42 results about "Edge strength" patented technology

The invention discloses a weldment welding seam automatic detection method and device based on machine vision, and relates to the technical field of machine vision intelligent detection. The weldment welding seam automatic detection method and device based on machine vision comprises the steps that S1, surface images and forming feature data of a weldment are collected and preprocessed to construct a standardized image feature data set; s2, the boundary clearness of the weld joint is evaluated by combining the edge strength and the contour coherence, and the main contour extraction range is dynamically adjusted; s3, analyzing abnormal focusing characteristics of the candidate area, and adjusting a defect labeling range and a detection priority; and S4, integrating the boundary definition and the abnormal focusing features, analyzing the structure abnormality, and dynamically controlling and verifying a resource allocation strategy. The problems that in the weldment detection process, obvious light reflection and texture blurring phenomena exist in a heat affected area at a weld joint, a traditional image enhancement and edge extraction algorithm is difficult to stably recognize microdefects, and the credibility of a detection result is reduced are solved.

The invention provides a three-dimensional map construction and camera trajectory estimation method based on edge information fusion, and belongs to the field of computer vision and robot perception. Collecting an image sequence in a scene, executing edge detection and then generating a two-dimensional edge intensity graph; initializing or updating a three-dimensional Gaussian point cloud: generating an initial point cloud by a first frame through color, depth and edge intensity values of the first frame through back projection; point clouds are newly added to the required areas of subsequent frames according to rendering differences; initializing a current frame of camera pose, generating a prediction map based on the pose and a previous frame of point cloud rendering, and constructing a loss function to optimize the camera pose; rendering is carried out again based on the optimized pose and the point cloud of the previous frame, a loss function is constructed to optimize the point cloud attribute, and the point cloud of the current frame is generated; and processing the sequence frame by frame, accumulating and optimizing the point cloud to form a global three-dimensional map, and accumulating the camera pose to output the camera motion trail. According to the method, the SLAM precision, stability and reconstruction quality in various complex scenes, especially in low-texture and strong-structure environments are improved.

The application provides a three-dimensional map construction and camera trajectory estimation method based on edge information fusion, and belongs to the field of computer vision and robot perception. An image sequence in a scene is collected, and a two-dimensional edge intensity map is generated after edge detection; a three-dimensional Gaussian point cloud is initialized or updated: an initial point cloud is generated by back projection of the color, depth and edge intensity value of the first frame; a point cloud is added in the required area according to the rendering difference in subsequent frames; the camera pose of the current frame is initialized, a prediction map is generated based on the pose and the point cloud of the previous frame, a loss function is constructed to optimize the camera pose; the point cloud attributes are optimized again based on the optimized pose and the point cloud of the previous frame, and the point cloud of the current frame is generated; the sequence is processed frame by frame, the optimized point cloud is accumulated to form a global three-dimensional map, and the camera motion trajectory is output by accumulating the camera pose. The application improves the SLAM precision, stability and reconstruction quality in various complex scenes, especially in low-texture and strong-structure environments.

The invention belongs to the technical field of image data processing, and particularly relates to an industrial robot efficient communication method and system based on edge computing, and the method comprises the steps: determining all narrow edge pixel points of each frame of image; determining an edge intensity factor of each frame of image; determining a structured edge factor of each frame of image; determining an optimization coefficient of an effective frame screening threshold value of each frame of image to obtain an optimized effective frame screening threshold value of each frame of image; and each frame of image is judged, and effective frames of the image are screened. According to the method, the gradient magnitude and gradient direction features of the narrow edge pixel points of each frame of image are analyzed, the effective frame screening threshold value of each frame of image is optimized, the limitation of the fixed effective frame screening threshold value is overcome, the effective frame images are screened, and the timeliness and robustness of industrial robot communication are improved.

