48results about How to "Achieve segmentation" patented technology

The invention discloses a thrombolysis prediction method based on deep learning and multi-modal fusion, and belongs to the technical field of thrombolysis prediction, and the method comprises the steps: S1, collecting an MRI image of a patient with ischemic cerebral apoplexy, and constructing a cerebral apoplexy automatic segmentation model; s2, carrying out image registration; s3, lesion segmentation; and S4, thrombolysis prediction. According to the cerebral apoplexy automatic segmentation model based on the gating parallel Mama network, the long-distance dependency relationship in the image can be effectively captured, ischemic cerebral apoplexy lesions can be rapidly and accurately segmented, and the lesion identification efficiency and accuracy are greatly improved; according to the algorithm provided by the invention, the linear time complexity is realized, the computing resource consumption is greatly reduced, and the bottleneck of the prior art is broken through; according to the cerebral apoplexy thrombolysis prediction model based on the multi-modal data, scientific automatic decision support is provided for judging whether the patient is suitable for thrombolysis treatment or not, and the quality and safety of medical decision are improved.

The invention relates to the technical field of burr detection, in particular to a high-speed rail wheel set oil filler hole burr detection and segmentation method which comprises the following steps: acquiring burr point cloud data around an oil filler hole in the surface of the inner wall of a high-speed rail wheel, and preprocessing the burr point cloud data to obtain a data set; point net + + is used as a basic network architecture, an L2 regularization item is introduced to construct a loss function, and an AdamW optimizer and a cosine annealing function are matched to construct a burr detection model; training a burr detection model based on the data set; and performing burr detection and segmentation on the point cloud data around the oil injection hole based on the trained burr detection model. According to the invention, the detection precision of the burrs around the high-speed rail wheel oil filling hole can be improved.

The invention provides a signal point identification method and a related product. According to a specific embodiment of the method, instance segmentation is performed on a cell nucleus image based on a cell instance segmentation model, and a cell segmentation result of each cell in the cell nucleus image is obtained; based on a fusion feature extraction model, carrying out image feature extraction on the fluorescence signal point image of each channel to obtain a fusion feature map; then, based on the cell segmentation result of each cell, extracting a cell region fusion feature map of each cell from the fusion feature map; then, based on a cell region mask feature extraction model, obtaining a masked cell region fusion feature map of each cell; and finally, based on a signal point identification model, performing signal point identification on the masked cell region fusion feature map of each cell to obtain a signal point identification result of the corresponding cell in each channel. Therefore, end-to-end cell segmentation and signal point identification can be realized, cell segmentation and signal point identification can be realized based on the cell nucleus image and the multi-channel fluorescence signal point image, cell segmentation and signal point identification can be completed at the same time with fewer model parameters, the calculation amount is saved, and the algorithm processing time is accelerated.

The present invention discloses a 2D / 3D detection and segmentation method, device and terminal equipment for orthopedic surgery registration, the method steps include: step S01. obtaining preoperative CT data and intraoperative perspective X-ray images; step S02. inputting the preoperative CT data into a 3D detection network for three-dimensional landmark detection and bone tissue segmentation, wherein the 3D image coding features of the input CT data are first detected, edge information is extracted, and decoding is performed to obtain preliminary detection and segmentation results, and the confidence map is calculated to obtain the intermediate segmentation results, which are input into a 3D prompt encoder to generate prompt information; step S03. inputting the intraoperative perspective X-ray image into a TransUNet network for two-dimensional landmark detection and bone tissue segmentation; step S04. outputting the landmark detection and segmentation results for registration. The present invention has the advantages of simple implementation method, strong real-time and interactivity, high detection and segmentation accuracy and efficiency.

The invention relates to the technical field of image segmentation, in particular to a pancreas CT image segmentation method and system and a storage medium, and the method comprises the steps: removing an edge interference region of a pancreas through analyzing the gray value features of the pancreas in a pancreas threshold segmentation image, and obtaining a pancreas interference-removed image; analyzing fluctuation characteristics in the direction of edge pixel points of the pancreas interference-removed image to obtain inflection points; performing linear cutting on the pancreas interference-removed image based on the inflection points to obtain paired cutting areas, analyzing gray difference and shape feature difference of the two cutting areas, and determining inflection points of an adhesion area and the adhesion area; based on the inflection points of the adhesion area, analyzing the distribution characteristics of pixel points in the adhesion area to obtain the possibility that the pixel points are pancreas edge pixel points; and classifying the pixel points according to the possibility to obtain a pancreas edge pixel point sequence. According to the invention, the accuracy of pancreas CT segmentation is improved, and the segmentation result is more authentic.

