311 results about "Image contour" patented technology

The invention relates to the technical field of automatic defect detection, and discloses a mechanical watch shell defect automatic identification method and system based on deep learning. The method comprises the following steps: acquiring original image data and original point cloud data; performing feature extraction to obtain image contour features and point cloud feature values; inputting into a preset initial defect detection model to obtain an initial defect detection result; carrying out geometric feature extraction to obtain key geometric features and minimum enclosing rectangle features; correcting, standardizing and screening chord lengths in the key geometric features to obtain a screened feature matrix; and carrying out comparison, updating and feedback adjustment on elements in the screened feature matrix and a preset threshold value to obtain a final classification result of the defect area. And inputting the standardized feature matrix into a preset fusion geometric feature model to obtain a preliminary classification result of the corresponding defect region position. The method can improve the classification effect of automatic defect identification and detection and improve the detection accuracy.

The invention discloses a three-dimensional finite element model generation and numerical simulation method and system for a wading engineering composite material XCT digital image, and belongs to the technical field of finite element analysis. Comprising the steps of collecting a composite material XCT image in an initial state and performing preprocessing; performing statistical analysis on the two-dimensional microstructure parameters; a virtual XCT image is generated based on the Pix2PixHD style migration network; the method comprises the following steps of: accurately segmenting each component region of a real XCT (X-ray computed tomography) image based on a Swinin-Transformer network; completing contour feature extraction and smoothing of the segmented image based on a mapping relation between pixels and nodes; constructing a space coordinate system and carrying out discrete derivation on all yarn center points of each segmented image; grid division is carried out based on a multi-material moving cube algorithm in a CGAL library, and a global cohesion unit is inserted; completing material attribute and local material direction distribution; and customizing proper boundary conditions, and carrying out numerical simulation based on a finite element method.

The invention relates to a clinical wound healing monitoring method based on a three-dimensional wound evaluation tool. A 3D image of a wound is constructed through the three-dimensional wound evaluation tool based on a 3D image generation technology; using an image contour recognition algorithm to recognize and mark a wound healing contour in the 3D image, and using the three-dimensional wound evaluation tool to measure and mark contour parameters of the wound healing contour; and sending the marked wound healing contour to an HIS system, and storing the marked wound healing contour in an electronic medical record of the patient by the HIS system. Wound parameters are marked on the constructed wound three-dimensional image, so that medical staff can monitor the wound form of a patient and supervise wound healing, the wound parameters of the patient are recorded through an HIS system, the wound healing condition of the patient is determined, effective wound healing evaluation and monitoring are achieved, and practical measures which can be implemented on the ground are provided for clinic. And wound supervision and nursing management can be conveniently carried out on patients in general surgery department and the like.

The invention discloses a fluorescence in-situ hybridization image contour enhancement method based on illumination fusion and local binary, relates to the field of image processing, and solves the problems of different dye concentrations and the like due to image defects such as non-uniform local illumination and relatively large noise in an acquired image during existing tissue cell fluorescence image preprocessing. The method comprises the following steps: removing noise in an image by adopting a median filtering method, calculating a linear equation to obtain a texture weight of the image, performing analysis processing by utilizing the texture weight and a smoothing function to obtain a brightness image and a smooth image, and performing gray processing to obtain gray distribution information; adopting a local binary algorithm to select a coefficient, and performing enhancement processing on the image to obtain an illumination enhanced image; the contour feature image and the illumination enhancement image are subjected to weighted addition, eight neighborhoods around eight maximum pixel values are analyzed, an area with a high energy gradient is recognized through a piecewise linear function and a thermodynamic diagram, and a high-frequency signal area is the contour area of the fluorescence in-situ hybridization tissue cell image.

The invention relates to the technical field of water area public safety, in particular to a cross-modal intelligent perception fusion early warning method and system, and the method comprises the steps: respectively constructing a dam leakage recognition model, a dam deformation recognition model and a dam water pressure recognition model through collecting the sound wave data, image data and water pressure data of a dam in real time; and fast Fourier transform is carried out on the sound wave data through the dam leakage identification model, frequency domain features are extracted, and dam leakage abnormal coefficients are obtained. Recognizing the image data according to the dam deformation recognition model, obtaining dam image contour features and dam image region features, and obtaining a dam deformation abnormal coefficient. And recognizing the water pressure data according to the dam water pressure recognition model to obtain a dam water pressure abnormal coefficient. And obtaining a comprehensive dam abnormal coefficient according to the three abnormal coefficients, comparing the comprehensive dam abnormal coefficient with a preset dam abnormal threshold value, and performing real-time early warning. Through cross-modal data fusion analysis, the intelligent sensing and real-time management and control capabilities of dam public safety are improved.

