28 results about "Minimum bounding box" patented technology

The invention discloses a man-machine interaction method, device and equipment based on an aircraft unstructured grid and a storage medium, and relates to the technical field of fluid mechanics, and the method comprises the steps: constructing a current directed bounding box as a root node based on a to-be-processed unstructured grid of an aircraft, equally dividing the current directed bounding box into a plurality of sub-bounding boxes from the root node, binding the sub-bounding boxes with sub-nodes of the root node, and carrying out layer-by-layer iteration until a preset hierarchy threshold value is reached to obtain a target space segmentation structure; determining a minimum bounding box of each interaction object in the generated interaction object grid set, performing intersection detection on the minimum bounding box and each node in the target space segmentation structure, and storing data of the interaction objects to a bottom leaf node; the capture state identifier is set by judging whether the interaction object data is stored in each bottom-layer leaf node, and collision detection with the bounding box of each bottom-layer leaf node provided with the capture state identifier is performed in the interaction range of the interaction instruction to obtain the target interaction object, so that the man-machine interaction efficiency is improved.

The invention discloses a neighbor soft package two-dimensional segmentation and three-dimensional positioning combined method, which comprises the following steps of: segmenting a two-dimensional picture by using a Segment Anything algorithm and carrying out preliminary screening according to the contour area of a segmented object; performing coordinate range screening, determining and calculating an external rectangle and an area thereof through processing of a two-dimensional minimum compact bounding box, and outputting a minimum area by rotating a line to be greater than 90 degrees; obtaining an included angle between an object center coordinate and x and y through a bounding box, calculating a result through a formula, and performing range screening to segment a recognition object; processing the soft package by using a two-dimensional minimum bounding box algorithm to obtain a center coordinate and an included angle; acquiring internal and external parameter matrixes, two-dimensional coordinates and depth of the camera, and converting two-dimensional image coordinates into three-dimensional coordinates in a camera coordinate system and a world coordinate system; and calculating a neighbor point cloud normal vector to obtain a three-dimensional attitude of the soft package center point. According to the positioning method disclosed by the invention, the problem of low positioning precision of the soft package is solved by a minimum compact bounding box algorithm, and a more accurate grabbing pose is conveniently provided for the robot.

The invention provides an object segmentation method based on multimodal data fusion and an image annotation tool. The method includes acquiring an initial RGB image, an initial infrared image, and an initial depth image that contain an object; aligning the initial RGB image, the initial infrared image, and the initial depth image to obtain a first RGB image, a first infrared image, and a first depth image; specifying an initial prompt point in the first RGB image, and acquiring first masks characterizing an object region from the images in different modalities; fusing pixel values of the images in different modalities based on the first masks of the first RGB image, the first infrared image, and the first depth image to obtain a second mask; and determining a minimum bounding box of the object, and calibrating the minimum bounding box to obtain a segmentation result of the object.

The invention provides a polygon clipping processing method based on GPU acceleration, comprising the following steps: forming a minimum frame: reading the shape of a to-be-processed graph to form the minimum frame surrounding the to-be-processed graph, the minimum frame having a plurality of extreme points; the center of a first segmented grid is determined according to one of the extreme points, the segmented grid is a regular polygon, coordinates of all vertexes of the first segmented grid are obtained through calculation according to the center of the first segmented grid, other segmented grids are formed according to center coordinate deviation of the first segmented grid, and the center of the first segmented grid is a regular polygon; the minimum frame is completely covered by the segmentation grids; and kernel clipping: grouping the segmentation grids, performing parallel processing on each group of segmentation grids by using a GPU calculation unit so as to clip each group of segmentation grids edge by edge, and reserving vertexes of the segmentation grids in the to-be-processed graph and intersection points formed by the outer contour of the segmentation grids and the outer contour of the to-be-processed graph.

