Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

90results about How to "Improve detail performance" patented technology

Deep learning-based corrugated pipe inner wall multi-type flaw detection method

The invention discloses a corrugated pipe inner wall multi-type flaw detection method based on deep learning, and the method comprises the following steps: S1, collecting an initial corrugated pipe inner wall image and corrugated pipe inner wall images in different illumination environments, and generating an original image; s2, preprocessing the original image to generate a preprocessed image; s3, extracting surface geometric overall features of the inner wall of the corrugated pipe from the preprocessed image through a pre-trained ResNet101 network; s4, inputting the preprocessed image and the surface geometric overall features into an improved StyleGAN3 network to generate an enhanced image; s5, inputting the enhanced image into the multi-scale adaptive deep network, and outputting a defect detection result; and S6, performing post-processing on the flaw detection result, and outputting a final flaw detection result. According to the method, deep learning and geometric characteristic embedding technologies are adopted, so that the defects of the inner wall of the corrugated pipe are accurately detected, and the method has the advantages of high precision, high robustness and comprehensive adaptability.
Owner:HEFEI MITSUKOSHI SEMICON TECH CO LTD

Three-dimensional model data management system and method for engineering surveying and mapping

The invention relates to the technical field of data processing and analysis, in particular to a three-dimensional model data management system and method for engineering surveying and mapping. The method comprises the following steps: acquiring engineering surveying and mapping data; preprocessing the engineering surveying and mapping data to obtain standard engineering surveying and mapping data; performing multi-dimensional feature recognition on the standard engineering surveying and mapping data to obtain multi-dimensional surveying and mapping feature data; performing key engineering surveying and mapping information extraction on the multi-dimensional surveying and mapping characteristic data to obtain key engineering surveying and mapping information; performing association analysis on the multi-dimensional surveying and mapping characteristic data and the key engineering surveying and mapping information to generate multi-dimensional engineering surveying and mapping associated data; monitoring the engineering surveying and mapping data in real time to obtain engineering surveying and mapping real-time monitoring data; through a data processing technology, a geographic information system technology and a three-dimensional mapping technology, integrated management of multi-source heterogeneous data is realized, and dynamic updating management of the data is realized, so that the presentation accuracy of the engineering surveying and mapping three-dimensional model is improved.
Owner:JINCHENG HEAVY IND MASCH (DALIAN) CO LTD

Open type earphone audio optimization method based on adaptive compensation algorithm

The invention provides an open earphone audio optimization method based on a self-adaptive compensation algorithm, and aims to improve the audio auditory experience of a user. The method comprises the following steps: firstly, acquiring an environmental noise signal through an earphone microphone, receiving a target audio signal, analyzing auditory perception of a user in a current environment by combining an ISO226 loudness curve and the like, and determining auditory sensitivity of the user to different frequency bands (including high frequency, intermediate frequency and low frequency); and then, enhancing the high-frequency part, deepening the low-frequency part and performing smooth transition compensation on the intermediate-frequency part by using a self-adaptive compensation algorithm. And then, dynamically adjusting the low-frequency part by using a hyperbolic filter, generating an adjusted audio signal, and generating a harmonic signal based on the amplitude and frequency characteristics to further optimize the tone quality. Finally, the optimized audio signal and the generated harmonic signal are provided for a user together, and the effects of automatically optimizing audio output in different environments and enhancing tone quality performance are achieved.
Owner:GUANGDONG MINGYANG SMART TECH CO LTD