The invention discloses a method and system for detecting the state of an intersection hole boring cutter, and belongs to the technical field of industrial big data. Detection is carried out based on multi-modal data, and the abrasion bandwidth of the rear tool face of the tool and the abrasion ratio of the front tool face image of the tool are correspondingly obtained through collected machining time sequence data in the one-tool machining process and the front tool face image of the tool after one-tool machining are completed and serve as detection indexes of the intersection point hole boring tool state. The machining precision condition is represented through the abrasion bandwidth, and when the abrasion bandwidth is large, it is indicated that the rear tool face of the tool is abraded greatly, and the machining precision is affected; the surface roughness condition is represented through the abrasion ratio, and when the abrasion ratio is large, it is indicated that the front tool face of the tool is abraded greatly, and the surface roughness of a workpiece is affected. According to the method, the requirements for machining precision and surface roughness are comprehensively considered, and accurate detection of the state of the intersection hole boring cutter is achieved.

The embodiment of the invention discloses an MRI (Magnetic Resonance Imaging) image segmentation method, device and equipment based on a Mask RCNN (Regional Convolutional Neural Network) algorithm and a storage medium. The method comprises the following steps: extracting features of a preprocessed MRI image to obtain a feature map; the RPN is used for generating a final Proposa box, and the RPN is used for generating a final Proposa box; the ROI is selected to be subjected to pooling operation, feature fusion and feature coding in sequence, and deep target features and a feature coding graph are obtained; performing classification and frame regression on the targets in the ROI, and determining a target frame and a classification label of each target; determining a prediction segmentation mask of each target through the segmentation network; carrying out post-treatment; non-maximum suppression processing is carried out in a classification mode, and a final target frame, a classification label and a post-processed mask image are determined. According to the scheme, focuses with different scales and different shapes, sizes and positions can be processed through the Mask R-CNN algorithm, compared with a traditional MRI image segmentation technology, higher accuracy and generalization ability are achieved, the two tasks of target detection and segmentation are achieved at the same time, and doctors are helped to provide more accurate diagnosis and treatment.

The invention relates to a computer vision technology, in particular to a medical image processing technology based on deep learning and a medical diagnosis auxiliary method. The invention provides an airway stenosis automatic segmentation method, system and device based on a CT image, and the method specifically comprises the following technical means: 1) building an airway stenosis automatic segmentation model based on a marked training set and a verification set, so as to achieve the automatic recognition, positioning and segmentation of a narrow airway; (2) reasoning by using the established airway stenosis automatic segmentation model, and automatically identifying, positioning and segmenting a stenosis airway from an input CT (Computed Tomography) image; and 3) performing quantitative parameter measurement based on the narrow airway obtained by automatic segmentation. According to the automatic segmentation system and device for the airway stenosis, whether the airway stenosis exists in the CT image or not can be rapidly and accurately recognized, if yes, automatic positioning and segmentation of the stenosis can be achieved, and the system and device assist imaging departments or clinicians in rapidly and accurately diagnosing the narrow airway; meanwhile, based on the airway stenosis obtained through automatic segmentation, the stenosis range and the maximum, minimum and average lumen diameters of the stenosis part can be rapidly and accurately calculated, on one hand, the method can be used for evaluating the severity of the stenosis, and on the other hand, the method can be used for assisting clinical treatment.

The invention discloses a laser radar-based efficient obstacle recognition method and terminal, and the method comprises the steps: carrying out the scanning through a laser radar with a preset number of wire harnesses, and obtaining the laser point cloud data in a preset range in front of a vehicle; filtering ground points in the laser point cloud data, and only retaining the ground points; with the laser radar as the center and the preset angle as the unit, a preset range in front of the vehicle is divided into fan-shaped areas; for each fan-shaped region, calculating a spatial distance between a point cloud point in the fan-shaped region and the laser radar, classifying the point cloud point according to the spatial distance, and identifying an obstacle according to a classification result; according to the method, operation is simplified, the obstacle recognition efficiency is improved, the hardware requirement is reduced, and the refresh rate stability of obstacle recognition is improved, so that the requirement for rapidness of obstacle emergency processing is met.

