17results about How to "Effective segmentation" patented technology

The application discloses a kind of for the method, device, equipment and medium of tab abnormality detection in battery cell production process, comprising: the tab area image collected in battery cell production process is segmented to generate tab mask;Based on the tab mask extraction multi-scale pixel-level feature vector, and the multi-scale pixel-level feature vector is input into unsupervised probability GMM model to identify pixel-level abnormal point;The pixel-level abnormal point is analyzed to obtain candidate abnormal area, and the parameter index of the candidate abnormal area is obtained;Determine whether the parameter index meets the set tab abnormality judgment rule, if meet, it is judged as tab abnormality, and tab abnormality result is output.Therefore, the application can be widely applied to the detection of tab folding, crease, fracture, misplacement and other abnormalities on battery cell production line.

The invention discloses a convex threaded hole measurement method based on a point cloud, and the method comprises the steps: collecting a point cloud A of the surface of a workpiece, and segmenting a point cloud B of a thread from the point cloud A; obtaining a normal vector of each point in the point cloud A; reversely projecting points in the point cloud A to a two-dimensional image plane, acquiring a normal change rate of each pixel point P, and judging whether the point P is a foreground point or a background point according to the normal change rate; storing the three-dimensional points corresponding to all foreground points in the point cloud B to complete point cloud segmentation; according to the method, based on the characteristic that the normal vector change rate of the thread point cloud is high, a segmentation scheme for point cloud screening based on the normal vector change rate is designed; the thread point cloud is segmented without the help of a theoretical value, the method is not restricted by workpiece machining deviation, and the segmentation precision of the thread point cloud is improved; and the thread point clouds of the stud, the planar threaded hole and the convex threaded hole can be effectively segmented, so that the universality is high, and the application scene is wider.

The application belongs to the technical field of image recognition segmentation, and discloses a medical image cell segmentation and tracking method, which comprises the following steps: step 1: data processing; feature extraction: the backbone part in the model is used to extract features from the preprocessed image, and the CSPDarknet structure is adopted in YOLOv8; step 3: FPN-PAN multi-scale feature fusion; step 4: Head prediction according to multi-scale features. The application realizes real-time tracking of the motion trajectory of cells by combining with a tracking algorithm such as deepsort. The method is mainly based on the YOLOv8 framework, and the Simam attention mechanism and the multi-scale proto method are adopted to optimize the model, so that the detection effect of YOLOv8 is further improved. The application can automatically complete the analysis and detection of medical images, is high in convenience and easy to use.

According to the tunnel face structural surface segmentation method and system, the tunnel face image is subjected to different-dimension remodeling, the tunnel face images of different dimensions after remodeling are obtained, the tunnel face images of different dimensions after remodeling are subjected to fusion processing, image features can be extracted from different scales by remodeling the image dimensions, and the segmentation accuracy of the tunnel face structural surface is improved. Enhancing recognition of details of the tunnel face; and feature mapping decoding processing is carried out on specific feature recognition results output by the two-stage specific feature extraction modules to obtain output features, pixel point recognition is carried out according to the output features, and effective segmentation of the structural surface in the tunnel face image can be realized.

The invention relates to the technical field of fuel cell production, in particular to a fuel cell bipolar plate adhesive tape segmentation method, system, medium and equipment based on gradient attention and a multi-scale void residual structure, a MobileNetV3-U-Net fusion architecture is adopted, lightweight MobileNetV3 is used as an encoder, high-level semantic features of an image are extracted step by step through a multi-layer inverse residual block, and the high-level semantic features of the image are extracted through a multi-scale void residual block. Meanwhile, a DG Attention module is embedded in the bottommost layer of the encoding stage, the DG Attention module guides attention weight calculation by extracting gradient features and long-strip-shaped features and strengthens feature capture of the edge and the slender structure of the adhesive tape, a decoder reserves multi-layer jump connection of U-Net, meanwhile, a traditional convolution operation is replaced with a Res2DilatedConv module, and therefore the decoding efficiency is improved. Multi-scale feature extraction is realized through channel grouping, multi-voidage convolution and cross-group feature fusion, the structural difference of the adhesive tape under different widths and bending degrees is adapted, and efficient and accurate segmentation of the adhesive tape region in the fuel cell bipolar plate is realized.

