24 results about "Edge segment" patented technology

The invention relates to the technical field of road engineering construction, and discloses a quality monitoring method and system for road construction, and the method comprises the steps: obtaining a time sequence image sequence of a construction region, carrying out the processing of the image sequence based on an improved Canny edge detection process, obtaining a binary image containing texture features, and carrying out the processing of the binary image. The method comprises the steps of calculating edge offset data of edge line segments between adjacent frames based on a dynamic time warping algorithm, constructing a pavement texture change model based on the edge offset data, identifying an abnormal region and calculating an area proportion of the abnormal region, and obtaining a construction quality evaluation index based on an edge expansion index and fusing data collected by an Internet of Things sensor. And comparing the construction quality evaluation index with a quality preset threshold value, triggering a grading early warning mechanism, and synchronously pushing the grading early warning mechanism to a highlight area of the construction site model. Based on the improved algorithm, the road surface texture noise interference is effectively inhibited, the edge detection stability is improved, the compactness prediction accuracy is improved, and the road construction quality monitoring efficiency is improved.

The invention relates to the field of integrated circuit manufacturing, in particular to an optical proximity correction method and device. Determining a target wafer grid image according to a preset grid size and the target wafer pattern; taking the target wafer pattern as a current mask pattern; inputting the current mask pattern into a pre-trained three-dimensional optical proximity correction model, and obtaining an emulational wafer grid image with the assistance of a grid size; calculating an edge height error corresponding to each edge section; judging whether the edge height error exceeds a preset adjustment threshold value or not; when the edge height error exceeds a preset adjustment threshold value, adjusting the position of each edge section in the corresponding normal direction according to the edge height error to obtain an updated current mask pattern; and the step 4 to the step 7 are cycled until the edge height error does not exceed the adjustment threshold value. According to the invention, the defects of photoresist height loss and the like are reduced, and the yield of finished products is improved.

The invention relates to the technical field of region segmentation, in particular to a splicing type prefabricated part splicing region identification method based on image semantic segmentation, which comprises the following steps of: acquiring a part image and extracting a light and shade change region, analyzing pixel arrangement and a jump edge, endowing a semantic label, tracking a contour path and extracting a trend and a direction. Recognizing turn-back, eliminating abnormal edge segments, recombining paths, extracting center line coordinates, adjusting directions, complementing point positions, and generating a label map to divide a splicing seam area. According to the method, boundary regions are divided through pixel arrangement and brightness edge connectivity, independent labeling of semantic blocks is completed, a structured edge segment set is constructed through path trend and turning point information, the coherence and connection relation sequence of the boundary direction is kept, direction sudden change positions are recognized, path anomalies are eliminated, and structural fracture and overlapping are reduced; central line pixels are extracted, a path trunk is formed through point supplementing operation at the turning position, an attribution area is delimited by combining contour extension and label coding, and the consistency and segmentation integrity of the abutted seam boundary are improved.

A method includes: providing a first design layout including a plurality of cells; updating a first cell of the plurality of cells using optical proximity correction to provide a first updated cell and a data set; and updating a second cell from remaining cells in the first design layout based on the data set and a model without involvement of optical proximity correction to provide a second updated cell, wherein the model includes hidden layers including nodes and is trained to obtaining converged values of the nodes of the hidden layers through providing a mapping of edge segments before lithography enhancement and edge segments after lithography enhancement using optical proximity correction, and wherein at least one of the providing, and updating is executed by one or more processors.

The invention provides an OPC correction method and system and a layout, and belongs to the field of semiconductors. The OPC correction method comprises the following steps: providing a layout, wherein the layout comprises a plurality of metal layer patterns; and performing OPC correction on the segmented line segment, expanding the first line edge segment outwards to the periphery of the corner, and shrinking the second line edge segment and the third line edge segment inwards to the interior of the line pattern. According to the method, OPC correction is carried out on the segmented line segments, the first line edge segment is expanded outwards to the periphery of the corner, and the second line edge segment and the third line edge segment are contracted inwards to the interior of the line pattern. When the pattern after OPC is used for manufacturing the photomasks, the distance between the two adjacent photomasks can be increased, so that the photomask manufacturing can be prevented from being restricted by manufacturing process conditions, the photomasks are prevented from generating defects, and the yield of photomask manufacturing is improved.