High-resolution positioning method for noise source imaging

The invention provides a high-resolution positioning method for noise source imaging. The high-resolution positioning method comprises the following steps: designing a sensor array layout and selecting sensors; a multi-channel acquisition synchronous acquisition sensor and records signals to obtain acquisition data; performing signal processing on the acquired data by using adaptive filtering, machine learning calibration, time-frequency analysis and intelligent beam forming technologies to obtain standard data; based on the standard data, performing high-resolution direction estimation by using an optimized multi-signal classification algorithm, and removing a sidelobe effect by using an optimized CLEAN algorithm to complete processing of the standard data; and based on the processed standard data, reconstructing a spatial distribution image of the noise source by using an inversion technology, and dynamically adjusting the resolution of the reconstructed spatial distribution image by using an adaptive algorithm to realize high-resolution positioning of noise source imaging. According to the invention, high-resolution positioning can be realized, the reliability and precision of noise detection are improved, and noise control and management are more effective.
Owner:101 INST OF THE MINISTRY OF CIVIL AFFAIRS

Three-dimensional reconstruction method and device based on structured three-dimensional Gaussian representation

The embodiment of the invention discloses a three-dimensional reconstruction method based on structured three-dimensional Gaussian representation. The method comprises the following steps: acquiring a target image containing a target object at multiple visual angles; generating an unstructured Gaussian primitive representation corresponding to the target object according to the target image; establishing a transmission matrix describing the unstructured Gaussian primitive distribution and the structured Gaussian primitive distribution, solving the transmission matrix by using an optimal transmission algorithm, and generating a structured Gaussian primitive representation corresponding to the target object according to the optimal transmission matrix; obtaining a text description for the target object, and complementing the structured Gaussian primitives according to the text description by using a diffusion model obtained by pre-training; and performing three-dimensional reconstruction by using the complete structured Gaussian primitive. According to the method, the optimal transmission algorithm is used for carrying out structured representation on the three-dimensional Gaussian primitives, and text description is used for complementing the Gaussian primitives through the diffusion model, so that the quality of three-dimensional reconstruction is improved.
Owner:ZHEJIANG WUWEN ZHIXING TECHNOLOGY CO LTD

Image restoration method based on multi-scale residual module and feature fusion

The invention discloses an image restoration method based on a multi-scale residual module and feature fusion, and belongs to the technical field of image processing. The multi-scale residual module designed by the invention can effectively capture feature information of different scales of the image, the quality of the repaired image is improved from a fine texture to a macroscopic structure, a feature fusion strategy fusing detail branches and semantic branches can fully integrate features of different levels, the detail expression and overall coordination of the image are enhanced, and the image quality is improved. Through fusion of multi-scale features, limitation possibly brought by single-scale features is avoided, and the repaired image is more real and fine in color, texture and structure. And the multi-scale residual error module can relieve the gradient disappearance problem, the calculation efficiency is high, the image restoration task can be completed in a short time, and the introduced and designed dual-channel global attention module has good robustness, so that the method can stably play a role in various complex image damage conditions.
Owner:ANHUI UNIVERSITY OF TECHNOLOGY

Smart city video monitoring adjustment method and system based on 5G network

The invention discloses a smart city video monitoring adjustment method and system based on a 5G network, and relates to the field of smart cities, and the method comprises the steps: 1, drawing a city map through a geographic information system technology, deploying a high-resolution monitoring camera supporting the 5G network, and an edge computing device, a video stream in an H.265 compressed format is captured in real time by using a built-in image sensor. The high-resolution camera is combined with high-speed transmission of a 5G network, the real-time performance of monitoring data is ensured, emergencies are quickly responded, real-time processing of video data and sensor data is achieved, the high-resolution monitoring camera is accessed through a Video Capture class of OpenCV, target detection is conducted on an area of interest in combination with a YOLO algorithm, and the real-time processing of the video data and the sensor data is achieved. The target identification accuracy and processing efficiency are improved, the edge computing device is used for carrying out data summary on the segmented frame area, the processing delay is dynamically adjusted according to the real-time network and the computing load, and efficient operation of the monitoring system under different conditions is ensured.
Owner:SMIC (HARBIN) CO LTD