The application discloses a kind of based on foreground perception's no ROI marking medical image classification method, comprising: the pre-processing of three-dimensional medical image data, obtain standardized three-dimensional image data;In the condition of no ROI marking, based on pixel intensity distribution and spatial connection relationship, self-adapting foreground mask is constructed and three-dimensional minimum bounding box is determined;Accordingly, the foreground three-dimensional image data is obtained by cutting;It is input three-dimensional convolutional neural network to extract three-dimensional convolutional feature, after channel attention weighting and multi-layer semantic feature fusion, input full connection layer and output the class probability of corresponding preset classification category by Softmax function.The application can automatically obtain foreground area without artificial ROI marking, reduce redundant background interference, and improve image classification accuracy and stability through three-dimensional convolution, channel attention and multi-layer semantic fusion.

The application discloses a kind of based on robot scanning viewpoint planning process product three-dimensional measurement method and device, the method comprises: the size information of measured object is acquired, the minimum bounding box is determined according to size information, scanning viewpoint is planned according to the minimum bounding box, and coarse grid model is generated based on scanning viewpoint;Coarse grid model is discretized into point cloud model, and the hole recognition of point cloud model is carried out using maximum angle criterion, and the hole boundary feature point is obtained, the clustering of hole boundary feature point is carried out, and the local hole in point cloud model is determined, and the secondary fine scanning of each local hole is carried out, and secondary fine scanning model is generated;Point cloud registration and fusion are carried out to coarse grid model and secondary fine scanning model, and the digitization model of measured object is obtained, the integrity, precision and efficiency of scanning can be guaranteed, and the automatic measurement of process product is realized.

The invention relates to the field of electric power, and discloses a substation control equipment terminal strip drawing identification method, device and equipment, and the method is used for converting a terminal strip drawing into a standardized intermediate file. The method comprises the following steps: acquiring an electronic image of a terminal block drawing, and analyzing and identifying space coordinates of connection points, space coordinates of connection lines, line types, line widths and space coordinates of short circuit symbols; identifying and extracting a plurality of character sequences and corresponding minimum bounding box coordinates; establishing a semantic binding relationship between the connection points and the character sequence based on the spatial coordinates, determining a spatial proximity relationship of the connection points in combination with a terminal arrangement rule, and identifying a short-circuit relationship based on the proximity relationship and the spatial coordinates of the short-circuit symbols; integrating the space coordinates, the line types and the line widths of the connecting lines, the space coordinates of the connecting points, the semantic binding relation and the short-circuit relation, and constructing a topological relation graph; elements and loops in the atlas are classified based on a power industry coding rule base, and are converted into a standardized intermediate file in combination with an engineering modeling rule;

Embodiments of the present application provide a large gantry type identification method and device based on 3D vision, which comprises: acquiring a point cloud set of the upper surface features of a workpiece moving at a constant speed through a 3D camera; filtering out background point clouds in the point cloud set to obtain preliminary extracted point clouds; through statistical filtering, removing random noise points distributed globally from the preliminary extracted point clouds, and then through clustering segmentation, eliminating invalid point cloud clusters, and obtaining pure point clouds; calculating the minimum bounding box of the pure point clouds, and determining the main direction of the workpiece and establishing a workpiece coordinate system; in the constructed workpiece coordinate system, performing feature calculation on the workpiece; comparing the results of the feature calculation with a pre-stored all-gantry-type-feature database, and outputting the gantry type according to the comparison result. Through complete point cloud reconstruction, multi-feature fusion identification, size and structure double verification, and man-machine collaborative error correction mechanism, high-precision, high-flexibility and high-robustness automatic identification of large gantry types is realized.

To address the shortcomings of existing methods for detecting surface texture defects in complex curved parts, such as low detection efficiency, poor stability and recognition, and difficulty in rapidly and accurately quantitatively characterizing sub-millimeter-level surface texture defects in three dimensions, this invention provides a rapid three-dimensional quantitative characterization method for surface defects in curved parts. This invention utilizes a light field camera to simultaneously acquire two-dimensional images and three-dimensional point cloud data, improving the consistency and stability of subsequent three-dimensional quantitative characterization of defects. It employs a deep learning semantic segmentation model to extract two-dimensional contour information of defects, providing precise semantic guidance for subsequent three-dimensional spatial mapping. It establishes a coordinate mapping relationship between pixels in the two-dimensional image and physical points in the three-dimensional point cloud coordinate system, enabling precise selection of subsets of the defect's three-dimensional point cloud. Finally, it uses principal component analysis (PCA) to establish a local coordinate system, and within this coordinate system, employs the minimum bounding box method to calculate the three-dimensional geometric dimensions of the defects, achieving rapid and accurate three-dimensional quantitative characterization of the defects.