The utility model relates to an anti-ice-falling device for the surface of a cross tie beam of a bridge tower, which comprises a plurality of anti-ice-falling devices arranged on the surface of the cross tie beam, each anti-ice-falling device comprises an anti-ice-falling clamp, a plurality of raised fixing pieces and a plurality of extended connecting pieces, and the anti-ice-falling clamps comprise a first anti-ice-falling clamp and a second anti-ice-falling clamp. The first anti-ice-falling clamp and the second anti-ice-falling clamp are fixedly arranged on the edges of the left side and the right side of the cross tie beam respectively, and the first anti-ice-falling clamp and the second anti-ice-falling clamp are connected into a whole through the extending connecting pieces and the protruding fixing pieces and fixed to the upper surface of the cross tie beam. The limiting baffles arranged at certain intervals prevent solid accumulated ice from sliding on the upper surface of the bridge tower cross tie beam, meanwhile, lateral drainage of the upper surface of the cross tie beam cannot be hindered, the phenomenon that a traditional drainage channel becomes a'water blocking channel 'due to icing is avoided, and the size of a single block of ice accumulated on the upper surface of the cross tie beam is limited within a safety threshold value as much as possible.

The present invention belongs to the technical field of LED display chips, and discloses a Micro LED display chip and a preparation method thereof. The present invention processes an MQW layer of an epitaxial wafer by ultrafast pulse laser, so that an EBL layer and a P-GaN layer that can be free of ICP etching are prepared by subsequent deposition processing, so that the preparation method of constructing a vertically distributed LED display chip is greatly simplified. At the same time, a larger current density can be achieved through a cathode shallow groove network and a deep through-hole design, and optical crosstalk can be effectively prevented.

The utility model provides an air conditioner indoor unit. The air conditioner indoor unit comprises a machine shell and a flow guide block. An air channel is formed in the machine shell and provided with an air outlet so that heat exchange airflow in the machine shell can be blown out of the machine shell through the air outlet. The flow guide block is arranged at the air outlet and comprises an air baffle and a partition plate, the flow guide face of the air baffle faces the interior of the air duct, and the partition plate extends out of the interior of the air duct from the flow guide face; the partition plate is used for dividing the heat exchange air flow into two strands, and the two strands of heat exchange air flow are guided towards the two sides of the air outlet through the parts, located on the two sides of the partition plate, of the flow guide face. The air conditioner indoor unit can prevent heat exchange airflow from directly blowing forwards.

The application discloses a kind of complex lighting conditions underwater artificial crack dataset construction method, it is related to image dataset production technical field, including, collection crack picture and extract crack curve generation coordinate information, based on this, construct crack 3D model and prepare concrete standard test block;The test block is placed in water, and underwater crack image under conventional, dim and uneven illumination is collected, and is classified after pre-processing according to illumination condition;Stable Diffusion model based on LoRA fine-tuning is used, and different illumination condition images are input respectively to train and optimize, and style expansion image is generated, and pre-processing image constitutes unlabeled data together;Through image enhancement algorithm processing the data, after the enhanced image is labeled, the label is corresponded with the original image, and the underwater artificial crack dataset under complex lighting conditions is constructed, the application solves the problem that actual underwater crack disease is difficult to collect, meets the core demand of engineering practice, and provides richer data support for underwater structure crack detection and segmentation model training.

In order to solve the problems of limited labeled data, differences between blood vessels and interference of lesion area, the present application discloses a retinal blood vessel image segmentation method based on a multi-scale dilated convolution residual network, which is still a challenge problem for accurately segmenting retinal blood vessels, especially fine blood vessels, on a retinal fundus image. The present application designs a multi-scale residual input and output module to make up for the loss of part of the blood vessel structure information due to down-sampling, combines the advantages of dilated convolution and DropBlock to relieve network overfitting and reduce the influence of the lesion area on blood vessel feature extraction, and further introduces a multi-scale mean pooling module to obtain high-level features and retain context information. Finally, by improving the way of jump connection, the dilated convolution is effectively used to improve the information transmission capacity of the jump connection. Compared with other algorithms, the present application can more accurately segment the fine blood vessels in the retinal image under complex conditions and has better robustness.