Nuclear medicine imaging method and system

The invention relates to the technical field of medical imaging, in particular to a nuclear medicine imaging method and system, and the method comprises the steps: patient preparation, tracer agent distribution detection, image enhancement, multi-target dynamic imaging, image quantitative analysis and data fusion, and result output and diagnosis suggestion. A preliminary image of tracer agent distribution in the body of a patient is obtained through a multi-modal imaging device, a high-resolution image is generated through a convolutional neural network noise suppression algorithm, a contrast ratio self-adaptive enhancement algorithm and a super-resolution reconstruction algorithm based on a generative adversarial network, and accurate dynamic analysis is carried out. In combination with a random forest algorithm and a time sequence analysis model, performing quantitative analysis on the radioactive intensity, the volume and the metabolic rate of the target spot, and generating personalized diagnosis suggestions; according to the invention, the precision and diagnosis effect of nuclear medicine imaging are greatly improved, and the method is suitable for early diagnosis, curative effect evaluation and disease progress monitoring of clinical diseases.
Owner:GUIZHOU PROVINCIAL PEOPLES HOSPITAL

Environment perception data processing method and system based on intelligent helmet

The invention relates to the field of data processing, in particular to an environment perception data processing method and system based on an intelligent helmet. The method comprises the following steps: acquiring a scene image in front of a user and multi-source environment real-time sensing data; performing scene attribute state analysis on the multi-source environment real-time sensing data, and constructing a real-time scene attribute state diagram; performing contrast dynamic balance adjustment on the scene image in front of the user, performing super-resolution fusion reconstruction, and constructing a super-resolution scene image; performing object visual identification on the super-resolution scene image, and performing object bounding box segmentation processing to obtain a plurality of accurate scene object bounding boxes; and carrying out optical flow motion change analysis on the plurality of accurate scene object bounding boxes, and carrying out space-time scene structure evolution analysis so as to obtain dynamic scene evolution characteristic data. According to the method, the data processing efficiency of the environment perception data is improved, and the environment perception data is visualized more visually.
Owner:ORIGJOY

Terrain-combined weather radar base data dynamic resolution three-dimensional structure reconstruction method

The invention discloses a terrain-combined weather radar base data dynamic resolution three-dimensional structure reconstruction method, which realizes three-dimensional structure reconstruction through combination of a digital elevation model (DEM) and weather radar base data. The method comprises the steps of cutting DEM topographic data and dumping the DEM topographic data into grid data in a binary form, generating a bitmap as a mapping source, establishing an elevation angle index array based on weather radar base data, establishing an azimuth angle index array based on a conical scanning surface so as to quickly map an angle, determining a rectangular reconstruction area and establishing a three-dimensional grid, traversing the three-dimensional grid to obtain an elevation angle, an azimuth angle and a distance combination array of grid points so as to obtain grid point data, smoothing the grid point data, re-sampling terrain grid data, constructing a terrain grid surface, obtaining a mapping image of the terrain grid surface, and rendering a three-dimensional body; the method has the advantages that the method does not need to depend on online service, and dynamic high-precision three-dimensional structure reconstruction of radar echoes in a specified area can be realized while the interaction performance is ensured.
Owner:NINGBO MUNICIPAL METEOROLOGICAL OBSERVATORY

Method and system for detecting inclusion defects on surface of wiredrawing steel plate based on machine vision

The invention discloses a machine vision-based wiredrawing steel plate surface inclusion defect detection method and system. The method comprises the following steps: S1, acquiring a multi-light-source image sequence on the surface of a wiredrawing steel plate and carrying out brightness standardization; s2, decomposing the image sequence after brightness standardization through a Gaussian-Laplacian pyramid, and performing Laplacian pyramid fusion and image reconstruction to obtain a fused image; s3, carrying out directivity analysis on the wiredrawing texture in the fused image, extracting main direction features of the wiredrawing texture, and obtaining a wiredrawing direction consistency feature map; s4, performing feature extraction and feature fusion on the fused image and the wiredrawing direction consistency feature map based on a double-flow feature extraction network, and performing global average pooling operation; and S5, constructing and training a defect classification network, and predicting the defect category of the surface of the wiredrawing steel plate. The problems existing in the existing wiredrawing steel plate surface defect detection technology can be overcome, and the defect of inclusion on the surface of the wiredrawing steel plate is accurately detected.
Owner:HUNAN SHUNXIN METAL PRODUCTS TECHNOLOGY CO LTD