The invention provides an adjustable modular electromagnetic shielding structure. The adjustable modular electromagnetic shielding structure comprises a plurality of L-shaped monomer modules, each L-shaped monomer module is composed of a first edge section and a second edge section which are vertically connected with each other, the length of each first edge section is larger than that of each second edge section, and the plurality of L-shaped monomer modules can be mutually spliced to form the modular electromagnetic shielding structure in a two-dimensional space and a three-dimensional space. The adjacent L-shaped monomer modules can generate relative displacement through a driving mechanism, so that the modularized electromagnetic shielding structure is switched between an unfolded state and a folded state. According to the adjustable modular electromagnetic shielding structure provided by the invention, through two-dimensional and three-dimensional orthogonal stacking of the L-shaped monomer modules, multi-dimensional adjustability of the porosity of the electromagnetic shielding structure is realized, and the structure has a self-locking characteristic, is high in stability and modular degree, is convenient to expand and maintain, and is suitable for various electromagnetic shielding application scenes.

The present application relates to the technical field of mask lithography, and discloses a hot spot layout correction method, device and equipment in a mask layout and a storage medium, the hot spot layout correction method comprises determining a hot spot layout in an initial mask layout; determining the environmental graphic parameters of each contour edge segment in the hot spot layout; constructing a basic optimization objective function according to the edge placement error varying with the displacement amount of the contour edge segment; superimposing an optimization compensation term on the basic optimization objective function according to each environmental graphic parameter to obtain an optimization objective function; optimizing the displacement amount of each contour edge segment according to the optimization objective function to obtain the corrected displacement amount corresponding to each contour edge segment, so as to determine the corrected layout of the hot spot layout. The present application realizes the automatic correction of the hot spot layout without manual operation, improves the correction efficiency of the hot spot layout, and ensures the accuracy and reliability of the correction result.

Provided is a method for manufacturing a semiconductor device including performing optical proximity correction on a layout. The optical proximity correction includes generating a contour of a target pattern of the layout, generating first to fourth error vectors between the contour and first to fourth edge segments of the target pattern, and generating a correction pattern by sequentially performing a first correction and a second correction on the target pattern. The first correction includes moving the first edge segment on the basis of the first error vector, and moving the second edge segment on the basis of the second error vector. The second correction includes moving the third and fourth edge segments on the basis of a sum of the third error vector and the fourth error vector. The first and second edge segments face each other, and the third and fourth edge segment face each other.

The invention belongs to the technical field of integrated circuit manufacturing and industrial detection, and provides a bonding pad via hole recognition and positioning method and system, and the method comprises the steps: carrying out the preprocessing of a bonding pad image, and extracting a continuous edge segment in the bonding pad image based on an edge detection algorithm; calculating a curvature inner side angle of the continuous edge section, and segmenting the continuous edge section based on the curvature inner side angle to obtain a plurality of arc sections; calculating the rough radius of the arc section; based on a preset radius matching condition and the rough radius pair arc section, if the ratio of the arcs in any arc group is greater than a preset threshold value, calculating the circle center and the radius of the arc group based on triangle fitting, otherwise, calculating the circle center and the radius of the arc group through least square fitting; and verifying the circle center and the radius by adopting a Helmholtz significance criterion pair, and outputting a corresponding circle center coordinate and radius as a via hole positioning result when the verification is passed. The method has better recognition performance, and reduces the misjudgment probability caused by a short arc at the same time.