The utility model relates to the technical field of flow guide plate provides a car radiator flow guide structure, including flow guide plate, the edge position department of flow guide plate inside is provided with mounting hole, the edge position department of flow guide plate is fixed with flexible lip, the surface of flow guide plate is provided with the turbulence structure, the inside of flow guide plate is provided with the protection structure. The utility model discloses is provided with turbulence structure, through horizontal convex strip and longitudinal convex strip perpendicular each other and presents equidistant distribution, forms the regular grid structure on the flow guide plate surface. Further can play the effective segmentation and guiding effect to the airflow, makes the airflow more evenly cover radiator surface, promotes the evenness of whole heat dissipation, and through the circular convex point and the recess, can further produce the turbulence to the airflow, makes the airflow form tiny vortex. Further enhanced the heat exchange efficiency between the airflow and radiator, help radiator to dissipate heat more quickly.

The application discloses a Tibetan prosody structure prediction method based on syntax information, relates to the technical field of speech synthesis, and is applied to the prosody structure prediction field in Tibetan speech synthesis. The Tibetan prosody structure prediction method based on syntax information comprises the following steps: S1, Tibetan text analysis and virtual word connection quantization are completed, syntax probability, hierarchical focus and pause difference are extracted, and a Tibetan prosody prediction database is stored and constructed after preprocessing; S2, boundary discrimination is performed based on virtual word connection relationship strength, influence range hit marking and word sequence distance; S3, entropy feature analysis is performed through syntax role probability data and boundary determination data; S4, three-layer prosody boundary type sequences are generated according to boundary assignment, and over-dense pauses are inhibited; and S5, rhythm intensity evaluation is performed through pause duration difference and syntax information entropy data. The problems of pause confusion and increased understanding burden caused by the misplacement of prosody and semantic hierarchy in Tibetan long speech synthesis are solved.

The embodiment of the application discloses a laboratory early warning method and device based on temperature sensing, relates to the technical field of laboratory monitoring, and comprises the following steps: acquiring a laboratory image segmentation model based on day and night images; acquiring laboratory monitoring images in real time, wherein the laboratory monitoring images comprise thermal imaging images; judging whether there is a high-temperature danger in the current laboratory based on the thermal imaging images; if yes, segmenting the laboratory monitoring images based on the laboratory image segmentation model to generate laboratory segmentation images; and generating corresponding early warning information based on the thermal imaging images and the laboratory segmentation images. Through the all-day heat source influence monitoring of the laboratory based on the deep learning model, the use safety of the laboratory is effectively improved.

This invention belongs to the field of spectroscopic instrument technology, and particularly relates to a snapshot-type FP interferometric array spectroscopic imager. The snapshot-type FP interferometric array spectroscopic imager includes: a front imaging mirror group for imaging target light onto a primary image plane and collimating the light emitted from the primary image plane into parallel light; an FP interferometric array for multi-beam interference modulation of the parallel light, outputting modulated light carrying interference information; a microlens array for aperture-splitting imaging of the modulated light to form an interferometric image array; and a relay imaging mirror group for secondary imaging of the interferometric image array and coupling the imaging results to a detector. This invention achieves a good balance between spatial resolution, spectral bandwidth, and spectral resolution, while enabling real-time snapshot detection of rapidly changing targets. This invention eliminates the need for moving mirror scanning, resulting in higher structural stability; and by eliminating the beam splitter, it effectively reduces the difficulty of system assembly and alignment.

The invention discloses an RGB-T semantic segmentation method and system based on multi-attention guidance and hierarchical fusion, and relates to the technical field of image processing. The system is composed of a double-flow encoder, a discriminative local texture perception unit, a semantic-driven cross-modal fusion unit, a semantic enhancement unit and a multi-scale layered refinement decoder, and efficient fusion and analysis of multi-modal features in a complex traffic scene are achieved. According to the discriminative local texture perception method, saliency features are learned through multi-attention guidance and a self-adaptive gating mechanism, accurate modeling of shallow texture information is focused, and the distinguishing ability of a region of interest and a target edge is improved; according to the semantic-driven cross-modal feature fusion method, efficient aggregation of global contexts is realized through high-level semantic guidance and cross-modal feature interaction, and feature complementarity is enhanced, so that the semantic-driven cross-modal feature fusion method has significant advantages in analysis of small targets, long-distance targets and boundary regions. The decoder adopts a progressive fusion mode, an additional edge detection module does not need to be added, and the overall segmentation precision is improved.

The invention belongs to the field of image region segmentation, and particularly relates to an AI-based colorectal tumor image region segmentation method and system, and the method comprises the following steps: S1, obtaining original 3D colorectal tumor image data; according to the method, the high-quality 3D colorectal tumor image data is processed through the surface parameterization technology, the obtained 2D feature map is more suitable for processing the hybrid FMU-attention network model, and the processing efficiency and the processing speed of the hybrid FMU-attention network model are greatly improved; in addition, an AI framework is formed by jointly fusing a hybrid FMU-attention network model, a Vision Transformers module and a DeepLabV3 + model, the AI framework enables the boundary to be more accurate and the multi-scale performance to be more excellent, and it is ensured that the tumor can be effectively recognized and segmented regardless of the size.