A digital human image reproduction method and system based on NeRF

The present invention discloses a method and system for replicating digital human images based on NeRF, and relates to the technical field of digital human image replicating. The method comprises: collecting digital human images to obtain training data sets; expanding and generating training samples; reconstructing a model using the expanded digital human image data training data; extracting features from the results of digital human image replicating output by the autoencoder, and inputting the feature vector after feature extraction into the classifier model to perform digital human image data screening. The present invention improves the details and authenticity of the generated samples, makes the expanded data closer to the real collected images in terms of fidelity, enhances the adaptability and stability when screening digital human image data, makes the screening model more adaptable when facing diversified replica data, improves the classification effect, and finally ensures the high quality and consistency of the replica data.
Owner:QUANZHOU JINTONG INFORMATION TECHNOLOGY CO LTD

Image optical anti-shake method for endoscope camera system

The invention provides an image optical anti-shake method for an endoscope camera system. The method comprises the following steps: acquiring motion data of an endoscope camera in real time; based on the collected motion data, using an image processing algorithm to analyze and identify the mode and degree of image shake, and calculating the compensation amount required by optical and electronic anti-shake; a high-precision optical anti-shake assembly is integrated, and according to the calculated compensation amount, an optical path or the position of a lens is rapidly adjusted through the optical anti-shake assembly so as to compensate for image offset caused by movement of the camera; an electronic anti-shake assembly is integrated, reading data of an image sensor or parameters of an image processing chip are adjusted through an electronic control system, and residual shake is further eliminated; sensor data and image quality are monitored in real time, optical anti-shake parameters are dynamically adjusted, and continuous image stability and clearness are ensured. Real-time, stable and clear output of images is achieved through the optical technology, and a better auxiliary diagnosis means can be provided for doctors.
Owner:ZHUHAI DI PU MEDICAL TECH CO LTD

Curtain wall reverse modeling method and system based on point cloud data

The invention provides a curtain wall reverse modeling method and system based on point cloud data, and is applied to the field of modeling data processing. The method can timely discover the defects of the model and improve the overall quality by accurately identifying irregular regions and performing error analysis, enhances the smoothness and detail performance of the model by applying a curved surface fitting algorithm and a hierarchical structure reconstruction technology, enables the model to be closer to a real object, and improves the accuracy of the model through structural feature-oriented local refinement. Geometric complexity is improved in a targeted manner, and uniformly distributed new vertexes are inserted to reduce zigzag and unsmooth areas, thereby improving overall fluency and visual effect.
Owner:SHENZHEN BODA CONSTR GRP CO LTD

Coal gangue detection method based on multi-angle perception and mixed scale Transform feature aggregation

The invention belongs to the technical field of artificial intelligence and coal mine production, and particularly relates to a coal gangue detection method based on multi-angle perception and mixed scale Transform feature aggregation, which comprises the following steps: firstly, acquiring a real-time video image of a conveying belt through a multi-angle camera device, and fully covering each area on the conveying belt; secondly, performing feature extraction and aggregation on the video image by adopting a mixed scale Transform model, fully fusing feature information under different scales, and enhancing the recognition capability of the model on the coal gangue; and finally, the aggregation features are analyzed by using a coal gangue detection algorithm, and high-precision coal gangue detection is realized. Real-time video images of the conveying belt are obtained through the multi-angle camera device, all areas of the conveying belt are fully covered, and therefore any blind area and any dead angle are eliminated. Advanced image preprocessing technologies such as de-noising and image enhancement are combined, so that the definition and detail performance of the image are remarkably improved.
Owner:SHANXI INFORMATION IND TECH RES INST CO LTD

