35 results about "Object contour" patented technology

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 relates to the technical field of sewing automation, and discloses an automatic sewing identification method and system based on image identification. The method comprises the steps of obtaining surface image data of a target sewing object, and performing feature extraction on the surface image data to generate a geometric contour data set; according to sewing parameter configuration information input by a user, executing path generation operation to obtain an initial sewing path coordinate sequence; based on spatial distribution characteristics of the geometric contour data set, performing path density adaptive adjustment on the initial path, and generating a corrected sewing path coordinate sequence; performing smoothness optimization processing on the corrected path to obtain a final sewing path coordinate sequence; and converting the final path coordinate sequence into visual graphic data, and rendering and outputting the visual graphic data in real time. According to the method, the contour of the sewing object can be accurately obtained, different sewing requirements can be flexibly met, the path density and smoothness are optimized, the accuracy and efficiency of sewing path generation are improved, the defects of a traditional method are overcome, and the method is suitable for various sewing processing scenes.

The invention discloses a TSTFP-Net-based semantic segmentation method for a low-illumination superposition rain image, and the method aims at an image segmentation scene shot in a weak light and rainy environment, and solves a problem that the semantic segmentation precision is reduced because of insufficient illumination, raindrop interference and a significant distribution difference between a source domain and a target domain. The method comprises the following steps: 1) processing an existing data set; 2) designing a semantic segmentation network based on low illumination and rain superposition of TSTFP-Net; 3) training the RIPLM network; 4) training a trunk segmentation network; 5) training a space domain and frequency domain double-teacher network; the method comprises the following steps of (1) training an FSFM module, (2) training a comparison learning module, (3) training a domain discriminator network, and (4) testing a semantic segmentation network with low illumination and rain overlay. The method can still accurately identify the contour and category of an object in complex weather, improve the domain generalization performance, reduce the domain gap, and further improve the precision of semantic segmentation in the low illumination and rain overlay environment.

The embodiment of the invention discloses a curve generation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a target point set; wherein the target point set at least comprises coordinate information of a first coordinate point of a target object contour; based on a pre-trained detection model, determining an index sequence of second coordinate points belonging to the boundary of the target object in the target point set; and determining a target boundary curve of the target object based on the index sequence. According to the technical scheme, the problems that a traditional curve generation method depends on manual parameter setting and is difficult to adapt to complex data distribution and the precision is difficult to guarantee are solved, high-precision automatic generation of the object boundary curve is achieved, and the technical effects of improving the curve generation accuracy and stability and enhancing the multi-scene applicability are achieved.

The application provides a warning zone identification method and device, electronic equipment and storage medium. The identification method comprises: performing yellow and white warning zone image preprocessing on a to-be-identified region image to determine a processed first image, and performing red warning zone image preprocessing on the to-be-identified region image to determine a processed second image; performing object contour extraction on the first image and the second image to determine a first object contour of the first image and a second object contour of the second image; performing preliminary contour screening processing and contour shape calculation processing on the first object contour and the second object contour to determine a candidate contour; wherein the candidate contour is a contour that meets the warning zone characteristics in the first object contour and the second object contour; and based on the number of pixel points of a target color in the candidate contour, the warning zone in the to-be-identified region image and the type of the warning zone are determined, thereby improving the accuracy of warning zone identification.

The invention discloses a 3DGS data acquisition method and related equipment. The method comprises the following steps: after responding to a data acquisition instruction, firstly initializing a camera of the electronic equipment, then calibrating an object contour range and a middle axis of a to-be-acquired object, determining the type of a two-dimensional object or a three-dimensional object to which the to-be-acquired object belongs, and then formulating a corresponding shooting path and shooting parameters based on the determined object type and object contour range; and finally, displaying the shooting path and the shooting parameters on a display screen of the electronic equipment so as to prompt a user to complete data acquisition of the to-be-acquired object according to the path and the parameters. By means of the object contour range, calibration of the center axis and judgment of the object type, an accurate object feature basis is provided for subsequent shooting, then an adaptive shooting path and shooting parameters are determined based on the features, related information is visually displayed through a display screen, and a user is effectively guided to standardize acquisition operation.