The application discloses a kind of indexable turning inserts, including center line, upper surface, lower surface and multiple side surfaces connecting upper surface and lower surface, two adjacent side surfaces are connected by circular arc surface, the height of each relative reference surface of circular arc cutting edge and main cutting edge is consistent, the height of each relative reference surface of transition cutting edge gradually reduces along the direction away from circular arc surface, the height of each of auxiliary cutting edge is consistent with the lowest point of transition cutting edge;The height of first protruding portion top surface, second protruding portion top surface and mounting positioning surface relative reference surface increases in turn;Upper surface is provided with third protruding portion symmetrically distributed about angle bisector, third protruding portion is located on third rake surface and is connected with second protruding portion.The present indexable turning insert has the advantages of facilitating to guide chip lateral curling, considering the edge strength of nose region, controllable chip shape and chip flow direction, reducing friction, avoiding chip blocking or chip holding and the like.

The application provides a dynamic anti-aliasing rendering method and system based on edge detection and motion perception, and relates to the technical field of computer graphics rendering. The method comprises the following steps: performing 8-direction edge intensity detection on a pixel point in an original scene image by using a 3*3 sampling matrix and calculating an edge intensity value; calculating a motion factor and dynamically adjusting an edge detection threshold; if the edge intensity value of the pixel point is greater than the edge detection threshold, performing adaptive blur processing based on a 9-point weighted average algorithm to generate a blurred pixel value, calculating a mixing factor, mixing the original pixel value of the pixel point and the blurred pixel value to generate a final pixel value; if the edge intensity value of the pixel point is less than or equal to the edge detection threshold, keeping the original pixel value of the pixel point unchanged, and the final pixel value is the original pixel value, and finally outputting a final scene image by collecting the final pixel value, so that the anti-aliasing processing strategy can be adaptively adjusted according to the static and dynamic characteristics of the scene, the rendering performance is optimized while the image quality is ensured.

The invention discloses an image processing method and device, equipment, a storage medium and a computer program product, and belongs to the technical field of image processing. The method comprises the following steps: mapping a target pixel point in a target image to an original image to obtain a coordinate of a target mapping point; determining a plurality of reference pixel points from the original image based on the coordinates of the target mapping point; determining the edge intensity of the target pixel point based on the gray values of the plurality of reference pixel points; based on the edge intensity of the target pixel point and the distances between the multiple reference pixel points and the target mapping point, weights corresponding to the multiple reference pixel points are determined; and determining the pixel value of the target pixel point based on the pixel values of the plurality of reference pixel points and the weights corresponding to the plurality of reference pixel points. As the edge intensity can well distinguish the image edge and noise, the weight of each reference pixel point is adjusted by adding the edge intensity of the target pixel point, the image noise can be reduced, the image edge can be enhanced, and the image contrast can be improved.

The invention belongs to the technical field of computer videos, and particularly relates to a monitoring video stream distortion correction and two-dimensional-to-three-dimensional fusion method and a monitoring video stream distortion correction and two-dimensional-to-three-dimensional fusion system. The non-uniform mapping relation is established by calculating the distortion offset of the pixel coordinates, and the interpolation weight is dynamically calculated based on the pixel distance and the local edge intensity, so that the non-uniform distortion of the lens can be accurately corrected, the edge detail features of the image are effectively reserved, image blurring or detail loss in the correction process is avoided, and the correction efficiency is improved. And the definition and the accuracy of the corrected image are greatly improved.

The invention relates to a water body rendering method and device, equipment, a storage medium and a program product, and the method comprises the steps: superposing periodically continuous waveform data and noise data to all vertexes of a horizontal grid, and obtaining a wave pattern; coloring the wave white mold to obtain a wave model; respectively carrying out derivation processing on the waveform data and the noise data at each vertex in the wave pattern to obtain a first space gradient corresponding to the waveform data and a second space gradient corresponding to the noise data; performing composite processing on the first space gradient and the second space gradient to obtain a composite gradient of each vertex; obtaining a vertex normal vector of each vertex based on the composite gradient of each vertex; based on the vertex normal vector and the line-of-sight direction vector of each vertex, generating edge intensity data of the wave contour; and superposing the edge strength data to the wave model to obtain a water body rendering result. According to the scheme, the vertex normal vector is dynamically determined based on the composite gradient, so that the rendered water body has a dynamic surge effect.