The invention relates to the field of environmental protection engineering, in particular to a device used in the water pumping and air pumping process in an underground water well, and the device can divide the internal space of a well casing into two independent parts. The device comprises an extraction pipe 1, an air bag 2, a one-way valve 3, a movable rubber plug 4, a one-way valve 5, a movable rubber sheet 6 and the like. According to the device, the air bag 2 is inflated by injecting air, the outer wall of the air bag is completely attached to the extraction pipe 1, the inner space of the well casing is divided, and therefore the function of taking water / air in a layered mode is achieved. During air exhaust, the one-way valve 5 and the movable rubber sheet 6 work, the one-way valve 5 prevents air in the air bag from being exhausted out of the air bag, the movable rubber sheet 6 is opened, and water / air below the air bag in the well casing is exhausted. The existing solution is complex and high in cost. According to the device provided by the invention, time and cost can be saved, and an expected effect can be achieved.

The present invention discloses a method for extracting roads from remote sensing images based on graph convolution. Graph convolution can aggregate feature information between adjacent nodes, extract features within a larger neighborhood of nodes, and effectively solve the problem of local position information loss. The designed method can be regarded as multi-task learning. First, CNN is used to realize feature extraction of remote sensing images. Secondly, a graph structure model is constructed based on the road features extracted by CNN, which is mainly composed of nodes and corresponding edge relationships. The road feature information extracted by the CNN branch is regarded as a node, and the difference between nodes is regarded as an edge. The local position information is obtained by obtaining the relationship between nodes. The present invention solves the problem of local road position information loss caused by the generalization effect of convolutional neural networks by using graph convolution, thereby effectively improving the accuracy of road segmentation.

The application discloses an AI identification method, system and model training method for millimeter wave image overlapping targets based on a three-branch network, and solves the security target detection problem in millimeter wave images with low resolution, mutual interference or mutual overlap through collaborative work of three independent branch networks, namely, an analysis branch network, an aggregation branch network and a prediction boundary branch network. The analysis branch network saves the detailed information in a high-resolution feature map; the aggregation branch network realizes the feature position offset alignment of high and low resolution feature maps through a recursive fusion module; and the prediction boundary branch network extracts high-frequency features and enhances the perception and prediction of target edges. The multi-branch network structure effectively realizes the segmentation and identification of images through the fusion between branches. While improving the segmentation performance of the model, the model significantly reduces the calculation complexity and the detection error and missed detection probability in millimeter wave images, and finally realizes the accurate detection of overlapping objects.

The present invention relates to an image segmentation method based on lightweight recursive non-local self-attention, comprising the following steps: step S1: collecting portrait segmentation data, and pre-processing to obtain a portrait segmentation data set; step S2: constructing an image segmentation model based on lightweight recursive non-local self-attention; step S3: based on the portrait segmentation data set, using a binary cross entropy loss function and back-propagating the loss gradient, and training until the model converges according to a preset training learning rate, optimizer and learning rate strategy; step S4: normalizing the image to be tested and inputting it into the trained image segmentation model to obtain a portrait segmentation map. The present invention can effectively realize portrait segmentation, and the imaging effect is significantly improved.

The present invention provides a speaker segmentation method based on an improved variational Bayesian algorithm, comprising: segmenting speech and extracting features of each segment; clustering using the improved variational Bayesian algorithm, first cross-calculating the feature similarity of speech segments, merging the two speech segments with the highest similarity until the maximum feature similarity is less than a threshold, and obtaining an initial clustering result; calculating the class vector of each class, and obtaining the probability of the voiceprint of each speech segment belonging to each class; calculating the voiceprint of each speech segment and the new weight of each class, and updating the initial state probability in a hidden Markov model at the same time, calculating the lower limit of evidence until a termination condition is met, and completing the classification of speech segments; merging speech segments of the same class, obtaining the start and end points of speech of different people, and realizing speaker segmentation; the method provided by the present invention combines the cosine algorithm with the improved variational Bayesian algorithm for clustering and speaker segmentation, and the segmentation is more accurate, the effect of speaker separation is improved, and it is more suitable for long dialogue scenarios.