The invention discloses a display image uniformity statistical method and device and a storage medium, which are used for improving the accuracy of gray uniformity analysis of images shot by a display. Generating a row / column direction minimum bounding rectangular frame for the normalized image; performing intersection processing and connected domain separation processing according to the minimum bounding rectangular frame in the row / column direction, and calculating a gray statistical information value of a separation region; generating a blank image, replacing the gray value of the blank image with the gray statistical information value, and generating a row curved surface image and a column curved surface image; circulating the array of the blank images, generating a dynamic row and column coordinate array of a neighborhood, and generating a vector field image; performing algorithm remapping processing on the normalized image and the vector field image to generate a result image; combining the result images, and filling the maximum pixel value of the multi-channel image into the to-be-processed image to generate the to-be-processed image; and generating a target filling image, and carrying out gray level distribution uniformity analysis according to the gray level value data.

The embodiment of the present application provides a kind of modeling method, system, equipment and storage medium for extracting foreground object, belong to three-dimensional reconstruction technical field.Modeling method for extracting foreground object includes: obtaining the image point cloud data of target object and camera pose information;The foreground and background in image point cloud data are separated, and the foreground object point cloud is obtained;According to the foreground object point cloud extracted, the minimum bounding box of foreground is determined;Based on the above data, foreground point data and background point data are determined;Based on foreground point data and background point data, the preset reconstruction network is used to determine foreground object reconstruction data and background reconstruction data;According to foreground object reconstruction data and background reconstruction data, generate target object model and display.The embodiment of the present application calibrates foreground and background using minimum bounding box, reduces the cost of manual annotation, also uses preset reconstruction network to predict model data, improves the accuracy and accuracy of model reconstruction.

The chip defect detection method and system based on kernel norm minimization and machine learning, the method comprises the steps of: using the background separation method based on kernel norm minimization to separate the image foreground and background for the collected chip defect data; the pad position of the foreground image is identified through Huffman transformation, and the corresponding pixels are set to 0 to eliminate the pad image; the image processing techniques such as corrosion and expansion are used to process and eliminate noise points, realize defect area connection; all defect areas in the image are searched to obtain the minimum boundary box of the defect area; the K-Means clustering algorithm is used for defect labeling of different types of defect images to form a defect data set; finally, an end-to-end deep neural network chip defect classification cluster algorithm model is formed by using the training set and the test set. The method can realize real-time and accurate integrated circuit chip surface defect recognition, classification and positioning.

The present application relates to OLED point glue detection technical field, especially in kind, a kind of OLED point glue coating specification detection method, storage medium, equipment, it includes the following steps: step S10, point cloud data acquisition;Step S20, based on the transmission belt point cloud elimination of straight-through filtering;Step S30, based on the product point cloud area segmentation of point cloud registration;Step S40, based on the point glue coating area segmentation of Euclidean clustering;Step S50, based on the point glue coating specification detection of point cloud minimum bounding box has been used to judge whether to pass detection.The present application designs relevant algorithm can realize efficient and accurate OLED point glue coating specification detection, improves production efficiency and quality control level.

This application discloses a batch texture mapping method for irregular top surfaces of 3D models, including: S1. Generating matching texture images based on GIS features; S2. Calculating the rotation angle of the minimum bounding box of the GIS features; S3. Adding the rotation angle of the minimum bounding box to the GIS features; S4. Creating texture mapping CGA rules in CityEngine; S5. Applying texture mapping to the irregular top surface of the 3D model and correcting it. This application enables batch texture mapping of the top surface of 3D models. Based on GIS vector and raster data, the generated texture images can strictly maintain consistency with the shape of each model's top surface, thus ensuring the accuracy of the texture files; at the same time, by calculating and adjusting the texture rotation angle, the mapping accuracy is further guaranteed. As an automated batch processing solution, compared with traditional manual texture mapping, this application significantly saves labor costs and improves work efficiency.