The application provides an image post-processing optimization management method for intelligent taking of a cabinet of electric power materials, and belongs to the technical field of electric power material management, and comprises the following steps: S1, taking and placing registration; S2, overlay image acquisition; S3, feature extraction; S4, recognition matching; S5, inventory management whole process and the like; the application collects multi-angle images of the materials through a high-resolution camera, constructs a multi-directional convolution framework by using a spatial gradient energy kernel, and strengthens the extraction capability of geometric contour, surface texture and identification features. The core technology adopts three-dimensional differential operation to analyze the edge strength, curvature change and deformation characteristics step by step, combines a multi-scale texture analysis and an adaptive light compensation algorithm, eliminates environmental interference and generates a unique feature coupling factor, matches a pre-stored feature fingerprint library in real time through an elliptical envelope detection algorithm, accurately identifies the types and quantities of the materials, and forms a closed-loop management from permission verification, accurate identification to intelligent early warning.

This invention relates to the field of hydropower station technology, and discloses a method, apparatus, equipment, and medium for detecting crack defects in hydropower station dams. The invention calculates the maximum value of the ridge variation coefficient through single-layer discrete two-dimensional wavelet decomposition and ridge transform, and then determines the true crack edge direction based on this maximum value, improving the edge feature enhancement effect and making the edge boundaries clearer. Furthermore, a crack edge image is constructed based on multiple pixels corresponding to the absolute values ​​of multiple gradients along the crack edge direction, and the crack defect detection result of the hydropower station dam to be detected is determined based on the constructed crack edge image. Therefore, by implementing this invention, the edge strength and recognition rate of image recognition are improved by enhancing edge features, thereby improving the accuracy of crack defect detection results.

The invention provides a dynamic anti-aliasing rendering method and system based on edge detection and motion perception, and relates to the technical field of computer graphic rendering, and the method comprises the steps: employing a 3 * 3 sampling matrix to carry out 8-direction edge intensity detection on pixel points in an original scene image, and calculating an edge intensity value; calculating a motion factor and dynamically adjusting an edge detection threshold value; if the edge intensity value of the pixel point is greater than the edge detection threshold value, performing adaptive fuzzy processing based on a nine-point weighted average algorithm to generate a fuzzy pixel value, calculating a mixing factor, and mixing the original pixel value of the pixel point and the fuzzy pixel value to generate a final pixel value; if the edge intensity value of the pixel point is smaller than or equal to the edge detection threshold value, the original pixel value of the pixel point is kept unchanged, the final pixel value is the original pixel value, the final pixel value is summarized, and a final scene image is output, so that the anti-aliasing processing strategy can be adaptively adjusted according to the static and dynamic characteristics of the scene, and the image quality is improved. And the rendering performance is optimized while the image quality is ensured.

The application discloses a kind of ice thickness measurement method and device based on visual perception of ice storage coil, and the method comprises: setting annular or fan ring-shaped calibration disc on ice storage coil.The calibration disc is provided with multiple different color stripe regions arranged concentrically and along radial direction.Judge whether each stripe region is completely covered by ice layer according to the brightness change of calibration disc covered by ice layer in image, and each stripe region is divided into completely covered interval, partially covered interval and uncovered interval.For partially covered interval, the ice layer covering ratio of partially covered interval is obtained, and thus the ice layer thickness is obtained.The application provides the accuracy of ice thickness detection by stage ice thickness analysis, specifically, the ice layer completely covered by ice layer is determined by using the segmentation and brightness of different color stripe regions, and then for the partially covered stripe region, the comprehensive edge strength of integrated gradient edge strength and phase consistency edge strength is used for coverage ratio calculation.