The invention discloses an underwater image enhancement method and computer equipment, and the method mainly comprises the steps: correcting the color of an underwater original image through employing a self-adaptive color compensation algorithm, enhancing the details through employing a Multiply Contour Bougie (MCB) morphological detail enhancement algorithm, and achieving the global enhancement through employing a self-adaptive gamma correction method. And finally, fusing the detail and global enhancement result through a weighted stationary wavelet visual perception fusion strategy to obtain a final underwater image. The minimum color loss (MCL) algorithm designed by the invention can solve the problem of color channel information loss, so that the mean value of each color channel is similar to the histogram distribution; the image contour can be restored based on an MCB morphological detail enhancement algorithm, and details are highlighted; according to the weighted stationary wavelet visual perception fusion strategy, image information of different scales can be integrated, the definition and the layering sense are improved, the defects of a traditional method in detail and noise processing are overcome, and a high-quality image is obtained.

The invention discloses an image contour recognition method and device for a cell cluster, and relates to the technical field of biological image processing, and the method comprises the steps: converting an input cell cluster image into a binary image; carrying out expansion operation and distance transformation on the image after morphological operation, and carrying out subtraction operation on the two obtained images to obtain a third image containing an unknown region; detecting connected regions in the accurate foreground region, and marking each independent connected region, background region and unknown region to obtain a marked image; performing region segmentation on the original image to obtain segmented marked regions, generating a corresponding binary mask image, and extracting the contour of each marked region by using the binary mask image; identifying and filtering the contour of each marked area to obtain an effective contour; the size of each effective profile is converted into an equivalent sphere diameter. According to the invention, accurate segmentation of adjacent or adhesion areas in the image can be efficiently realized, and accurate counting and size measurement of cell clusters can be obtained.

The invention relates to the technical field of image processing and automatic cutting, and discloses an image contour extraction and cutting method, which comprises the following steps of: setting shooting equipment at a target position, placing an actual length reference mark, shooting a reference mark image through the shooting equipment, measuring the number of pixels and determining a pixel density reference parameter; collecting an image containing the target object and the background area through the shooting equipment; identifying a to-be-replaced pixel matched with the background color based on the background color, replacing the to-be-replaced pixel with a transparent pixel, and generating a processed image after the background is eliminated; performing edge detection on the processed image, extracting contour information of a target object, generating a vector contour file of an actual size in combination with a pixel density reference parameter, and generating a cutting path; and the cutting path is sent to the cutting device, and cutting operation of the target object is completed. According to the method, the image of the target object is collected in real time, the vector contour file of the actual size is generated in combination with the pixel density reference parameters, and dynamic and accurate target object cutting is achieved.

The invention discloses a joint dilator of an orthopedic knee arthroscope and a control method thereof, and relates to the technical field of medical instruments, the method comprises the following steps: acquiring a geometric center of a preset distraction range and a rotating fulcrum of the dilator, and constructing a preset Y axis according to the geometric center and the rotating fulcrum; when the spreader performs initial rotation by taking a preset Y axis as a rotation axis, the graph acquisition piece acquires an actual image contour in real time; comparing the actual image contour with the target image contour in real time; when the coincidence degree between the actual image contour and the target image contour is a second coincidence degree, the position of the preset Y axis is confirmed again according to the coincident contour between the actual image contour and the target image contour in the first coincidence degree, and an actual Y axis is set; the angle control device is combined with the actual Y axis to control the dilator, through the arrangement, it can be ensured that the dilated range of the dilator coincides with the range corresponding to the actual injury characteristic, and then the wound area of a patient is reduced.