The invention discloses a corrosion state analysis method and system for a transformer substation grounding grid. The method comprises the following steps: acquiring a standard state image and a to-be-detected state image; clustering the to-be-detected images based on the similarity to obtain an image set; target images with the maximum bounding box area and the minimum bounding box area are selected from the set; obtaining corrosion data of the target image through the image recognition model; calculating boundary frame area deviation degrees of other images and the target image; and determining corrosion data of other images by adopting a weighted fusion strategy based on deviation degree threshold judgment. According to the method, the image difference is quantified through the bounding box area difference, and the intelligent weighted fusion is combined, so that the influence of the shooting condition difference on the detection result is effectively overcome, the accurate and efficient evaluation of the corrosion state of the grounding grid is realized, and the accuracy and robustness of the detection result are remarkably improved; and reliable technical support is provided for intelligent operation and maintenance of the transformer substation grounding grid.

This invention discloses a hybrid bounding box method and system for baggage detection based on point cloud image planar projection. The method includes: acquiring 3D point cloud data of a target scene; establishing a dynamic quality benchmark based on the extreme value distribution of the 3D point cloud data in a preset direction; segmenting the 3D point cloud data using a cascaded segmentation strategy, wherein the cascaded segmentation strategy includes at least two segmentation algorithms with different segmentation characteristics; after each segmentation algorithm is executed, performing real-time quality evaluation on the current segmentation result based on the dynamic quality benchmark, and generating a quality evaluation result; determining whether to trigger the next level segmentation algorithm or output the current segmentation result based on the quality evaluation result; and generating the minimum bounding box of the target baggage based on the final output segmentation result. This invention, by constructing a dynamic quality benchmark and a cascaded segmentation strategy, achieves efficient and robust segmentation of baggage point clouds in complex scenes, significantly improving the accuracy and adaptability of baggage detection.

The present application relates to a kind of plate point cloud defect detection methods based on tensor voting principal component analysis, comprising the following steps, the three-dimensional point cloud data of the plate to be detected is obtained, and the plate point cloud data obtained is preprocessed;Plate point cloud kdtree is built, all point neighborhood point set is found, and is saved;Normal voting tensor is constructed;The voting tensor of point is constructed;The spectral component of normal voting tensor is calculated;The spectral component of point voting tensor is calculated;Defect index is calculated;Defect index threshold is set, and defect region is obtained;Defect point cloud is clustered, and delete threshold is set, and interference defect region is deleted, and finally defect region point cloud minimum bounding box is obtained.The present application has obvious advantages compared with traditional two-dimensional image defect detection, and the detection result is more accurate;Detection speed is faster.

This application discloses a method, system, smart terminal, and storage medium for quantitative assessment of open space between three-dimensional buildings, relating to the field of data processing technology. The method includes: obtaining the minimum bounding box corresponding to a three-dimensional building model in three-dimensional space; obtaining observation point control parameters; generating three-dimensional observation viewpoints based on the minimum bounding box and observation point control parameters; generating a corresponding pushbroom point set on the surface of the three-dimensional building model using a pushbroom method, and removing invisible points to obtain a visible pushbroom point set composed of visible pushbroom points; calculating the building's visual field volume corresponding to each three-dimensional observation viewpoint based on the visible pushbroom set; obtaining the theoretical visual field volume corresponding to each three-dimensional observation viewpoint; and calculating the visual field volume openness index corresponding to each three-dimensional observation viewpoint based on the building's visual field volume and the theoretical visual field volume, using this as the result of the open space quantitative assessment. This application enables quantitative assessment of open space at specific spatial points.

This application relates to the field of intelligent healthcare, specifically to a method, device, and program product for constructing a discretized model of human tissue regions. It includes: S1, acquiring an image of human tissue, calculating the minimum bounding box of the tissue region (MUB), and obtaining a region boundary map; S2, generating a mesh covering the region boundary map based on any vertex of the region boundary map, and obtaining a region mesh map; S3, filtering the pixels of the region mesh map for effective pixel blocks to obtain effective pixel block areas; S4, calculating the total area of ​​the region mesh map and minimizing the absolute value of the difference between the total effective pixel block area and the total area of ​​the region mesh map; S5, updating vertices in the neighborhood of the current vertex to obtain updated vertices, and repeating S2-S4 based on the updated vertices to obtain an absolute value set; S6, filtering the region mesh map corresponding to the vertex with the largest absolute value in the absolute value set. This application has significant application value.