This invention discloses a phase-height mapping calibration method and apparatus in fringe projection, relating to the field of three-dimensional object contour measurement technology. The method includes: controlling a projector to project a set of phase-shifted sinusoidal fringe patterns onto a micromirror array with known height values; simultaneously triggering a camera to acquire fringe image sequences; performing phase shift calculations on the fringe image sequences to obtain a wrap-around phase map; constructing a global optimization model of absolute phase using the known height value of the micromirror array as an absolute constraint; solving for the absolute phase field corresponding to the height of the micromirror array; extracting the center coordinates, absolute phase values, and known height values ​​of each micromirror element; constructing a three-dimensional control point set; fitting the three-dimensional control point set; calculating the continuous model coefficient field; and forming a phase-height lookup table. This invention can significantly improve the real-time calculation efficiency of three-dimensional measurement, meeting the practical needs of high-precision, high-stability industrial three-dimensional shape measurement.

The invention discloses a 2.5 D convolution-based feature extraction method and system, and the method comprises the steps: carrying out the useless region cutting of an original 3D image, and obtaining a preprocessed 3D image; determining a plurality of slice directions according to the object contour of the original 3D image; according to the plurality of slice directions, segmenting the preprocessed 3D image into a plurality of 2D image sets; performing fusion processing on all the 2D images in the plurality of 2D image sets to obtain a 2.5 D image output set; multi-channel processing of a convolutional neural network is carried out on the 2.5 D image output set, and internal and external structural features of the object are extracted; and determining an abnormal region of the object according to the structural features. The 2.5 D image set is obtained by carrying out cutting preprocessing and multi-direction slice segmentation on the 3D image, non-time dimension correlation processing on the 3D image is achieved, the calculation complexity and the requirement for computing power and memory resources are reduced, and the image recognition efficiency and credibility are improved.

The application aims to provide a method and device for jet sorting objects. The application determines real physical coordinates of all points of an object contour according to pixel coordinates of the object; converts the real physical coordinates of all points of the object contour into time coordinates using a time stamp; calculates a range covered by each object on a corresponding time coordinate axis according to the real physical coordinates of all points of the object contour, a speed of a conveying belt and a position of a photographing position to a jet valve position; calculates a jet time and a jet position according to the range covered on the corresponding time coordinate axis; combines the jet position and the jet time; and controls the jet valve to jet sort according to the combined data. Thus, the object can be accurately positioned, the jet time can be calculated with low positioning complexity, the jet is accurate and the object can be effectively sorted.

The invention relates to a modeling machine, an arrangement, and a method for detecting a danger to a person in a current working area of ​​a manipulator during a volume change of an object by means of the manipulator, comprising the steps of selecting a first, in particular memory-stored, field data set from a plurality of field data sets based on a first current position of the manipulator, wherein the first field data set is in particular assigned a first predefined object contour, determining a current second object contour at the first position based on second sensor data from a second measurement by means of a sensor arranged indirectly and / or directly on the manipulator, detecting that the second object contour deviates from the first object contour and, in response thereto, outputting a signal which represents the danger to the person in the current working area of ​​the manipulator.

The invention discloses a pure machine vision distance measurement method and system, and relates to the technical field of environmental perception. The method comprises the following steps: a calibration stage: a plurality of concentric arc marks are preset on the flat ground, the marks take the projection of a preset reference point on the ground as the circle center, and the radius is the known distance; a first camera and a second camera are installed on a support with pitch angles capable of being synchronously adjusted at a fixed non-zero relative pitch angle difference, calibration image pairs are synchronously collected at a plurality of preset overall pitch angles, center lines of arc marks in the images are extracted, and a pixel-distance mapping relation pair library in one-to-one correspondence with the overall pitch angles is established. In the distance measuring stage, a first camera and a second camera are fixedly installed on a carrier according to the fixed non-zero relative pitch angle difference, a to-be-measured image pair is collected in real time, the contour of a target object is detected, all preset overall pitch angles are traversed, and a first mapping relation and a second mapping relation under the corresponding overall pitch angles are obtained from a mapping library respectively; and querying the minimum distance of the contour pixels to obtain a first candidate distance and a second candidate distance, selecting the overall pitch angle of the target according to the consistency of the first candidate distance and the second candidate distance, and determining a final distance measurement result based on the two candidate distances corresponding to the overall pitch angle of the target. Through double-view multi-overall pitch angle calibration and consistency matching, the method is adaptive to complex three-dimensional terrains, does not need camera internal reference, does not depend on target priori, does not need any active sensor, and realizes high-precision, low-cost and low-calculation-amount real-time distance measurement.

The invention provides an end-to-end target object frame extraction method and device, equipment and a medium. The method comprises the following steps: marking an original remote sensing image by using a real frame to obtain sample data, and training a constructed target object frame extraction model by using the sample data to obtain a trained target object frame extraction model; then, testing the trained target object frame extraction model to obtain a final target object frame extraction model; and finally, processing a remote sensing image to be extracted by using the final target object extraction model, and outputting a corresponding target object frame prediction result, so that an accurate target object contour in the remote sensing image to be extracted can be obtained, and the obtained target object frame prediction result is subjected to geographic data marking or analysis.