The invention discloses an image alignment method and device, equipment and a storage medium, and relates to the field of image processing. According to the scheme, a to-be-detected circuit board scanning graph is obtained, and any unit area in the to-be-detected circuit board is recognized and extracted; based on a preset segmentation rule, segmenting the unit area into a plurality of subareas, and generating a window image set; and extracting window images in the window image set, mapping the window images to a mother layout corresponding to the Pcs particle image one by one according to image contours, and carrying out image alignment according to the image contours so as to carry out defect detection according to the mother layout. The method comprises the following steps: carrying out rasterization segmentation on a circuit board area image, and mapping the image to a target position in a mother layout in a manner that each window image is matched and aligned one by one; and each window image can be compared with the contour line of the mother layout as much as possible, the problems of element missing, dislocation, addition and damage are detected, and the subsequent defect detection precision is improved.

The invention discloses an intelligent cabin environment sensing and control method based on a vehicle sensor, and relates to the technical field of intelligent automobiles, and the method comprises the following steps: in the operation process of an intelligent cabin, an infrared thermal sensing system collects thermal imaging data in an automobile in real time, extracts multi-dimensional feature data based on the obtained thermal imaging data, and sends the multi-dimensional feature data to the intelligent cabin; and forming an infrared thermal image characteristic data parameter set. According to the method, the key features of the infrared thermogram are extracted, and the machine learning model is combined, so that accurate identification and dynamic suppression of the thermo-optical interference state are realized. According to the method, sampling and filtering strategies can be adaptively adjusted in complex environments such as strong light reflection, a real thermal image contour is reconstructed, a non-human body heat source is effectively eliminated, the recognition accuracy and the operation stability of an infrared thermal sensing system are remarkably improved, the misjudgment rate is reduced, and the safety and the reliability of an intelligent cabin are enhanced.

The invention provides a non-contact steel pipe end face size measurement method and system, and the method comprises the steps: traversing recognized contour points on the basis of obtaining an image and recognizing the contour of the image, calculating the distance between the contour points, judging the contour points which are too far away from each other as burrs, eliminating the influence of the points, carrying out the ellipse fitting of the contour, and obtaining the size of the end face of a steel pipe. And parameters such as the inner diameter, the outer diameter and the roundness can be calculated according to an ellipse fitting result. According to the invention, based on the original image of the pipe end, the contour of the pipe end is identified by combining image edge detection and a graphic method, the position relation between contour points and the contour is calculated through a self-research algorithm, pipe end burrs are accurately positioned, and the influence of the burrs on the measured pipe end size is eliminated. The problems that a manual measurement method is insufficient in precision and high in maintenance cost are solved, the burr sensitivity is improved, and the effective pipe end size is accurately obtained.

The invention discloses an accurate segmentation method and system for an abdominal organ image contour, and relates to the field of medical image processing, and the method comprises the steps: obtaining a to-be-segmented abdominal medical image sequence, CT image data at least containing one phase, and a corresponding Hensi unit matrix; based on an abdominal organ positioning network, obtaining a region of interest containing a target small organ and a preliminary space prior probability graph; a local HU goodness of fit diagram is generated through the dissection focusing path, a boundary response intensity diagram is generated through the boundary enhancement path, and a refined feature diagram is output; inputting the refined feature map into a main body segmentation network; fusing with corresponding hierarchical features of the decoder network, performing up-sampling and reconstruction, and generating an initial segmentation probability graph of the target small organ; and generating a final segmentation result graph and a comprehensive confidence graph. According to the method, the technical problems that in the prior art, due to the fact that inherent physical attributes are not fully mined and utilized, boundary positioning is fuzzy, the image segmentation precision is insufficient, and the confidence coefficient is low are solved.

The invention discloses an intelligent rice seed screening technical method based on deep learning and a control system, and the method comprises the steps: collecting continuous frame RGB images, rolling on an inclined collection plate, of each batch of rice seeds in a preset time period through a high-speed camera, and analyzing the continuous frame RGB image information of all batches of rice seeds; extracting image contour information and an image pixel ratio of each batch of rice seeds, judging the batch with the optimal representation of the rice seeds, and recording the batch as a target sample; the method comprises the following steps: acquiring spectral data of a target sample through a hyperspectral imaging system, preprocessing the spectral data and extracting a spectral marking signal; constructing a convolutional neural network model, inputting the spectral marking signal and the RGB image pixel information of the current frame number of the target sample into the model for training, and outputting an adjustment result; according to the method, the characterization information of the rice seeds is accurately obtained, and batch intelligent screening of the rice seeds is carried out, so that the screening efficiency is improved.