The invention relates to a texture processing area segmentation method and device, computer equipment and a storage medium. The method comprises the following steps: binarizing a texture grey-scale image based on a multi-level grey-scale range to obtain a multi-level binary image; randomly selecting seed points from a connected domain in the binary image; segmenting the connected domain based on the seed points to obtain segmented regions; determining a first mapping region where the segmentation region is mapped to the surface of the three-dimensional object; determining a first pose of the three-dimensional object when the center of a seed triangular patch corresponding to the seed point in the first mapping area is vertically penetrated by a central ray in the laser processing range; if the first mapping area is matched with the laser processing range when the three-dimensional object is in the first pose, determining that the first mapping area is a texture processing area, and determining that the first pose is a corresponding processing pose; otherwise, the first mapping area is segmented into a texture machining area matched with the laser machining range, and the corresponding machining pose is determined. By adopting the method, the texture processing quality can be improved.

The embodiment of the present invention provides a multi-sensor fusion human intention recognition method, system, electronic device and medium, which belongs to the field of human-machine collaboration and intelligent manufacturing. The method includes: using inertial sensors worn on the upper limbs of personnel to collect continuous action signals and perform online segmentation to obtain a set of candidate segmentation points; based on information gain, the continuous action signals are divided into segments with minimum entropy to determine the optimal segmentation point in the set of candidate segmentation points, and the legitimacy of the segmented action segments is checked based on meta-actions; based on a fully convolutional bidirectional long short-term memory network, the dependency relationship between the action segments is determined, and the behavioral intention of the personnel is identified by matching according to the action library. By combining the action library and the fully convolutional bidirectional long short-term memory network, the segmentation of complex continuous human actions in human-machine collaboration scenarios is realized, the accuracy and versatility of personnel intention recognition in the process of human-machine collaboration are improved, and the basis for the real-time implementation and later optimization of human-machine collaboration tasks is provided.

The invention relates to an etching method and a semiconductor manufacturing method, and the etching method comprises the steps: depositing a first oxidation layer and a first mask layer on a to-be-etched layer, forming a plurality of metal channels in the first mask layer through patterning, depositing a mask layer group on the first mask layer, and carrying out the etching of the mask layer group through patterning, so as to form a mask block, and etching the first oxide layer and the to-be-etched layer by taking the first mask layer and the mask block as masks to form a target pattern. And etching the first oxide layer and the to-be-etched layer by taking the first mask layer and the mask block as masks, thereby obtaining the target pattern on the to-be-etched layer, realizing segmentation of the metal layer in the target pattern, realizing effective insulation between the metal layers, reducing function failure caused by crosstalk or short circuit between the metal layers, and improving the performance of the device. Therefore, the overall performance and reliability of the chip are improved, and the production rate of defective products is reduced.

The embodiment of the present invention relates to the field of computer vision and information technology, and discloses a method, device and medium for segmenting and optimizing jewelry pictures in a detection certificate. The method comprises: collecting jewelry pictures, framing part of the jewelry pictures and attaching labels of corresponding types to obtain pictures to be trained; training a target detection model based on the pictures to be trained to obtain a trained target detection model; detecting the jewelry pictures using the trained target detection model to obtain pictures to be segmented; inputting the pictures to be segmented into an image segmentation model, identifying the picture type of the pictures to be segmented using the image segmentation model, and marking the pictures according to the picture types; performing image segmentation based on the marked pictures to obtain segmented pictures; optimizing the segmented pictures using a predetermined optimization algorithm, and uploading the optimized pictures to the detection certificate. Through the above-mentioned manner, the embodiment of the present invention can improve the accuracy and efficiency of jewelry picture segmentation.

The invention relates to an image processing method and device, computer equipment, a storage medium and a program product. The method comprises the following steps: acquiring an image acquired by a robot; performing semantic segmentation and mask generation on the image to obtain a task area, a non-task area and mask image data of the non-task area; performing disturbance processing on the non-task area to obtain a disturbed non-task area; and based on the task area, the post-disturbance non-task area and the mask image data, generating a target image for robot imitation learning. By adopting the method, the problem of distraction in the imitation learning process can be effectively avoided.

