13results about How to "Accurate and effective extraction" patented technology

The invention relates to the technical field of fault detection, and discloses a remote electric energy meter fault detection and identification system and method, and the system comprises a data collection module, a data integration module, a data inspection module, a feature extraction module, a fault analysis module, and a fault alarm module. According to the method, multiple electric energy meter target data are collected and integrated into a corresponding data set, and a complete and ordered data basis is provided for analysis; abnormal values are recognized and replaced by calculating the average value and the standard deviation, the data quality is improved, and the accurate analysis result is guaranteed; key parameters covering multiple aspects are accurately extracted from a data set, and a powerful basis is provided for fault analysis; from metering, hardware and communication fault dimensions, according to comparison of extracted parameters and threshold values, fault types and positions are accurately judged in multiple directions; the fault alarm module remotely transmits an analysis result to a monitoring center, and generates an alarm message containing detailed information such as an electric energy meter number, a fault type, a parameter value, a threshold value and the like when a fault occurs, thereby facilitating timely maintenance by maintenance personnel.

The invention provides a low-illumination image enhancement method based on adaptive brightness enhancement and high-fidelity color correction. Comprising the following steps: constructing a training data set, and performing data enhancement on paired low-illumination / normal-illumination images to obtain an image data set for training; designing a low-illumination image enhancement network based on adaptive brightness enhancement and high-fidelity color correction; designing a loss function for training the designed network; training the designed network by using the training data set; and inputting a to-be-measured low-illumination image into the trained network to generate a normal-illumination image. According to the method, the adaptive brightness improvement strategy and the color correction under the multi-color space constraint are utilized, the image brightness can be adaptively enhanced, the image color is subjected to high-fidelity correction, and a high-quality image with proper brightness and high-fidelity color is obtained; the problems of uneven brightness, color distortion, detail loss and the like of the low-illumination image are effectively solved.

The invention provides a naked eye light field 3D display image fusion method and system, and the method comprises the steps: collecting a target image and a reference image, and carrying out the depth extraction and viewpoint image generation based on the target image and the reference image, so as to obtain a first image and a second image; performing mapping filtering processing on the first image and the second image to obtain a first mapping image and a second mapping image; performing preliminary fusion processing on the first mapping image and the second mapping image to obtain an initial fusion image; performing iterative segmentation and image parallax correction processing on the initial fusion image to obtain a corrected image; according to the method, the depth information in the target image can be effectively and accurately extracted, and meanwhile, the influence of noise on image fusion can be reduced, so that the finally obtained naked eye light field 3D display image is closer to a real 3D effect and is more accurate.

The invention relates to the technical field of computer vision, and provides a body surface three-dimensional point cloud target area extraction method and system, and the method comprises the following steps: carrying out the three-dimensional scanning and shooting of the body surface of a target patient equipped with a windowing film, and obtaining the three-dimensional point cloud data and two-dimensional image of the body surface of the target patient; constructing a mapping relation based on the three-dimensional point cloud data and the two-dimensional image; inputting the two-dimensional image into an image segmentation model to obtain an ROI recognition result about the outer contour of the diaphragm windowing area in the two-dimensional image; three-dimensional points in the three-dimensional point cloud data are traversed, the three-dimensional points are mapped to two-dimensional points of a two-dimensional image based on the mapping relation, whether the three-dimensional points are located in an ROI recognition result of a view angle where the two-dimensional image is located or not is judged, if yes, the three-dimensional point # imgabs0 # is added into a target area set, and if not, the next three-dimensional point is judged; and after traversal is completed, outputting the target region set as a target region extraction result.

The invention discloses a data model conversion method and device and a computer readable storage medium, and relates to the technical field of computers and the technical field of software. The method comprises the steps that a model transfer relation graph and a model conversion rule are obtained, and the model transfer relation graph is used for representing object entities in a to-be-converted data model and a target data model and the relation between the object entities in a graph mode; the model conversion rule is a rule on which the to-be-converted data model is converted into a target data model; converting the to-be-converted data model into an intermediate data model according to the model transfer relation graph and a model conversion rule; detecting model difference information of the intermediate data model and the target data model; and adjusting the intermediate data model according to the model difference information until the intermediate data model is adjusted to be the target data model. The technical problem that in the prior art, in the upgrading and reconstruction process of an information system, the data model conversion efficiency is low is solved.

The present invention relates to the field of computer vision technology, and proposes a method and system for extracting a target area of ​​a body surface three-dimensional point cloud, wherein the method comprises the following steps: performing three-dimensional scanning and photographing on a target patient's body surface equipped with a windowed diaphragm to obtain three-dimensional point cloud data and a two-dimensional image of the target patient's body surface; constructing a mapping relationship based on the three-dimensional point cloud data and the two-dimensional image; inputting the two-dimensional image into an image segmentation model to obtain an ROI recognition result of the outer contour of the diaphragm windowed area in the two-dimensional image; traversing the three-dimensional points in the three-dimensional point cloud data, mapping the three-dimensional points to two-dimensional points of a one-dimensional image based on the mapping relationship, and judging whether they are located within the ROI recognition result of the viewing angle of the two-dimensional image, and if so, adding the three-dimensional point to a target area set, otherwise judging the next three-dimensional point; after completing the traversal, outputting the target area set as a target area extraction result.