The invention discloses a defect detection method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a transmission image and a reflection image; respectively inputting the transmission image and the reflection image into an image contour extraction model to obtain contour images of the transmission image and the reflection image; wherein the image contour extraction model is trained by using a plurality of transmission images, reflection images and contour diagrams of the transmission images and the reflection images in advance; matching the generated contour image of the transmission image with the contour image of the reflection image to obtain contour matching coordinate points; matching the transmission image and the reflection image based on the contour matching coordinate points; pixel values of all pixel points on the transmission image and the reflection image are calculated and added to obtain a composite image; and comparing the pixel value of each pixel point on the synthesized image with a threshold value to determine a defect pixel point and a position coordinate thereof.

This technology provides systems and methods for tracking facial movements and reconstructing facial expressions by learning skin deformation patterns and facial features. Frontal view images of a user making a variety of facial expressions are acquired to create a data training set for use in a machine-learning process. Head-mounted or neck-mounted wearable devices are equipped with one or more camera(s) or acoustic device(s) in communication with a data processing system. The cameras capture images of contours of the users face from either the cheekbone or the chin profile of the user. The acoustic devices transmit and receive signals to calculate a representation of the skin deformation. A data processing system uses the images, the profile of the contours, or skin deformation to track facial movement or to reconstruct facial expressions of the user based on the data training set.

The invention relates to a robot drawing method, device and equipment and a storage medium, and is applied to the field of robot control. Contour extraction processing and skeleton extraction processing are carried out on the original image, a contour image is generated, the contour image comprises a plurality of contour segments, and each contour segment comprises a plurality of contour points; a Douglas-Peucker algorithm is adopted to simplify the number of contour points of each contour segment in the contour image, and a contour optimization image is generated; converting the image coordinate system of the contour optimization image into an actual coordinate system of preset paper according to the size of the preset paper; according to a contour screening threshold value, screening an optimized contour section in the contour optimization image, and according to the optimized contour section, generating a contour secondary optimization image; and controlling the robot to finish drawing in the actual coordinate system according to the outline secondary optimization image and preset drawing parameters. The method has the technical effects that the drawing precision is guaranteed while the drawing efficiency of the robot is improved.

The invention relates to the technical field of image contour extraction, and provides an image edge contour extraction method and device, electronic equipment and a storage medium. Edge detection and region division are carried out on a target image to obtain a region matrix, scanning and contour point screening are carried out on each sub-region to obtain a contour point coordinate chain table of each sub-region, and single-region adjacent point screening is carried out on contour points of each sub-region in combination with a first distance threshold to obtain an initial contour track chain table set; performing cross-region contour fitting point extraction in combination with the boundary of each sub-region to obtain a to-be-verified point data set, and performing cross-region adjacent point screening on the to-be-verified point data set according to a second distance threshold and the region arrangement data to obtain a fitting point combination set, and carrying out cross-region merging on the initial contour track chain table set to obtain an edge contour track chain table set. Through multi-core parallel scanning, multi-distance screening and cross-block fitting technologies, the precision and efficiency of image edge contour extraction are improved.

The proposed technique introduces embodiments of a computer-implemented method for interpreting image delineations as vector objects and topological extraction from segmentation by visual computational methods applied to sections (slices) of seismic volumes, in order to aid geological interpretation and sampling of parameters originating from the fracture network and its topology. Embodiments of a developed method integrates a software / application that allows the loading and generation of statistical data related to the fracture network while maintaining georeferencing and scale of the two-dimensional input data. In addition, a fracture segmentation method is shown that uses pyramid image smoothing (decomposition into hierarchical levels of resolution) in order to reduce the amount of details and aid the identification of main faults or fractures. The segmentation after this smoothing is based on adaptive thresholding segmentation.