The present invention belongs to the technical field of photovoltaic modules, and specifically relates to an integrated thin-film photovoltaic battery group, a double-sided photovoltaic module and a preparation method. The battery group includes a substrate, a first electrode layer, a photovoltaic functional layer including a first charge transport layer and a second charge transport layer, and a second electrode layer. The first surface of the substrate has a series groove, and the series groove has a mask wall B with an angle of 45° to 90° with the substrate reference plane A. In the z direction of the substrate, the depth of the mask wall B is greater than the sum of the thickness of the first electrode layer and the photovoltaic functional layer not including the second charge transport layer and is less than the total thickness of the first electrode layer, the photovoltaic functional layer and the second electrode layer. The first electrode layer, the photovoltaic functional layer and the second electrode layer are sequentially prepared on the projection surface of the x-y plane of the first surface of the substrate, and each monomer thin-film photovoltaic cell is formed by means of the series groove. The first electrode layer of the monomer thin-film photovoltaic cell and the second electrode layer of the adjacent monomer thin-film photovoltaic cell are connected in series to form a battery group, which significantly improves the photoelectric conversion efficiency.

The invention discloses a spatial frequency domain imaging method, which comprises the following steps: acquiring first optical parameter diagrams of a to-be-detected tissue under different wavelengths, and demodulating a diffuse reflection image of the to-be-detected tissue to obtain the first optical parameter diagrams under the different wavelengths; inputting the first optical parameter diagrams under different wavelengths into a semantic segmentation network to obtain second optical parameter diagrams only containing the to-be-detected tissue under different wavelengths; taking the second optical parameter diagram under one wavelength as a reference image, and registering the second optical parameter diagrams under other wavelengths to obtain registered optical parameter diagrams with correct relative positions under different wavelengths; combining the registration optical parameter images under different wavelengths into channels to form an RGB image, and inputting the RGB image into a spectrum reconstruction network to obtain a hyperspectral image of the tissue to be detected; according to the invention, a high-quality spatial frequency domain imaging result can be quickly obtained.

This invention relates to the field of computer technology and discloses a method, apparatus, device, and storage medium for developing web pages. The method includes: receiving a web page development request; responding to the request by invoking a pre-built business code library and a pre-built component code library; generating a web page framework based on the business code library; performing conversion processing on the component code library to generate component links corresponding to the component code library; introducing the component links into the web page framework; and generating the web page using the Render function of the Vue framework library. Through this method, the invention can achieve the separation of component code and business code, eliminating the need for further separation of component code, reducing code coupling among multiple developers and departments, lowering communication costs and maintenance management risks, improving page development efficiency, and solving the problem of low web page development efficiency caused by developers repeatedly reviewing code when collaborating with multiple people and departments.

The present application relates to the field of image recognition technology. The present application discloses a method, device, equipment and medium for fine-grained semantic segmentation of window images, which can realize accurate segmentation of building facades at a fine-grained level. The method for fine-grained semantic segmentation of window images includes acquiring window images; preprocessing the window images to obtain binary images; generating a unified expression paradigm based on the binary images; denoising the binary images using morphological opening operations combined with the unified expression paradigm to obtain denoised images; extracting the denoised images using connected component analysis to obtain fine-grained segmentation results of window glass.

The present invention discloses an image processing method and device, wherein the method comprises: acquiring image information, and converting the image information into a transmittable image signal; converting the image signal into a spike signal through an oscillating artificial neuron device; encoding the spike signal to obtain a counting pulse; wherein the number of the counting pulses is used to characterize the size of the spike; inputting the counting pulse into a pulse coupled neural network for segmentation processing to obtain a segmented target image. The present invention can realize the spike coding function and visual perception function in the human brain visual system.

The application discloses a small sample semantic segmentation method, comprising the following steps: extracting features of support set pictures through a basic backbone feature extraction network to obtain feature layers at various levels; multiplying the feature layers of the support set pictures with support set real masks to segment objects in the support set pictures; generating basic discrete cosine transform (DCT) vectors, and expanding the generated vectors to the same size as the feature layers to obtain total frequency domain vectors; multiplying corresponding feature vectors and frequency domain vectors, and then dividing by object pixel sums in the pictures to finally obtain multiple prototypes; wherein the lowest frequency domain prototype can represent global understanding of the object, and other frequency domain prototypes can represent local cognition of the object. The small sample semantic segmentation method of the application extracts object information in pictures in the frequency domain, and describes the objects in the pictures in detail from two aspects of global and local.