The present application discloses a method and apparatus for sample adaptive compensation in video coding. The method comprises: obtaining coding depth information and quantization information of a video coding unit, and calculating a boundary information value based on the coding depth information and the quantization information; skipping the sample adaptive compensation calculation if the boundary information value is less than a preset threshold; calculating edge strength values ​​for different preset directions using an edge direction estimation algorithm if the boundary information value is not less than the preset threshold, and screening boundary compensation modes based on the edge strength values; and traversing preset categories in the screened boundary compensation modes to perform sample adaptive compensation. This solution optimizes the sample adaptive compensation mechanism and improves overall coding efficiency and speed.

The invention relates to a method for efficiently producing ultra-thin elements (2, 4) of glass-based material, the material thickness (d) of which is between 5 m and 100 m, into which filiform damages (14) are introduced by irradiation with an ultra-short laser pulser, the distance (A) between adjacent filiform damages (14) being between 0.4 m and 4.0 m. This advantageously achieves high edge strength. The invention also relates to ultrathin elements (2, 4) of glass-based material and to the use thereof.

The application provides a sub-pixel edge positioning method based on conversion intensity and stable edge domain, comprising: S101, performing edge detection on an original image to obtain an edge detection result; S102, traversing each edge pixel based on the edge detection result to determine whether each edge pixel is in a current stable edge domain; S103, if the edge pixel is not in the current stable edge domain, obtaining a stable discrete direction sequence of the edge pixel, expanding the discrete direction sequence along a tangent direction to obtain an expansion result; S104, updating the current stable edge domain based on the expansion result to obtain an updated stable edge domain, repeating steps S102-S104 to obtain a standard stable edge domain; S105, solving effective statistical parameters in the standard stable edge domain; and S106, performing calculation based on a conversion relationship between the statistical parameters and an integral of an edge intensity curve and a step function under a continuous space to obtain a sub-pixel coordinate. The method improves positioning precision on the basis of improving operation efficiency.

The invention relates to a method for producing ultrathin elements and elements made of glass-based material (2, 4) and their use, in particular made of glass and / or glass-ceramic, with a material thickness (d) of 5 to 100 µm, using at least one ultrashort pulse laser (16) with a laser wavelength at which the glass-based element is at least substantially transparent, wherein the beam profile of the laser beam (20) is shaped such that it has a laser preferred direction LVR. This introduces defects (14) into the ultrathin glass-based element (2, 4) along a parting line (12, 121, 122), wherein adjacent defects (14) have a spacing (A) of 15 µm to 50 µm. This results in particularly good edge strength.

An electronic apparatus performs a method of decoding video signal, including: receiving, from the video signal, a picture frame that includes a first component, and a second component; selecting an edge direction across a collocated sample of the first component or a current sample of the second component relative to the respective sample of the second component; calculating a first edge strength determining a difference between values of two samples from the group consisting of the collocated or current sample, a first neighbouring sample, and a second neighbouring sample; quantizing the first edge strength into M segments according to M−1 threshold values; determining the classifier for the respective sample of the second component based on the M segments; determining a sample offset for the respective sample of the second component according to the classifier; and modifying the respective sample of the second component based on the determined sample offset.

The invention discloses a three-dimensional point cloud feature detection method and device, equipment, a medium and a product, and relates to the technical field of point cloud processing, and the method comprises the steps: obtaining ordered point cloud data which are three-dimensional point cloud data which are orderly arranged in an X direction and a Y direction; determining a first distance between the target point and a previous point of the target point in the X direction in the ordered point cloud data; according to the first distance, determining edge strength corresponding to the target point; according to the edge strength and a preset edge threshold range, determining edge points in the ordered point cloud data, the edge points being target points whose edge strength belongs to the preset edge threshold range; and determining three-dimensional geometric features according to the edge points. According to the method, the efficiency of determining the edge points in the three-dimensional point cloud can be improved, so that the detection speed of the three-dimensional geometric features is improved, and the applicability of a three-dimensional scanning detection technology in high-speed detection is ensured.