Image acquisition method and system of dual-wavelength circular polarization camera

The invention discloses an image acquisition method and system for a dual-wavelength circular polarization camera, and the method comprises the steps: setting two independent nanosecond pulse laser sources, and respectively transmitting lasers with different wavelengths; respectively carrying out left-hand circular polarization and right-hand circular polarization treatment on the two lasers; integrating the two nanosecond pulse laser beams subjected to circular polarization treatment into one beam, wherein the two beams are completely overlapped and are coaxial; the integrated nanosecond pulse laser beams are irradiated to a target object, and reflected light is collected through a double-spectrum camera to form two images with different wavelengths; and distributing weights for the image according to the importance of the polarization information, and generating a comprehensive image by adopting an image fusion technology. According to the image acquisition method and system of the dual-wavelength circular polarization camera, two independent nanosecond pulse laser sources are arranged to emit short-wavelength laser and long-wavelength laser respectively, the two beams of laser are subjected to left-handed and right-handed circular polarization processing and then integrated into one beam, and efficient imaging in a complex environment is achieved.
Owner:BAORUI LASER TECH (CHANGZHOU) CO LTD

Three-dimensional geographic target construction analysis system based on neural network model

The invention discloses a three-dimensional geographic target construction analysis system based on a neural network model, relates to the technical field of three-dimensional geographic target construction, and adopts a generative adversarial GANs network in deep learning to carry out super-resolution reconstruction so as to improve the detail performance of a three-dimensional model; a confrontation training mode of a generator and a discriminator is adopted, the generator is responsible for generating a high-resolution three-dimensional model, and the discriminator is responsible for judging whether the generated model is real or not; in combination with a progressive generation strategy, the resolution and complexity of the model are gradually increased; real-time data processing and analysis are carried out by using a stream processing framework Apache Kafka, so that the model can reflect topographic changes in time; sensor data and satellite images are collected and processed in real time through a flow processing pipeline, and topographic change detection and analysis are carried out; and an incremental updating strategy is adopted, so that the delay of data processing and the consumption of computing resources are reduced.
Owner:BEIJING GUANTIAN TECH CO LTD

Low-illumination image enhancement method and device based on improved normalized stream

The embodiment of the invention discloses a low-illumination image enhancement method and device based on an improved normalized stream. The method comprises the following steps: acquiring a low-illumination image to be enhanced; performing normalization processing on the to-be-enhanced low-illumination image to obtain a normalized spectrum; performing edge extraction processing on the normalized spectrum to obtain an edge spectrum; performing feature fusion on the to-be-enhanced low-illumination image, the normalized spectrum and the edge spectrum to obtain a feature fusion image; performing feature extraction on the feature fusion image by adopting a feature extraction module to obtain a feature extraction image; performing feature mapping on the feature extraction image by adopting a feature mapping module to obtain a feature mapping image; and inputting the feature mapping image into the normalized stream reversible network, and outputting a normal illumination image. By enhancing the low-illumination image through the method, the enhancement effect can be improved.
Owner:XIDIAN UNIV

Animation motion capture device for virtual reality technology

The invention discloses an animation motion capture device for a virtual reality technology, and particularly relates to the technical field of virtual reality, which comprises a data acquisition module for capturing and collecting user motion and environment information, and a data processing module for analyzing, converting and optimizing the acquired data. Through cooperative work of the data acquisition module, the data processing module, the rendering feedback module and the like, accurate capture and real-time feedback of user actions are achieved, it is ensured that physical attributes of a virtual environment and a real environment are highly consistent, the data acquisition module can obtain multi-dimensional information such as somatosensory, expressions and environments of a user at high precision, and the user experience is improved. And the data processing module accurately identifies and optimizes user actions through a deep learning algorithm and a skeleton mapping technology, and the rendering and feedback module generates vivid visual, tactile and sound effect feedback, so that high immersive experience is brought, and fluency and consistency of actions and feedback are guaranteed.
Owner:WUHAN INST OF TECH