The invention discloses a cable detection method based on picture processing, which is used for detecting dust and dirt on the surface of a cable, and comprises the following steps of: acquiring a sequence image of a fixed cable in a box body without external light interference for each section of cable by an area-array camera which rotates by 180 degrees by taking the cable as a central axis, converting a gray image, intercepting a rectangular image, and determining the dust and dirt on the surface of the cable. Gaussian fuzzy filtering, contrast enhancement, threshold segmentation, edge detection, contour detection, calculation of the contour surrounding area of the cable image and judgment of whether dust and dirt exist in the section of the cable are performed. According to the method, whether dust and dirt exist or not is judged through the cable image contour surrounding area, the radian size of the cable arc-shaped surface and details of the arc-shaped surface do not affect judgment, and meanwhile the dust and dirt details of the cable edge can be better highlighted through backlight shooting. The influence of the external environment and the arc-shaped surface of the cable on dust and dirt detection can be greatly reduced, and the detection precision, accuracy and working efficiency are improved.

The invention discloses a contour extraction method based on edge detection, and belongs to the technical field of image processing, and the method comprises the following steps: obtaining a plurality of images to be subjected to contour extraction, and real labeled contours corresponding to the images to be subjected to contour extraction; performing noise level quantization and edge ambiguity quantization analysis on each image to be subjected to contour extraction, and calculating to obtain an image complexity level of each image to be subjected to contour extraction; performing complexity level classification on each image to be subjected to contour extraction, and constructing a complexity classification image data set; constructing a contour extraction model based on edge detection; a trained contour extraction model is obtained; and obtaining a new image to be subjected to contour extraction and an image complexity level, inputting the image to be subjected to contour extraction into the trained contour extraction model for contour prediction, and obtaining a predicted image contour of the image as a contour extraction result. The problem that it is difficult to accurately extract the object contour in a complex image scene is solved.

The invention provides a monocular vision pose tracking method, which comprises the following steps: according to a camera imaging principle, establishing a constraint relation between three-dimensional coordinates of the center of a ball target and coordinates of pixel points on an imaging contour, optimally solving the three-dimensional coordinates of the center of the ball target, realizing spatial positioning of a single ball target, and forming a tracking target by a plurality of ball targets. After the center position of each ball target is captured in real time, the six-degree-of-freedom space pose of the tracking ball target is obtained according to singular value decomposition, and for positioning errors caused by ball target contour extraction, ball target diameter deviation and the like, a calibration method based on feature points is provided to calibrate internal parameters of a system, and compared with a traditional PnP method, the method has the advantage that the positioning accuracy is improved. The positioning process does not need to pre-specify the corresponding relation between the feature points and the image points, does not depend on the pose relation of other feature points, does not relate to other auxiliary constraints, and is easy to realize automatically.

Using a computer-implemented intermediary by which contouring performed by two participants, such as two physicians, can be compared. First, contouring performed by each participant can be compared to contouring performed by the intermediary. Then, by way of the common intermediary and a transitive analysis, contouring performed by each participant can be compared.

The invention belongs to the technical field of image processing, and particularly discloses a polygon contour optimization method and system based on annular point density. The method comprises the following steps: calculating polar coordinates of each contour point in a polygonal contour; setting a detection interval, calculating the density of contour points in the detection interval according to the polar coordinates, and deleting all contour points of which the density is greater than or equal to a preset threshold value; setting a circulation step length delta, calculating the density of the contour points in the next detection interval in the circulation step length delta, deleting all the contour points with the density larger than or equal to a preset threshold value, and traversing the polygon contour for one circle to complete polygon contour optimization. According to the method, the problems that redundant information can be extracted in the traditional image contour extraction process and the outer contour of the polygon cannot be accurately extracted are solved, and the geometric contour conforming to the polygon features can be effectively obtained.

The invention relates to the technical field of image processing, in particular to a method and a system for extracting an algae microscopic image contour. The method comprises the following steps: acquiring a multi-view image flow of algae cells under a microscope, and carrying out frame-level scheduling and semantic splicing to obtain a multi-angle fusion image unit; constructing a topological contour tensor graph for the multi-angle fusion image unit; performing multi-scale positive and negative contour comparison training on the topological contour tensor diagram to obtain a contour embedded spectrum; holographic algae sample matching back-pushing is carried out on the contour embedded spectrum, and a candidate matching spectrum set is obtained; analyzing an edge feature contribution degree according to the candidate matching spectrum set, and carrying out dynamic fusion on the edge feature contribution degree and the multi-angle fusion image unit to generate an edge saliency heat map; screening independent algae samples by utilizing an edge significance heat map; and carrying out data modeling on the independent algae samples, and synchronizing a data modeling result to an algae sample library. According to the invention, the image contour extraction precision and efficiency can be improved.