This invention belongs to the field of image segmentation technology and discloses a method, system, medium, device, and terminal for extracting identification documents. The system utilizes an embedded device to acquire multispectral images of the identification document; it models the document image edge segments based on their straight-line geometric properties and infers from the contextual information of the global image; it extracts global image features using a ResNet network and then encodes them using a Deformable DETR encoder; in the first-stage decoding process, it predicts the document edge segments using an attention mechanism and learnable line segment and position queries; in the second-stage decoding process, it compares the correlation between line segment features and image features to predict the relative order between line segments, and obtains the complete edges, vertices, and bounding boxes of the document image through perspective transformation. This invention uses a two-part matching method to predict line segments, avoiding pre- and post-processing, simplifying the detection channel, and achieving true end-to-end processing.

The application discloses a kind of based on deep learning's forging defect feature identification method, it is related to intelligent manufacturing technical field, the edge pixel coordinates of candidate defect area are read, and edge sequential numbering and edge trend sequence are generated;Three point direction variation is calculated, and curvature transition point is determined according to threshold;With the edge pixel sequence between adjacent curvature transition point generates defect profile segmentation label;The direction variation sequence of edge section, the sequence of internal and external gray difference and the trend of section head and section tail are called, and curvature direction symbol sequence, gray difference synchronous change state and edge section closure trend are generated, and each segmented exclusive selection is selected to stable defect curvature channel, texture interference exclusion channel or to be judged curvature channel, and form edge section level supervision signal.The application can make neural network with defect profile segmentation as supervision unit to carry out channel learning, reduce the feature of real defect boundary and texture, noise edge is mixed, improve the delicacy and explainability of forging defect identification.

The present application relates to the field of integrated circuit manufacturing, in particular to an optical proximity correction method and device, by receiving a target wafer pattern; determining a target wafer grid image according to the target wafer pattern according to a preset grid size; taking the target wafer pattern as a current mask pattern; inputting the current mask pattern into a pre-trained three-dimensional optical proximity correction model, and assisting with the grid size to obtain a simulation wafer grid image; calculating the edge height error corresponding to each edge segment; judging whether the edge height error exceeds a preset adjustment threshold; when the edge height error exceeds the preset adjustment threshold, adjusting the position of each edge segment in the corresponding normal direction according to the edge height error to obtain an updated current mask pattern; repeating steps four to seven until the edge height error does not exceed the adjustment threshold. The present application reduces defects such as photoresist height loss and improves product yield.

The invention relates to a microsensor (10) for measuring a measurement variable, having an edge region (14), at least one membrane (16) having a membrane surface (22) which adjoins the edge region (14), is self-supporting, is spanned by a first main axis (18) and a second main axis (20), and has a total length (26) along the first main axis (18), and which can be deflected in a normal direction (24) of the membrane surface (22) depending on the measurement variable, and stiffening means (34) connected to the membrane surface (22) which are arranged, running along the first main axis (18), at least in a first section (36) between a first edge section (38) of the edge region (14) as far as a centre (40) of the total length (26), and which have a width (52) in the direction of the second main axis (20), wherein the width (52) of the stiffening means (34) increases precisely once at least in a first direction (54) along the first main axis (18), starting from the first edge section (38) to the centre (40), and then decreases again precisely once.

The invention relates to the technical field of image pattern recognition, in particular to a diopter parameter analysis model optimization method and system, and the method comprises the following steps: collecting image sequence recognition region features, constructing edge segments, judging the structural continuity, executing concurrent recognition and trajectory analysis, and outputting a diopter parameter analysis optimization graph. According to the method, through constructing curvature section and gray vector distribution in an image sequence, extracting continuous features of a boundary structure, screening stable graph frames in combination with inter-frame direction consistency, recording corresponding boundaries by using a dual-channel recognition rhythm alignment mode, establishing a structure difference value and generating an error trajectory graph; the stable nodes are identified based on the offset trend, and the key graph segment is backtracked, so that the restoration coherence and response consistency of the image structure are enhanced, and the expression ability of the graph result to the change characteristics of the crystalline lens and the accuracy of parameter extraction are improved.