This invention discloses a method for identifying photomask defects, belonging to the field of photomask inspection technology. The method includes: S1, acquiring a surface defect image; S2, preprocessing the surface defect image based on the imaging characteristics of the defect; S3, correcting the preprocessed surface defect image using a filtering algorithm; S4, performing colorimetric analysis on the corrected surface defect image to obtain particle regions; S5, segmenting the particle regions based on the morphological features of the defect to obtain defect islands; S6, performing morphological analysis on the defect islands to obtain the defect type and size. The photomask defect identification method provided by this invention automatically processes surface defect images and extracts defect islands by combining morphological and colorimetric analysis methods, improving the accuracy of defect identification, overcoming the limitations of differences in imaging systems across different devices, achieving automatic batch processing, and saving manpower and material resources in mass production.

The present application provides a wafer processing method and a wafer processing apparatus capable of efficiently processing a wafer into individual chips regardless of the kind of wafer or the kind of processing. According to the present application, a wafer processing method and a wafer processing apparatus are provided. The wafer processing method includes a holding step of holding a wafer (10) in a holding unit (4), a damage layer forming step of positioning a convergence point (P1) of a shock wave (PL3) at the wafer (10) held in the holding unit (4) to form a damage layer (S1) in a region to be divided, and a dividing step of dividing the wafer (10) into individual chips (12') with the damage layer (S1) as a starting point. The wafer processing apparatus includes a holding unit that holds a wafer, and a damage layer forming unit that positions a convergence point of a shock wave at the wafer held in the holding unit to form a damage layer in a region to be divided.

A method is provided for studying the behaviour of an animal in an experimental area using a camera to obtain a video stream having a plurality of images of the area, removing background noise from the images, extracting parameters from the images to form multidimensional data points, clustering data points so that each cluster represents and animal behaviour and outputting a matrix of results. Apparatus for performing the method is also provided.

The application discloses a train wheelset fluorescent magnetic powder flaw detection image crack identification, positioning and measurement method, which comprises the following steps: partitioning a train wheelset and collecting a crack magnetic powder image of the train wheelset; pre-processing the collected crack magnetic powder image to highlight crack features; segmenting and extracting the crack features; and identifying, positioning and measuring the crack according to crack feature information. According to the shape characteristics of the crack, the application proposes a crack feature segmentation algorithm combining the crack area and the area of its circumscribed rectangle and the aspect ratio of the circumscribed rectangle of the connected domain, realizes accurate extraction of the wheelset crack features, and identifies, positions and measures the crack. Compared with other crack detection methods, the application not only realizes nondestructive detection of the surface crack of the train wheelset, but also effectively segments the features of the train wheelset crack, facilitates identification, positioning and measurement of the crack of the train wheelset, and thus realizes automatic wheelset crack magnetic powder flaw detection.

The utility model relates to the technical field of house building construction, disclose a kind of big slope concrete pouring fixing device, including two fixed plates, two the fixed plate is arranged along vertical direction with interval, a plurality of fixed rib is provided on the fixed plate located at bottom, the top of fixed rib is fixedly connected after being passed out from upper layer the fixed plate with upper layer the fixed plate;Blocking net is arranged between the fixed rib of upper and lower sides, and blocking net extends to both sides along horizontal direction;The utility model can reduce the fluidity of big slope roof concrete, fix the board surface concrete and subsequent concrete protective layer, improve roof pouring quality, reduce potential safety hazard.

The invention provides a distributed seismic data segmentation and extraction device, method and application, and belongs to the field of seismic exploration. The method comprises the following steps: acquiring original seismic data by adopting a plurality of distributed seismic prospecting instruments, and storing the original seismic data by taking time as a label; the data processing center issues a single-shot segmentation extraction command to the plurality of distributed seismic detectors through the narrowband Internet, and the plurality of distributed seismic detectors segment and extract seismic data according to the single-shot segmentation extraction command; when a segmentation and extraction ending command is received, the plurality of distributed seismic detectors transmit the segmented and extracted seismic data to a data processing center through a 5G network; and the data processing center synthesizes and stores a shot gather file. According to the invention, while the integrity of the seismic data is ensured, the transmission of invalid seismic data is effectively reduced, and the processing demand of a data processing center is reduced, so that the seismic data processing efficiency is greatly improved.