Inspection blurred image recognition method and device based on regional target detection algorithm

The invention relates to the technical field of image processing, and relates to an inspection fuzzy image recognition method and device based on a region target detection algorithm. The method comprises the following steps: acquiring an original image shot by a to-be-identified device; carrying out image preprocessing, and inputting the image into the trained region target detection network for processing; and carrying out deblurring processing by adopting a multi-Wiener filtering algorithm to obtain a reconstructed clear image, and finally carrying out image recognition. According to the invention, a region target detection network based on YOLO network modification is adopted, so that the to-be-deblurred target region in the equipment image can be accurately identified while the model is lightened; according to the method, a multi-Wiener filtering algorithm is adopted, so that the to-be-deblurred target area in the equipment image can be effectively deblurred; according to the method, the clear image can still be generated under the condition that the inspection robot moves, the definition and detail performance of the image are greatly improved, the recognition accuracy is further improved, and misjudgment caused by blurring is reduced.
Owner:HUAIHE ENERGY HUAINAN PANJI POWER GENERATION CO LTD

Machine vision unit applied to intelligent security management system of intelligent community

The invention discloses a machine vision unit applied to a smart community intelligent security management system, and relates to the technical field of image processing, and the machine vision unit comprises an image obtaining module which obtains a real-time video stream monitored at night, and carries out the frame cutting to obtain a plurality of frames of images; the multi-frame image fusion module is used for processing the multi-frame image by using a multi-exposure fusion algorithm to generate a fused image; the self-adaptive correction module is used for self-adaptively adjusting the brightness and the contrast ratio of the fused image and carrying out detail enhancement and color correction; the image processing module is used for removing noise and enhancing details through bilateral filtering and a multi-scale Retinex algorithm; and the image post-processing module is used for performing color balance adjustment to ensure that the images are natural and consistent. Through multi-frame image fusion, adaptive correction, multi-scale detail enhancement and color balance processing, the definition, detail expression and recognition precision of the image in a night monitoring scene are remarkably improved, the adaptability and robustness of the system are enhanced, and meanwhile redundant information and resource consumption in the processing process are reduced.
Owner:SHANGHAI TIANQI INTELLIGENT BUILDING CO LTD

Coal mine underground video dust filtration monitoring method based on model fusion

The invention provides a coal mine underground video dust filtration monitoring method based on model fusion, and belongs to the technical field of coal mine underground video dust filtration monitoring. The technical problem to be solved is to provide the coal mine underground video dust filtration monitoring method based on model fusion. According to the technical scheme, video data of an underground operation area environment are collected in real time through a camera installed in an underground coal mine, the video data are uploaded to a server arranged in a roadway in real time, the server firstly intercepts image frames from the underground coal mine video, timestamp marking is conducted on the image frames, and the image frames are sent to the roadway; carrying out histogram equalization preprocessing on each image frame, and improving the brightness and contrast of the collected video image to realize image enhancement; the server further adopts a dark channel prior method to carry out data recovery on the video image, a fuzzy effect caused by dust on the image is removed, and the real definition of the image is recovered; the method is applied to coal mine underground video dust filtration monitoring.
Owner:YUWU COAL CO LTD OF SHANXI LUAN GRP

Front-end polymerization type 3D printing ink and method for preparing hydrogel through front-end polymerization-3D printing