The invention provides an extraction method, an extraction system and extraction equipment for a key fragment of a metal organic framework material. The extraction method comprises the following steps: obtaining a crystal structure file of the metal organic framework material; converting the crystal structure into a crystal diagram; obtaining a key fragment sub-graph according to the crystal graph and an extraction rule; based on the key fragment subgraph, identifying a hydrogen deficiency site, obtaining a hydrogen deficiency site coordination environment, complementing hydrogen atom information, and obtaining a key fragment; and associating the crystal structure of the metal organic framework material with the key fragment to obtain a database, thereby completing the extraction of the key fragment of the metal organic framework material. According to the technical scheme provided by the invention, the key fragments of the metal organic framework material can be accurately and effectively extracted, so that the metal organic framework structure database is established and is used for calculating and analyzing an electronic structure and screening the metal organic framework material with special performance.

The application discloses a bearing fault classification method and related device. The method comprises the following steps: inputting a first vibration signal collected under a plurality of known fault types of a bearing into a preset dictionary learning algorithm model after processing, and constructing a sparse dictionary containing a plurality of fault type training samples; processing a second vibration signal collected for fault classification to obtain a test sample, performing sparse optimization solving on the test sample, and obtaining a first sparse coefficient; respectively reconstructing the first sparse coefficient and the training sample under a plurality of fault types, and obtaining a plurality of fault type corresponding reconstruction signals; comparing the test sample with the plurality of fault type corresponding reconstruction signals, and determining the fault type corresponding to the test sample according to the comparison result. The application can accurately and effectively realize bearing fault classification diagnosis.

The present invention relates to a drainage pipe defect recognition model training method and a drainage pipe defect recognition method. The technical solution of the present invention is a drainage pipe defect recognition model training method, which is characterized by: obtaining a training sample, the training sample has a drainage pipe image and a defect type corresponding to the image, and there are n defect types in the training sample; preprocessing the drainage pipe image, including graying the drainage pipe image; extracting image features corresponding to each defect type from the preprocessed drainage pipe image corresponding to each defect type; training the initial binary classification model based on the image features corresponding to the i-th defect type and the image features corresponding to the j-th defect type among the n defect types, and obtaining defect recognition models for the i-th and j-th defect types, and obtaining defect recognition models in total. The present invention is applicable to the field of image recognition technology.

The present invention relates to a method for identifying a live face, the method for identifying a live face comprising the following steps: (1) collecting face image information; (2) generating at least one set of face image sequences from the face image information; (3) extracting local face features from the at least one set of face image sequences; (4) filtering out a low-dimensional feature vector of the face image information from the local face features; (5) inputting the low-dimensional feature vector into a trained classifier for live face image recognition; and (6) outputting a recognition result. Accordingly, the present invention also relates to a system for identifying a live face. The method and system according to the present invention can not only quickly and accurately identify a live face but also meet the high real-time requirements of embedded devices.

The present invention provides a naked-eye light field 3D display image fusion method and system, the method comprises collecting a target image and a reference image, performing depth extraction and viewpoint image generation based on the target image and the reference image to obtain a first image and a second image; performing mapping and filtering processing on the first image and the second image to obtain a first mapping image and a second mapping image; performing preliminary fusion processing on the first mapping image and the second mapping image to obtain an initial fused image; performing iterative segmentation and image parallax correction processing on the initial fused image to obtain a corrected image; performing projection transformation on the corrected image to obtain a naked-eye light field 3D display image. The present invention can effectively and accurately extract depth information from the target image, and can also reduce the influence of noise on image fusion, so that the finally obtained naked-eye light field 3D display image is closer to the real 3D effect and more accurate.

The invention belongs to the technical field of end plug welding seam detection, and aims to solve the problems of large calculation amount, relatively low accuracy and efficiency and the like caused by directly processing an image by a traditional image processing method when a region of interest of an end plug welding seam ray digital image is extracted. The invention discloses an end plug welding seam ray digital image region-of-interest extraction method, which comprises the following steps: carrying out image preprocessing and image enhancement on an acquired image, then carrying out semantic annotation on the image, sending the image into an improved residual module network, introducing an attention mechanism, and using a cross entropy loss function to accelerate the calculation speed and improve the model efficiency. And the improved network model is trained and is used for testing a set observation effect. According to the method, the region of interest can be effectively and accurately extracted, the detection complexity is reduced, and the detection efficiency is improved.

The invention provides an Alzheimer's disease risk prediction system and method based on continuous data, and the system comprises a data collection module which is used for collecting the multi-dimensional continuous data of an individual; the data analysis module is used for analyzing the multi-dimensional continuous data, collecting data features of the multi-dimensional continuous data and outputting posture features of the individuals according to the data features; and the prediction module is used for determining a behavior path change point of the individual under the target time sequence based on the attitude features, and performing Alzheimer's disease risk prediction based on the behavior path change point. The accuracy and reliability of risk prediction of the Alzheimer's disease are ensured, and great convenience is provided for disease treatment.