The invention relates to the field of part size measurement, in particular to a part size measurement method and system for a mechanical part, and the method comprises the steps: obtaining a to-be-measured grayscale image of a to-be-measured mechanical part and a template grayscale image of a corresponding mechanical part with a known size; processing the to-be-measured grayscale image to obtain a to-be-measured enhanced grayscale image, and extracting all single edge fragments in the to-be-measured enhanced grayscale image; and processing the template grayscale image to obtain a template enhanced grayscale image, extracting all single edge segments in the template enhanced grayscale image, determining a size mapping relation according to the single edge segments and known sizes of the single edge segments, and calculating the actual size of each single edge segment in the mechanical part to be measured based on the size mapping relation. According to the invention, high-precision extraction of multiple directions and weak edges of the mechanical part is realized, and the precision of dimension measurement is improved.

Provided are an optimal near-field generation method and a mask manufacturing method comprising the optimal near-field generation method. The optimal near-field generation method may include obtaining a mutual interference complex diffraction pattern formed by mutual interference between a plurality of spherical waves formed as a certain plane wave incident on each of a plurality of edge segments differentiated from an edge of a design layout is scattered on each of the plurality of edge segments, obtaining a complex near-field by applying a Kirchhoff boundary condition to the mutual interference complex diffraction pattern, and obtaining an optimal near-field by optimizing the complex near-field so as to reduce a difference between the complex near-field and a rigorous near-field of the design layout.

The invention provides a machine vision appearance defect detection method and system for a wire harness connector, and the method comprises the steps: carrying out the edge contour extraction of a collected wire harness connector image, obtaining the edge contour of a crimping region and the edge contour of a cable insulation region, and determining a crimping deformation region; determining a contour anomaly coefficient based on the curvature difference value and the connection continuity between the adjacent edge sections; extracting an insulating layer edge section at the junction of the crimping area and the cable insulating area, and determining a damage risk value of the insulating layer; performing flash detection, crack detection and recess detection on the edge contour of the crimping area to obtain a flash area, a crack length and a recess depth, and determining a crimping appearance quality value; and performing weighted summation on the contour abnormal coefficient, the insulation layer damage risk value and the crimping appearance quality value to obtain an appearance defect comprehensive index, and determining the crimping appearance quality grade of the wire harness connector according to the appearance defect comprehensive index. According to the scheme of the invention, the intelligent monitoring of the crimping appearance defect of the wire harness connector is realized through the multi-source feature fusion.

The invention relates to the technical field of plane planning layout, in particular to a hot-line work toolbox tool placement optimization design method, which comprises the following steps of: acquiring tools required by each stage and counting frequency, classifying common tools, calculating a distance between a placement area and an operation orientation, determining a preferred area and a corresponding task group, and correcting tool orientation to generate a matching table. An operation chain is formed according to the sorting connection sequence, and the direction and the position are integrated to generate an optimized layout diagram. According to the method, high-frequency tools are preferentially included into task group function division through tool use frequency and stage repetition rate analysis, short edge sections are screened as preferential placement areas according to the distance between the boundary and the operation direction, and the orientation of the tools is corrected in combination with included angle measurement, so that the tool orientation tends to be consistent in the boundary direction; the method comprises the following steps: extracting a point position relation of adjacent tools according to a function connection sequence to construct a communication path, optimizing tool arrangement according to logic and direction, compressing an operation path, reducing orientation interference, and improving pick-and-place coherence, space moving line coordination and task rhythm matching.

This invention discloses a machine vision-based method for detecting the size of lithium battery electrode sheets. The method includes acquiring an image of the lithium battery electrode sheet and roughly locating the edges; extracting edge distribution features containing edge gradient directions and position coordinates, matching them with a pre-stored template to determine the specifications and model, and setting the search area using detection parameters; acquiring candidate edge points within the search area, and employing a dual strategy to filter and eliminate false edge points: when the maximum cluster proportion reaches a preset threshold, neighboring points are filtered using the maximum cluster anchor point; if the threshold is not reached, the optimal point is selected based on gradient filtering conditions and a preset side direction; and calculating the final edge segment and size based on the true edge points using sub-pixel precision calculation and iterative fitting. This invention achieves adaptive registration for specifications and highly robust extraction of complex edges, improving detection accuracy and automation.