The application discloses a collision avoidance method and system of an all-weather ship loader and electronic equipment, and relates to the field of collision detection. The application obtains point cloud data of the ship loader by adopting a millimeter wave radar, and establishes a three-dimensional space coordinate system based on cycle information of the ship loader. The point cloud data is unified in the three-dimensional space coordinate system through coordinate conversion and coordinate information. After normalization and clustering operations are respectively performed on the point cloud data, the required categories are obtained. The point cloud data corresponding to the categories matching the shape of the ship cabin is unified in the three-dimensional space coordinate system, and the ship cabin boundary is efficiently and accurately determined. A minimum bounding box is constructed according to the size of the roller and the platform where the roller is located, the minimum bounding box and the ship cabin boundary are projected onto the xy plane and the xz plane by using a projection method, a region of interest is set, and whether the projection of the minimum bounding box and the projection of the ship cabin boundary in the region of interest intersect or overlap is judged, so that the purpose of detecting and avoiding collision is quickly and accurately achieved.

The invention discloses a multi-beam sonar point cloud-based underwater target detection method. The method comprises the steps of preprocessing an original multi-beam point cloud; segmenting the ground point cloud by adopting a cloth simulation filtering algorithm to obtain candidate target point clouds; a DBSCAN clustering algorithm based on density is adopted to extract a target point cloud cluster; and performing three-dimensional bounding box fitting on the target point cloud cluster by adopting a directional minimum bounding box method, and outputting a final underwater target three-dimensional positioning result. The method has the remarkable advantages of no need of a prior model, high adaptability, high detection precision, automatic processing flow and the like. Statistical filtering is introduced to realize high-efficiency noise elimination, a distribution simulation algorithm is adopted to accurately separate the ground from a target, a density clustering algorithm is utilized to adaptively extract a target area, and a directional minimum bounding box is combined to realize high-precision target positioning, so that the robustness and the intelligent level of point cloud processing are remarkably improved; a solution which is low in cost, high in performance and high in practicability is provided for the fields of underwater detection, seabed structure recognition and the like.

The application discloses an aircraft non-structure grid-based human-computer interaction method and device, equipment and a storage medium, and relates to the technical field of fluid mechanics. The method comprises the following steps: constructing a current oriented bounding box as a root node based on an aircraft to-be-processed non-structure grid, dividing the current oriented bounding box into a plurality of child bounding boxes from the root node, and binding the child bounding boxes with the child nodes of the root node; iteratively layer by layer until a preset level threshold is reached to obtain a target space segmentation structure; determining the minimum bounding box of each interaction object in the generated interaction object grid set, performing intersection detection on each node in the target space segmentation structure, and storing the data of the interaction object to a bottom leaf node; and setting a capture state identifier by judging whether the interaction object data is stored in each bottom leaf node, performing collision detection on the bounding boxes of each bottom leaf node provided with the capture state identifier in the interaction range of the interaction instruction to obtain a target interaction object, so that the efficiency of human-computer interaction is improved.

This invention provides a method, apparatus, device, and readable storage medium for determining the damage level of blade cracks. The method includes: acquiring the centroid coordinates, width, and image of the blade; performing crack detection on the blade image to obtain the total number of cracks and the minimum bounding box coordinates of each crack; cropping all crack images from the blade image based on the minimum bounding box coordinates of each crack, and obtaining crack feature information for each crack based on each crack image; obtaining a blade crack feature index value based on the centroid coordinates, width, total number of cracks, minimum bounding box coordinates of each crack, and crack feature information of each crack; and determining the blade crack damage level based on the total number of cracks and the blade crack feature index value. This invention, with equipment as the primary driver, automatically determines the blade crack damage level, reducing manual intervention and improving the efficiency of blade flaw detection testing.