This invention belongs to the field of image data processing technology, specifically relating to an efficient communication method and system for industrial robots based on edge computing. The method includes: determining all narrow-edge pixels in each frame of an image; determining the edge intensity factor of each frame; determining the structured edge factor of each frame; determining the optimization coefficient of the effective frame filtering threshold for each frame, thus obtaining the optimized effective frame filtering threshold for each frame; and judging each frame to achieve effective frame filtering. This invention optimizes the effective frame filtering threshold for each frame by analyzing the gradient magnitude and gradient direction characteristics of the narrow-edge pixels in each frame, overcoming the limitations of a fixed effective frame filtering threshold, and filtering effective frames to improve the timeliness and robustness of industrial robot communication.

A method for separating a substrate of a brittle-hard material is provided. The method includes the steps of introducing defects into the substrate at a spacing from one another along a separation line using at least one pulsed laser beam; selecting an average spacing between neighboring defects and a number of laser pulses for generating a respective defect such that a breaking stress (σB) for separating the substrate along the separation line is smaller than a first reference stress (σR1) of the substrate and such that an edge strength σK of the separation edge obtained after separation is greater than a second reference stress (σR2) of the substrate; and separating the substrate after introducing the defects by applying a stress along the separation line.

A high-resolution remote sensing image edge repairing method based on tracking, first, a deep convolutional neural network is designed according to the remote sensing edge detection task, samples are made and edge detection model and edge tracking model are trained; the remote sensing image is input into the edge detection model to extract the ground feature edge, so as to obtain the ground feature edge intensity map. The remote sensing image and the random starting point file are input into the edge tracking model to iteratively generate the edge tracking graph, and for the discontinuous ground feature edge, the edge detection result and the edge tracking result are fused to automatically determine the edge width and complete the edge. The method completes the broken line repairing work through the topological information provided by the edge tracking algorithm for the discontinuous edge part in the prediction result of the edge detection algorithm, and simultaneously uses the edge binary graph to assist in connecting the broken edge tracking graph, or creates an edge in the position where the edge does not exist to complete the edge tracking graph. The method can improve the edge extraction accuracy and efficiency, and improve the accuracy of the edge tracking model.

Methods and devices are provided for video coding. In one method, a decoder may obtain an edge component syntax element that indicates a first component for classifying a second component, where the first component and the second component are different components of a plurality of components in a picture frame. Furthermore, the decoder may obtain a classifier for a current sample of the second component according to the edge component syntax element, an edge direction, and one or more edge strengths of a collocated sample of the first component, obtain a class corresponding to the current sample of the second component by classifying the current sample of the second component according to the classifier, and obtain a sample offset for the current sample of the second component according to the class and modify the current sample of the second component based on the sample offset.

The invention relates to the technical field of geological exploration and rock core analysis, in particular to a drilling slag sample recognition statistical method and system based on image recognition and a storage medium. The main color feature, the color uniformity degree, the texture contrast and the edge strength of each row of pixel points serve as feature vectors, a real layering boundary is screened out through feature vector weighted scoring and local maximum value detection, layering boundary optimization is achieved in combination with a minimum thickness threshold value, the risk of layering misjudgment and superfine segmentation is remarkably reduced, and the layering accuracy is improved. The accuracy and the stability of layered discrimination are comprehensively improved; according to the method, the physical thickness of each layered interval is further associated with the multi-dimensional feature vector sequence of all rows in the interval, quantitative statistics and digital expression of the stratum structure are achieved, and a high-quality data basis is provided for stratum internal structure analysis and rock core archiving.

The application discloses an image scaling method based on edge perception and content adaptive interpolation. Firstly, edge detection is performed on an input original image to obtain edge strength representing the richness of image details. Secondly, according to the edge strength, the distribution of the two values with the highest weights in an interpolation function is adaptively adjusted. For a position with higher edge strength, a larger weight distribution is adopted to better preserve edge and detail features. For a position with lower edge strength, a smaller weight distribution is adopted. Finally, pixel resampling of the whole image is completed by using the interpolation kernel after adaptive adjustment, and an output image with a target resolution is generated.