This application relates to the field of robotics technology and discloses a robot visual target detection method and system, including acquiring visual image information of the target to be detected; identifying multiple edge segments that are broken due to occlusion from the visual image information; logically connecting the edge segments located on the same straight line to construct the reconstructed edge of the target to be detected; acquiring preset geometric specification information that matches the category of the target to be detected, and determining a reference benchmark that meets preset stability conditions in the reconstructed edge; performing geometric extrapolation based on the reference benchmark and the preset geometric specification information to complete the missing contour features of the target to be detected in the occluded area; determining the working pose information of the target to be detected based on the complete contour features of the target to be detected after completion, and restoring the complete contour of the target when the target is partially occluded and the visual information is incomplete, accurately determining its working pose, and improving the accuracy and robustness of target detection in complex occlusion scenarios.

This specification discloses a method, apparatus, storage medium, and electronic device for dense-state data statistics. The method includes: first, arranging the determined dense-state data to be statistically analyzed to obtain a dense-state data sequence; then, for each preset segmentation rule, dividing the dense-state data sequence into subsequences according to that rule; and performing dense-state statistics on the dense-state data contained in each subsequence in parallel to obtain the segmented statistical value of each subsequence. Next, for each preset edge segmentation rule, dividing the dense-state data sequence into edges according to that rule; and determining the edge statistical value of each edge segment in parallel based on the segmented statistical value of each subsequence corresponding to each segmentation rule; and finally, determining the statistical result of the dense-state statistics for each dense-state data based on the edge statistical value. Using a combined serial and parallel approach for dense-state data statistics can reduce serialization and decrease network interaction latency.

This invention relates to a method and apparatus for analyzing images of microlithographic microstructure components, wherein each of a plurality of pixels in the image is assigned an intensity value under various conditions. The method according to the invention includes the following steps: separating a plurality of edge segments in the image (step S130); classifying each of the separated edge segments as a relevant edge segment or an irrelevant edge segment (step S140); and determining continuous segments in the image based on the relevant edge segments (step S150).

The present application relates to the technical field of machine vision, and discloses a mechanical watch pointer reading detection and recognition method and system based on machine vision, which comprises the following steps: acquiring an image containing a dial and a pointer; determining a pointer rotation center and extracting edge features; when the edge features are missing, identifying visible edge segments and determining local center features by using geometric symmetry; combining the rotation center and the local center to construct a pointer pointing reconstruction line; extending the reconstruction line to the scale area to obtain a virtual pointing position; determining the reading by angle mapping and comparing it with the qualified range, outputting a detection conclusion and saving data. The present application can effectively deal with the case of partial occlusion or incompleteness of the pointer, improve the reading accuracy and robustness, and is suitable for automatic quality inspection scenes.

The invention relates to the technical field of semiconductor photoetching processing, and discloses an optical mask pattern correction method, device and equipment and a storage medium, the correction method comprises the following steps: obtaining an initial mask pattern, segmenting the contour edge line of the initial mask pattern to obtain segmented edge segments; performing corresponding EPE error operation on the evaluation points on the segmentation edge sections to obtain EPE error values corresponding to the segmentation edge sections; and comparing the EPE error value corresponding to each segmentation edge segment with an EPE threshold value, and performing optimization iteration on the edge segment number, the edge segment length and the edge segment position of the segmentation edge segment according to a comparison result until the EPE error value corresponding to each segmentation edge segment is smaller than the EPE threshold value, and determining the segmentation edge segment after the initial mask pattern is corrected. In the process of correcting the optical mask pattern, optimization iteration is carried out on three different dimensions including the number, the length and the position of the segmentation edge sections, and the exposure precision of the photoetching mask plate and the semiconductor photoetching precision are effectively improved.