The invention provides a packaging volume determination method, electronic equipment, a storage medium and a program product, and relates to the technical field of logistics. The method comprises the steps of obtaining a target image of an area where a to-be-packaged article is located, wherein the target image comprises a two-dimensional image and depth information corresponding to each pixel point in the two-dimensional image; based on the two-dimensional image, the depth information and camera parameters of a camera for collecting the two-dimensional image, determining a three-dimensional point cloud of the to-be-packaged article; based on the three-dimensional point cloud of the to-be-packaged article, a minimum surrounding frame of the to-be-packaged article is determined; performing rendering display on the target image and the minimum bounding box; and when it is determined that the displayed minimum surrounding frame is matched with the to-be-packaged article in the target image, the packaging volume of the to-be-packaged article is determined based on the volume parameter of the minimum surrounding frame. According to the invention, the speed of determining the packaging volume can be improved, and a user can perceive the determination process of the packaging volume.

The invention relates to the technical field of drawing, in particular to an automatic hole digging and filling method and device for a drawing and a storage medium. According to the method, each visible geometric element can be automatically read from the drawing to be filled, corresponding ItemType instance data are obtained, and a minimum bounding box is calculated; classifying the geometric elements according to the instance data, and setting the color and filling style of each class; according to the minimum bounding box and the geometric center point, judging whether the geometric element is an internal element of another geometric element or not; if yes, generating a hole digging area of external geometric elements; and according to the color and the filling style, filling the filling area or the hole digging area of the geometric elements to generate a filled drawing. It can be understood that according to the technical scheme, the hole digging area is jointly recognized through the minimum bounding box and the geometric center point, and the recognition precision is high; and during filling, rapid filling can be carried out according to colors and filling styles corresponding to the geometric elements, so that the plotting efficiency of a complex structure is remarkably improved.

The present invention relates to the technical field of image segmentation, and particularly relates to a target segmentation method based on multi-modal data fusion, and an image labeling tool. The method comprises: acquiring an initial RGB image, an initial infrared image and an initial depth image, each of which includes a target; performing alignment processing on the initial RGB image, the initial infrared image and the initial depth image, so as to obtain a first RGB image, a first infrared image and a first depth image in the same coordinate system; specifying an initial prompt point in the first RGB image, and respectively acquiring, by means of the initial prompt point, first masks representing target regions in the images of different modalities; on the basis of the first masks of the first RGB image, first infrared image and first depth image, fusing pixel values of the images of different modalities, so as to obtain a second mask; and on the basis of the second mask, determining a minimum bounding box for the target, and calibrating the minimum bounding box, so as to obtain a segmentation result of the target. The present invention significantly improves the precision and robustness of target segmentation by means of multi-modal data fusion and an iterative optimization strategy, and is applicable to a plurality of complex application scenarios.

System and method are provided for generating training data for feature matching among images of a building structure. The method includes obtaining a model of a building that includes a camera solution and images used to generate the geometric model. The method also includes, for facades of the model: applying a minimum bounding box to a respective facade to obtain a respective facade slice that is a 2-D plane represented in a 3-D coordinate system of the model; and projecting visual data of at least one camera in the camera solution that viewed the respective facade onto a visibility mask associated with the respective facade slice. The method also includes photo-texturing the projected visual data facade slice to one of the facade slices or the geometric model to generate a visual 3-D representation of the building; and generating a training dataset by perturbing the visual 3-D representation.

The application relates to the field of electric power, and discloses a substation control device terminal strip drawing paper identification method, device and equipment, which is used for converting a terminal strip drawing paper into a standardized intermediate file. The method comprises the following steps: acquiring an electronic image of the terminal strip drawing paper, and identifying spatial coordinates of connection points, spatial coordinates of connection lines, line types, line widths and spatial coordinates of short-circuit symbols; a plurality of character sequences and corresponding minimum bounding box coordinates are identified and extracted; a semantic binding relationship between the connection points and the character sequences is established based on the spatial coordinates, the spatial adjacent relationship of the connection points is determined in combination with a terminal strip arrangement rule, and a short-circuit relationship is identified in combination with the adjacent relationship and the spatial coordinates of the short-circuit symbols; the spatial coordinates of the connection lines, the line types and the line widths, the spatial coordinates of the connection points, the semantic binding relationship and the short-circuit relationship are integrated, and a topological relationship atlas is constructed; and elements and loops in the atlas are classified based on an electric power industry coding rule library, and are converted into the standardized intermediate file in combination with engineering modeling rules.