The invention discloses a method for preparing hydrogel through front-end polymerization-3D printing and front-end polymerization type 3D printing ink, and relates to the technical field of functional polymer material preparation. Ink formed by combining Carbomer 940 particle hydrogel, acrylamide and acrylic acid is prepared on the basis of blocking phase change, the printability of the ink is studied, front-end polymerization and a 3D printing technology are combined, a gel precursor solution and Carbomer 940 serve as main components of the 3D printing ink, a front-end polymerization method is adopted, a gel support of a specific structure and morphology is printed, and the 3D printing ink has the advantages of being simple in preparation process, low in cost and high in practicability. A direct ink writing (DIW) technology is coupled with a front-end polymerization reaction, a polymer three-dimensional model is rapidly prepared with minimum energy input, meanwhile, the front-end end face is spread forwards along with the moving track of a printing needle in the printing process, ink is cured into a polymer in situ through high heat release of the polymerization reaction, and the self-supporting capacity of the model is improved.
Owner:NANJING TECH UNIV

Three-dimensional imaging reconstruction system in minimally invasive interventional operation based on adaptive neural network

The invention relates to a three-dimensional imaging reconstruction system in a minimally invasive intervention operation based on an adaptive neural network. Converting the two-dimensional slice data into a three-dimensional image layer by layer through a self-adaptive neural network, and identifying the structural relationship between the layers through the network and performing three-dimensional reconstruction; marking the paired two-dimensional image and three-dimensional model; the two-dimensional image data is enhanced by applying an image transformation technology, so that the data diversity is increased; generating new two-dimensional image data by using a generative adversarial network (GAN), and combining the new two-dimensional image data with the existing data; in the operation process, a three-dimensional preview of the contour is rapidly generated through a downsampling technology; a plurality of sensors and a computer vision technology are used for tracking the position and the movement track of the surgical instrument in real time; adjusting transparency and colors of different tissues in the reconstructed image according to operation requirements; calculating a shortest path between two points in an operation area by using a Dijkstra algorithm, marking the operation area, and controlling the segmentation and demonstration of the operation area by using an area growing algorithm and a threshold segmentation technology.
Owner:HENAN CANCER HOSPITAL

Digital human image copying method and system based on NeRF

The invention discloses a digital human image duplicating method and system based on NeRF, and relates to the technical field of digital human image duplicating, and the method comprises the steps: collecting a digital human image, and obtaining a training data set; expanding and generating a training sample; training a data reconstruction model by using the expanded digital human image data; and performing feature extraction on a digital human image duplicating result output by the auto-encoder, and inputting a feature vector after feature extraction into the classifier model to perform digital human image data screening. According to the method, the details and the trueness of the generated sample are improved, the expanded data are closer to the actually acquired image in the aspect of trueness, the adaptability and the stability when the digital human image data are screened are enhanced, the screening model is more adaptive when facing diversified duplicated data, the classification effect is improved, and the classification efficiency is improved. And finally, high quality and consistency of the duplicated data are ensured.
Owner:QUANZHOU JINTONG INFORMATION TECHNOLOGY CO LTD

Pulsar candidate body generation method and system based on improved DCGAN

According to the pulsar candidate body generation method and system based on the improved DCGAN, pulsar candidate body images are generated through the improved DCGAN, channel information of different scales is extracted, and then channel attention weights are generated through convolution processing of different scales; multi-scale features of output features are extracted, and then a space attention map is obtained based on different pooling operation processing; local details and global features of the image can be enhanced, so that the overall quality and detail performance of the generated image are remarkably improved.
Owner:SHANDONG COMP SCI CENTNAT SUPERCOMP CENT IN JINAN +1