The present application belongs to the technical field of image data processing, and particularly relates to a gas pipeline root welding quality detection method and system based on image recognition, which comprises the following steps: performing median filter pretreatment on the pipeline root image to remove noise points; analyzing the edge positioning strength of each point based on the second derivative of the pixel point and the process weight, and calculating the average edge strength of the sampling path; combining the physical distance of each pixel point and the energy dissipation coefficient to correct the light path loss, and obtaining the root welding degree; using the root welding degree, the average edge strength and the hazard sensitivity based on the pipeline design pressure to obtain the hazard score; and sending a brake instruction to the robot in response to the score exceeding the limit. The present application introduces physical compensation and pressure equal mapping mechanism, realizes quantitative evaluation and closed-loop control of the weld quality, and improves the detection accuracy.

This invention discloses an optimization device and method for laser spot edge diffraction, relating to the field of optical instrument technology. The optimization device includes two sets of guiding mechanisms, two sets of driving mechanisms, and two sets of aperture pairs. Each aperture pair includes two opposing aperture plates, the edges of which are used to define the light transmission boundary and have a serrated structure. The driving mechanisms drive the two aperture plates in the same set to move synchronously in opposite directions along the guiding direction. The positions of the two aperture pairs are independently controlled to combine and form different light transmission boundaries. In the optimization method, the positions of the two aperture pairs are adjusted according to the edge morphology of the laser beam to be optimized, and the laser beam is made to pass through the target light transmission boundary for edge intensity modulation. Through the technical solution of this invention, the edge diffraction caused by traditional straight-edge apertures is reduced, and the edge smoothness of the laser beam, the flexibility of light transmission boundary adjustment, the adaptability to different spot morphologies, the shaping consistency, and the processing stability are improved.

The application discloses an image fusion method, device and computer readable storage medium. The image fusion method comprises the following steps: extracting high-frequency components and low-frequency components of a first image to obtain first high-frequency components and first low-frequency components; extracting high-frequency components and low-frequency components of a second image to obtain second high-frequency components and second low-frequency components; wherein the resolution of the first image is lower than that of the second image; determining the edge intensity of each pixel position in the second high-frequency components, and determining the local difference measure value of each pixel position in the high-frequency components of the first image and the second image; performing pixel-by-pixel weighted fusion on the first high-frequency components and the second high-frequency components to obtain fused high-frequency components; and obtaining a fused image based on the fused high-frequency components and the fused low-frequency components. Through the above method, the application can solve the problem of loss of edge details in the fused image when the current image fusion method is used for image fusion.

The application discloses a binocular depth estimation method based on scene structure information reinforcement and monocular-binoctamer collaborative optimization, which comprises the following steps: acquiring left and right view images, obtaining left view monocular depth prior and left and right view multi-scale scene structure features; converting the features into matching features, combining shallow texture features to perform high-frequency boundary enhancement, and constructing a binocular matching cost volume; modeling cross-view high-frequency response of middle and high layer features, generating conditional features to guide geometric aggregation, and obtaining a geometric coding volume; obtaining initial parallax and matching confidence based on the geometric coding volume, fusing high-frequency edge strength, matching confidence, spatial structure attention and current binocular matching residual to generate multi-source dynamic attention, and guiding selective gated recurrent unit (GRU) update; and through three-stage collaborative iteration, upsampling and residual fine-tuning, a final parallax or depth result is obtained. The application can improve the accuracy and stability of binocular depth estimation.

Disclosed is a method for generating a panning image. The method includes receiving an image frame comprising one or more objects and a background portion, identifying, in the received image frame, a first area corresponding to a salient object amongst the one or more objects, generating a histogram map based on the received image frame and the first area, the histogram map indicates a distribution and orientation of edge intensities corresponding to the salient object in the received image frame, estimating, using a set of models, a speed parameter and a direction parameter corresponding to the salient object based on the received image frame, the first area, and the histogram map, and generating the panning image by applying a blur to a second area corresponding to the background portion of the received image frame based on the estimated speed parameter and the estimated direction parameter.