The invention relates to the technical field of image recognition, in particular to an automatic detection grabbing system of an earphone cover assembly line, which comprises an earphone cover image acquisition module for capturing an earphone cover image; according to the invention, by calculating the edge sharpness integral of the earphone cover image and extracting the contour coordinate sequence, the image contour positioning precision is improved, and the ambiguity problem of earphone cover boundary identification is solved; main harmonic morphological characteristics are obtained based on low-order descriptor coefficient energy distribution, the expression mode of the morphological characteristics of the earphone cover is optimized, the accuracy degree of matching of the form of the earphone cover and a standard template is further improved, and the appearance misjudgment probability is reduced; in addition, morphological opening operation is carried out by applying multi-scale structure elements, the structure scale entropy is calculated to reflect change characteristics of fine textures on the surface of the earphone cover, finer texture particle size distribution is established, the texture consistency deviation degree is measured and calculated, and the error of surface texture detection is reduced.

The invention relates to the technical field of rice phenotype extraction, and discloses a rice fine phenotype extraction method based on three-dimensional vision, a terminal and a storage medium. The method comprises the following steps: acquiring a multi-view image of a rice plant by using a depth camera, and performing camera attitude estimation on the multi-view enhanced image to obtain camera internal and external parameters corresponding to each view; constructing a neural radiation field network, training the neural radiation field network by taking the enhanced image, the contour mask and internal and external parameters of the camera as input, and learning color and density distribution of a scene by utilizing a volume rendering technology; extracting surface points based on a mixed threshold strategy by using the trained neural radiation field network, and generating a rice three-dimensional point cloud; inputting the rice three-dimensional point cloud into a point cloud instance segmentation network, and outputting an instance mask containing each organ of the rice and a segmentation result of a three-dimensional bounding box; and counting phenotypic parameters of the rice according to a segmentation result. According to the method, the fine instance segmentation of the high-density point cloud is realized while the reconstruction precision is ensured, and the phenotypic analysis of the spike fraction is supported.

The invention provides a curvature-enhanced large-displacement image variational optical flow method for motion estimation, relates to the field of image processing, and aims to describe the local structure complexity of an image by introducing an image contour curvature. On the basis of the curvature, limited self-adaptive weighted adjustment is carried out on the brightness invariant constraint and the gradient invariant constraint on the data item level of the opto-rheological model, so that the interference of unreliable matching in a complex structure region on optical flow estimation is inhibited; the robustness of optical flow estimation in illumination variation, complex texture and large displacement scenes is improved; and the numerical stability and convergence of the model in the multi-scale calculation process are ensured.

The invention discloses a data enhancement method and device based on a multi-scale feature adaptive focusing network, and relates to the technical field of image generation. The method comprises the following steps: carrying out adversarial training on a multi-scale feature adaptive focusing generative adversarial network; the generative adversarial network comprises a generator network and a discriminator network; the discriminator network comprises a multi-scale feature extraction module, a dynamic channel attention sub-module and a space attention sub-module, and the feature extraction capability of image contours, textures and details is enhanced; in the training process, multi-scale feature confrontation optimization is achieved through multi-scale Wasserstein distance and dynamic weight updating, the capturing capacity of a discriminator and a generator for details of different scales is enhanced, and the convergence of the model is ensured; according to the method, the required weight parameter can be obtained after the model is trained to be converged, and the high-quality image generation task is completed by using the weight parameter, so that the problems of low quality, lack of diversity and the like of the generated image in a small sample scene can be solved.

The invention relates to the technical field of printing detection, solves the technical problems of low recognition accuracy and low recognition efficiency caused by the influence of external environment in the prior art, and particularly relates to a printing quality detection method based on machine vision, which comprises the following steps: S1, acquiring an original image of a printing image, preprocessing the original image to obtain a preprocessed image; and S2, calculating a first feature value and a second feature value according to the preprocessed image, and obtaining a preposition vector based on the first feature value and the second feature value. Defect identification is performed by combining the morphological characteristics of the image contour, so that the robustness of the algorithm can be improved, the training of a large amount of data is avoided, and the method has high robustness and high accuracy. Compared with a traditional area threshold value, the method is more accurate and efficient, and the calculation efficiency is greatly improved.