The application discloses a rolling mill tapping bending state detection method and device, equipment and medium, relates to computer technology field, including determining each single frame image by using video stream, generating reference object image, carrying out coordinate calculation, obtaining reference object contour pixel coordinates and rolling piece contour pixel coordinates; generating reference object axis pixel coordinate point set and rolling piece axis pixel coordinate point set, generating rolling piece axis fitting equation and reference object axis fitting equation; calculating the rolling piece axis fitting equation, obtaining the bending measure, carrying out coordinate conversion on the reference object axis pixel coordinate point set, obtaining the plane size coordinates, calculating the rolling piece chord height according to the plane size coordinates and the bending measure; based on the rolling piece axis fitting equation, the reference object axis fitting equation, the bending measure and the rolling piece chord height, the rolling mill tapping bending state detection result is generated. The application can reduce the labor intensity of artificial visual inspection, reduce the production safety accident rate, improve the rolling mill tapping product quality and efficiency.

The application discloses a plane target contour reconstruction method and device suitable for geographic information data processing, and is used for recovering a contour shape of a target geographic element from an unordered discrete point set. The method comprises the following steps: acquiring a plane discrete point set representing the target geographic element, wherein the plane discrete point set is unordered coordinate data; processing the plane discrete point set, and structuring the plane discrete point set into a line segment set group representing local contour features of the plane discrete point set through an iterative splitting and merging operation based on an ellipse model, wherein the splitting and merging operation is controlled by an adaptively generated linearity threshold and a disconnectedness threshold; processing the line segment set group, and integrating the line segment set group into closed polygon contour data representing the target geographic element by establishing a spatial index for end points of the line segment set group and performing a nearest neighbor connection; and outputting the closed polygon contour data. The method and device improve the contour reconstruction accuracy, automation degree and algorithm robustness.

The invention relates to the technical field of image processing, in particular to a target object contour marking method based on a large model and display equipment. According to the embodiment of the invention, the to-be-processed image is acquired, the target object in the to-be-processed image is determined, the contour of the target object in the to-be-processed image is determined based on the large model, and the contour of the target object is marked according to the sparse degree of the stored marking points and is displayed. According to the method, the contour of the target object does not need to be drawn manually in detail, only the contour of the target object in the to-be-processed image needs to be determined based on the large model, and then the contour of the target object is marked according to the sparse degree of the pre-stored marking points, so that the efficiency of marking the contour of the target object is improved.

The present application relates to the technical field of spectral detection, in particular to a method and system for agricultural product quality detection based on big data, which obtains a sequence of spectral image frames of a plurality of continuous detection objects on a conveyor belt, performs detection object contour segmentation on the spectral image frames and divides them into a plurality of detection area bands, calculates the standard deviation of the mean value of the spectral reflectance intensity of each area band and the maximum value ratio as an edge distortion value, determines a first frame distortion coordinate and a second framing coordinate through the edge distortion value of the distorted object obtained from the edge distortion value, calculates the corresponding time difference and obtains the optimal value through neural network training. The present application solves the problem of repeated detection and waste of processing resources caused by edge distortion of the hyperspectral camera in continuous detection, and improves the online detection efficiency.

The application discloses a contour extraction method and system based on optimized clustering and fractional order error back propagation network, relates to the technical field of contour extraction, and comprises the following steps: acquiring sampling points, generating the boundary of a convex hull by using the sampling points, and calculating the points in the convex hull according to the convex hull boundary; calculating a Voronoi diagram and constructing a convex polygon according to the points in the convex hull and the sampling points, judging whether the Voronoi diagram is stable, re-calculating the convex hull until the Voronoi diagram is stable, extracting Voronoi points if the Voronoi diagram is stable, and calculating a Voronoi circle by using the Voronoi points to find an inflection point; calculating MinPts by using the inflection point, finding cluster points, calculating the contour coefficients of the cluster points, determining optimal cluster points according to the contour coefficients of the cluster points; judging whether the optimal cluster points converge, re-calculating the convex hull if the optimal cluster points do not converge, and obtaining a data sequence if the optimal cluster points converge; and training the data sequence in a pre-established initialization fractional order error back propagation model to output a final smooth object contour.