Intelligent tobacco leaf baking method and system based on image recognition

The invention belongs to the technical field of image data processing, and particularly relates to an intelligent tobacco leaf curing method and system based on image recognition, and the method comprises the steps: obtaining a to-be-cured tobacco leaf feature map and a cured tobacco leaf feature map, and carrying out the imaging evaluation, thereby obtaining an imaging quality evaluation value; analyzing the imaging quality evaluation value, carrying out imaging adjustment, obtaining an imaging characteristic standard value, obtaining a to-be-flue-cured tobacco leaf quality parameter set based on the to-be-flue-cured tobacco leaf characteristic graph, calculating the to-be-flue-cured tobacco leaf quality evaluation value, carrying out to-be-flue-cured tobacco leaf pre-screening based on the to-be-flue-cured tobacco leaf quality evaluation value, and matching a tobacco leaf first-flue-curing mode, image segmentation is carried out on the characteristic graph of the flue-cured tobacco leaves to form a plurality of research and judgment units, quality monitoring values of the flue-cured tobacco leaves are obtained, a tobacco leaf redrying judgment result is obtained by comprehensively analyzing a tobacco leaf first-drying mode and the quality monitoring values of the flue-cured tobacco leaves, and the tobacco leaf redrying judgment result is matched with a tobacco leaf redrying mode. According to the method, accurate image recognition and classification in the tobacco leaf curing process are realized, and the applicability and scientificity of curing mode matching are improved.
Owner:TOBACCO RESEARCH INSTITUTE OF CHINESE ACADEMY OF AGRICULTURAL SCIENCES (QINGZHOU TOBACCO RESEARCH INSTITUTE OF CHINA NATIONAL TOBACCO COMPANY)

Infrared image generation method based on multi-scale perception

The invention relates to an infrared image generation method based on multi-scale perception. Comprising the steps that image data are acquired and preprocessed, and the image data comprise a visible light image and an infrared image; constructing an infrared image generator and an infrared image discriminator; constructing a visible light image generator and a visible light image discriminator; a deep learning method is utilized to train an infrared image generator, an infrared image discriminator, a light image generator and a visible light image discriminator based on acquired image data. And determining the trained infrared image generator as an infrared image generation model. The method can generate a high-quality infrared image from a visible light image, and can be widely applied to the fields of night vision monitoring, remote sensing detection and the like.
Owner:SHENYANG INST OF AUTOMATION - CHINESE ACAD OF SCI

Power operation point cloud map construction method and system based on inspection robot

The invention relates to the technical field of video survey, in particular to an electric power operation point cloud map construction method and system based on an inspection robot. The method comprises the following steps: acquiring multi-source sensor data of a power operation environment; carrying out synchronous calibration on the obtained multi-source sensor data, and checking the time consistency of the multi-source sensor data; carrying out point cloud registration by using an improved point cloud registration algorithm on the premise of passing the time consistency verification to obtain an optimized weight matrix; performing pose estimation on the inspection robot based on the pose data from the inertial measurement unit, and calculating the pose data of the inspection robot; fusing the optimized weight matrix and the pose data of the inspection robot, and carrying out filtering updating on the fused point cloud matrix by adopting a filtering algorithm; and constructing a power operation point cloud map of the power operation environment based on the filtered and updated point cloud matrix. Accurate three-dimensional visualization support can be provided for electric power operation, and the operation efficiency and safety are improved.
Owner:STATE GRID JIANGSU ELECTRIC POWER CO LTD CHANGZHOU BRANCH

Capsule endoscope image high-definition display method

InactiveCN119110141ALighten processing tasksReduce potential impactImaging qualityColor correction
The invention discloses a capsule endoscope image high-definition display method. The method comprises the following steps: S1, performing JPEG compression on an original image P1 shot by a capsule endoscope to obtain an image P2; s2, the image P2 is transmitted to an external receiver through a wireless transmission module in the capsule endoscope to obtain an image P3, and no image processing operation is carried out from the image P2 to the image P3; and S3, in an external receiver, performing denoising, white balance, contrast improvement and color correction operation on the image P3 by using an OpenCV tool to obtain an image P4. And S4, using a local implicit image function super-division model based on deep learning to perform super-division of any magnification on the image P4 to obtain an image P5 finally displayed in the upper computer. According to the invention, image processing of the capsule endoscope is transferred to the external receiving device, so that a series of defects caused by image processing in the capsule endoscope are effectively avoided; moreover, an image with any resolution can be generated, image details can be reconstructed more finely, and the image quality is improved.
Owner:WUXI FUSHENG SMART MEDICAL TECH CO LTD