shooting angles, image depth of field and the like can be effectively eliminated, and the measurement precision is improved.

The invention discloses a self-adaptive partition surface layout method based on asynchronous frame scheduling, which is suitable for partition panel components in Web application. The method comprises the following steps: rendering a multi-panel area divided by a separation line in a page, initializing the panel size, registering a drag event for the separation line and registering size change monitoring for a parent container and / or a browser window; in the dragging process, the temporary size state is updated according to pointer displacement, when the temporary size changes, a size updating task is registered through a browser animation frame scheduling interface, and when a drawing frame arrives, the latest temporary size is synchronized into the actual size state at a time, and interface rendering is driven; and when the dragging is finished or the size change of the parent container is detected, adaptively adjusting the size of each panel according to a preset constraint rule. According to the method, the rendering overhead can be reduced in a high-frequency interaction scene, the dragging fluency is improved, the consistency of the component state and the interface display is ensured, and the self-adaptive capability of the partition surface layout is enhanced.

The application discloses a kind of concrete apparent bubble defect quantitative analysis method and system, it is related to computer vision technical field, method includes: collecting hardened concrete surface sample image;Using first semantic segmentation model to sample image is segmented, obtains concrete area mask;Hardened concrete surface sample image is cropped, and the image after cropping is input to second semantic segmentation model and depth estimation model, obtains bubble and the segmentation mask of size reference object and the depth information diagram of concrete area;Plane fitting is carried out to concrete surface using depth information diagram, and perspective correction is carried out;According to the proportional relationship of the pixel area of corrected size reference object mask and real size, or according to the proportional relationship of the real area of concrete area and pixel area, the real world area of bubble is calculated, and the equivalent diameter of bubble is calculated.The application realizes non-contact high-precision quantification.

The application discloses a real-time inventory method for a container yard, an electronic device and a computer program product. The method comprises: acquiring images of a target yard area collected by a camera carried by a drone and a takeoff height of the drone in a geospatial reference coordinate system; identifying four corner points of a container in each image, and determining length information and width information of the container in the image in an image coordinate system, to further determine actual size information of the container in each image, coordinate information of a geometric center point of the container in the image coordinate system, and an azimuth angle of the container in the image coordinate system; determining a height of the geometric center point in the geospatial reference coordinate system; and determining a total number of containers in the target yard area based on the takeoff height of the drone, the height of each geometric center point in the geospatial reference coordinate system, and the actual size information of the container in each image.

An efficient processing system for plates, comprising: a moving station (P) having a conveying direction (T); a device (1) for cutting out a section (C) having a shape (F) and an actual dimension (D '') from a sheet (L) having a surface (S) and comprising an optimization algorithm integrated with the cutting tool; machining means (2) configured to perform machining of the surface of each section (C) of the shape (F) or of the thickness throughout each section; a printing unit (3) configured to receive as input a set of images (I) having a desired size (D ') and to dispense ink onto each segment (C) of the shape (F), where the printing unit (3) is located downstream of the cutting device (1) taking into account the conveying direction (T); a drying device (11) configured such that the ink dispensed onto each section (C) is dried; an identification unit (4) comprising a plurality of sensors configured to detect the shape (F) and the actual size (D ") of each segment (C); a matching unit (5) configured to match images (I) belonging to the group of images with the shape (F) of each segment (C) detected by the identification unit (4); an alignment unit (9) configured to detect the presence of a rotation angle (alpha) between the section (C) and the main transport direction (T); and a control unit (10) configured to receive as inputs the image (I) or the trimmed image (Ir), the actual size (D ''), the expected size (D ') and the angle (alpha) of each segment (C), and to send operation commands to the printing unit (3); at least one of: a scaling unit (7) integrated with a scaling algorithm and configured to adapt an expected size (D ') of the image (I) to an actual size (D' ') of each segment (C) detected by the identification unit (4) and transmitted to the scaling unit (7); a trimming unit (8) integrated with a trimming algorithm, operating alternately with the scaling algorithm of the scaling unit (7), and configured to receive as input the actual size (D ") of each segment (C) transmitted by the detection unit (4) and to calculate a trimmed image (Ir) based on the image (I).

The invention discloses a mold matching system for magnetic material processing, which belongs to the technical field of magnetic material processing, and comprises a mold library management module for storing and managing attribute information of molds; the technological parameter input module is used for inputting technological parameters related to each working procedure of magnetic material production; the intelligent matching engine is used for calculating and dynamically adjusting the influence weights of the process parameters on the sizes of the blank in the die width direction, the orientation direction and the pressing direction according to a historical production data cluster; based on the target blank size and the process parameters input in real time, an optimal mold and a matched process scheme are obtained through matching; and the closed-loop feedback optimization module is used for collecting the actual size of the sintered blank, carrying out difference comparison on the actual size and the preset size, and carrying out parameter updating and optimization on the three-dimensional weight dynamic optimization model through a deep reinforcement learning mechanism based on difference data. According to the method, collaborative optimization of the mold and process parameters is realized, and the problem of low efficiency of traditional manual matching is solved.