The application provides a light field saliency object detection method and device based on implicit graph learning, and the method comprises the following steps: inputting a full-focus image to be tested and a focal stack image to be tested into a pre-trained double-flow deep convolution network to obtain a full-focus feature map and a focal stack feature map; inputting the full-focus feature map and the focal stack feature map into a pre-trained contour perception model to obtain a saliency prediction result and an object contour prediction result. By inputting the full-focus image to be tested and the focal stack image into the pre-trained double-flow deep convolution network, the application promotes the mining of saliency features, and by inputting the full-focus feature map and the focal stack feature map into the pre-trained contour perception model, the application can effectively aggregate information of mixed scales, and then obtain an accurate light field saliency object detection result.

The invention relates to the technical field of image communication, and discloses an edge end semantic perception video compression and transmission method and device and electronic equipment, and the method comprises the steps: obtaining a to-be-transmitted video frame, a semantic mask matrix, semantic confidence data and transformation residual data; performing physical alignment on the semantic mask matrix and a maximum coding tree unit grid of a video encoder; identifying a target grid in an edge crossing state by calculating a pixel space variance metric value in the grid; sub-coding unit self-adaptive division based on semantic gradient is carried out on the target grid, and a sub-coding unit set matched with the object contour is generated; according to the method, the foreground contour sharpness is locked through a semantic edge pre-division mechanism, coding prediction redundancy is eliminated by using the priority score, and the coding prediction accuracy is improved. And realizing correlation between local space-time physical entropy change and semantic importance.

The invention discloses a three-dimensional object morphology measurement method based on a Newton-like ring, and relates to the technical field of phase shift interference measurement. The method comprises the following steps: constructing a conical lens Newton ring-like model by using a conical lens and plate glass; single-color fringe patterns with different wavelengths are obtained by using incident light with different wavelengths; the air film thickness is used for calculating the optical path difference to obtain the phase change, the wavelength modulation phase shift technology is used for controlling the phase deviation by changing the wavelength of the incident light, the interference light intensity of the incident light with different wavelengths is obtained, and the phase change controlled by wavelength modulation is obtained by using the interference light intensity of the incident light with different wavelengths and the corresponding phase deviation; and calculating the object plane modulation phase of the measured object based on the phase change difference before and after the measured object is placed, and obtaining the object contour information of the measured object. According to the method, phase shift is modulated by adjusting the wavelength of incident light, vibration interference and step length calibration errors caused by mechanical phase shift can be directly avoided, and control is simple.

The embodiment of the invention discloses an object contour display method and device in a panorama, and the method comprises the steps: obtaining a three-dimensional model of a target object in a preset scene under a world coordinate system; determining a conversion matrix between a panoramic cube chartlet of a panoramic image corresponding to the preset scene and the world coordinate system; performing coordinate alignment on the three-dimensional model and the panorama based on the conversion matrix, and embedding the three-dimensional model into the panorama; rendering the three-dimensional model based on color information in the panorama to obtain a target three-dimensional model embedded in the panorama; performing dynamic lighting effect rendering on the target three-dimensional model, and displaying at least one geometric surface contour in the target three-dimensional model; according to the embodiment, each point position does not need to be identified respectively, and the object identification cost is reduced.

An image augmentation method and device, the method comprising: extracting a first object contour by a first image mask, wherein the first object contour corresponds to a first label; superimposing the first object contour to a superimposition region in a second image mask according to an augmentation parameter to generate a third image mask, wherein the augmentation parameter comprises a contour scaling parameter, a contour moving distance, a contour rotation angle, and a superimposable range, the third image mask containing the first object contour and a second object contour in the second image mask; and generating a corresponding sample image according to the first object contour and the second object contour in the third image mask by a generative adversarial network model for machine learning. Thus, the problem of requiring a large amount of manpower to collect a large number of samples for labeling a large number of samples in the preprocessing stage can be effectively solved.

The application discloses a picture deep learning preprocessing method based on feature compression, and relates to the technical field of computer vision and image processing.The preprocessing method comprises the following steps: frequency domain decomposition, parameter dynamic adjustment compression encoding and reconstruction of an approximate original picture.The application has the advantages that: through frequency domain decomposition and selective compression, while the core visual information (such as object contour and texture distribution) of the picture is retained, the calculation burden of the deep learning model is greatly reduced; compared with traditional spatial domain down-sampling, the problem of global feature loss is avoided; the application is especially suitable for resource-limited scenes such as mobile terminals or edge devices; and the deployment efficiency and practicability of a lightweight model are significantly improved.

This invention discloses a method and apparatus for detecting the tilt of finished grain stacks based on computer vision, belonging to the field of grain depot management technology. The method includes: using a binocular vision camera as the measuring device and performing stereo calibration; acquiring left and right images of the stack at fixed points and times; using deep learning-based stereo matching technology to obtain disparity maps of the left and right images, and acquiring image depth information and point cloud information; comparing the point cloud information of the current stack image with the point cloud information of the reference image, and determining whether the stack is tilted based on the comparison result. This invention employs a domain adaptive technique based on color distribution transfer, enabling the model to maintain efficient feature extraction capabilities even under complex lighting conditions in grain depots. It can be deployed without manual annotation, reducing application costs. The invention introduces a residual network to extract multi-scale features and combines it with an adaptive aggregation module, resulting in clearer object contours in the output disparity map and improved detection accuracy.

This application discloses a grape segmentation method based on multi-source feature fusion, belonging to the field of image processing technology. It establishes an instance segmentation network, MFFNet, based on multi-source feature fusion. MFFNet collaboratively utilizes powerful semantic features extracted by the Swing Transformer and explicit boundary information generated by parallel edge branches. ASFF adaptively fuses multi-scale information for grape segmentation, employing a staged feature fusion strategy to perform targeted feature fusion at different stages, improving the accuracy and robustness of grape segmentation in complex scenes. MFFNet effectively solves the segmentation challenges of dense grape berries, occlusion, and blurred boundaries in complex field environments. The staged feature fusion strategy enhances the model's ability to perceive and depict object contours, effectively addressing the problems of instance adhesion and unclear boundaries.

The application discloses a positioning method and a positioning device. The method comprises the following steps: collecting a depth map and preprocessing, removing outliers based on depth gradient smoothness analysis, and inferring and interpolating the object contour through gradient direction; extracting accurate single-pixel edges based on gradient features; converting the depth map into a point cloud and performing denoising correction; on the basis of traditional geometric centers, introducing multi-objective optimization of grabbing stability, upgrading the positioning center to an optimal grabbing target point considering contact area, gravity center load, surface flatness and clamping force constraints, and more suitable for high-precision stable grabbing of small objects without surface features. The whole process does not depend on object surface texture, gray or contour features, and only uses depth gradient geometric properties as the positioning basis. The application realizes high-precision and high-robustness three-dimensional positioning of objects without surface features, effectively improves the industrial grabbing success rate, and is suitable for automatic sorting and assembly scenes of micro electronic components, precision mechanical parts and the like.

The invention discloses a multi-scale mixed hierarchical bounding box collision detection method driven by a large model. According to the method, firstly, semantic perception is realized by using an SAM large model, potential collision objects in a scene are automatically recognized, unnecessary geometric calculation is reduced, meanwhile, a collision detection scale division standard is determined according to user prompt word text semantics, and collision object contour data and relative distance data obtained through screening are combined, so that a collision detection result is obtained. Determining a collision detection scale in the scene; and then aiming at different collision detection scales, different modes are adopted to construct a collision object hierarchical bounding box structure, and different detection algorithms are adopted to carry out collision detection, so that the real-time performance of the collision detection is ensured, the detection accuracy is improved, the effective balance between the collision detection precision and the real-time performance under different scales is realized, and the collision detection accuracy is improved. And the dependence of the collision detection process on high computing power is reduced.

The application discloses an image generation method and device, electronic equipment and computer readable medium. The method comprises the following steps: acquiring first object contour description data, first object posture description data and second object posture description data, wherein the first object contour description data and the first object posture description data are determined according to a first object description image; correcting the first object contour description data according to difference representation data between the two posture description data to obtain corrected contour description data; and performing image generation processing according to the corrected contour description data and the second object posture description data to obtain a generated image, so that the posture of the object described in the generated image is consistent with the posture of the object described in the second object posture description data, and the generated image is consistent with the first object description image in at least one object feature except the posture of the object, thereby improving the image generation effect.

The invention relates to a scanning image histogram conjoint analysis method and system. The method comprises the following steps: acquiring a user scanning image; determining a target scanning image; expanding the marked contour of the object of interest to obtain an expanded object; generating a first gray histogram of the interested object and a second gray histogram of the extended object, and obtaining quantitative parameters; and comparing the quantitative parameter with the quantitative parameter template corresponding to each classification to determine the classification, and sending the classification as a conjoint analysis result to the user terminal. According to the method, hierarchical scanning images are fully utilized, auxiliary information contained in each layer of scanning images is fully utilized in a weight decreasing mode, and the conjoint analysis efficiency is greatly improved.