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

517results about How to "Easy to keep" patented technology

Adaptive underwater image enhancement algorithm

The invention relates to the technical field of image processing, and discloses an adaptive underwater image enhancement algorithm, and the algorithm specifically comprises the following steps: S10, constructing a network architecture: designing a conditional diffusion model as a main frame of image enhancement, the model comprises a forward noise adding process and a reverse denoising process, and in the denoising process, the forward noise adding process and the reverse denoising process are carried out; a denoising network composed of a Transform block and an improved residual state space block RSSB is introduced for encoding and refining image features, the RSSB is combined with a discrete state space equation of Mamba, a long dependency relationship in an image is modeled in a recursive form, channel redundancy is avoided, a visual state space module VSSM is introduced, important pixel information in the image is captured in a self-adaptive mode, and the image quality is improved. And the modeling capability of spatial details is enhanced. According to the adaptive underwater image enhancement algorithm, the influence of underwater environment complexity on image quality is overcome, and the image contrast, the color rendition degree and the detail retention capability are improved.
Owner:HUNAN NORMAL UNIVERSITY

Remote sensing target detection method based on convolutional neural network

The invention discloses a remote sensing target detection method based on a convolutional neural network, and the method comprises the steps: 1, constructing a full-scale enhanced alignment detection network model OSEADet comprising Backbone, Neck and Head, the Neck is an enhanced feature pyramid network EFPN, adding a P2 detection layer on the PAFPN network, enabling a feature map P2 to pass through SPDConv, outputting a feature map rich in small target information, fusing the feature map with a feature map P3, and obtaining a feature map P3; carrying out feature integration by using a CSP thought and a CSPF module; the Head comprises three dynamic task alignment sharing detection heads DSTA, the DSTA comprises a bounding box regression branch and a classification branch which are respectively used for predicting a target position and a target category, the tasks of the three DSTA are combined through group convolution, task interaction features are learned from a plurality of convolution layers, joint features are obtained, interaction among the tasks is increased, and the tasks are dynamically adjusted and aligned through a TaskDecomposition layer; 2, carrying out training and verification on the OSEADet; and 3, the trained OSEADet is used for predicting the position and category of the remote sensing target, so that the detection precision is high, and the remote sensing target can be detected in real time.
Owner:XIJING UNIV

Carbon dot-based high-water-content tough hydrogel and preparation method thereof

The invention discloses carbon-dot-based high-moisture-content tough hydrogel and a preparation method thereof, and belongs to the technical field of hydrogel. Specific carbon dots (hydroquinone carbon dots) are introduced, the hydroquinone carbon dots are used as nano-filler and are connected to a double-crosslinking network of sodium alginate / polyvinyl alcohol hydrogel through hydrogen bonds, effective crosslinking polymerization is promoted, and the hydrogel is prepared. The hydroquinone carbon dots are introduced, so that the conductivity can be improved, the mechanical property of the high-water-content hydrogel can be enhanced, the preparation process is simple, and the performance is excellent.
Owner:SOUTHWEST FORESTRY UNIVERSITY

Image super-resolution generation model construction method and image super-resolution generation method and system

The invention discloses a method for constructing an image super-resolution generation model. The method comprises the following steps: constructing a mixed data set containing various types of images; preprocessing the images in the mixed data set to obtain initial low-quality images and corresponding image text descriptions; determining a first initial learning framework, and generating a corresponding high-quality reference image according to the preprocessed initial low-quality image; training the first initial learning framework to obtain a first-stage repair model; determining a second initial learning framework, and generating a high-resolution image according to the random noise-added original image, a high-quality reference image generated by the first-stage repair model and image text description prediction; and training the second initial learning framework to obtain a target second-stage image super-resolution generation model. According to the method, a two-stage network framework model is designed and developed based on a potential diffusion model, rich prior of the diffusion model is utilized, and the generalization ability of the model is greatly improved.
Owner:NANJING UNIV OF POSTS & TELECOMM

Abnormal sound detection method and system of neural network based on multiple receptive fields

ActiveCN119091909ARealize motion captureFlexible adjustment of receptive field sizeSpeech analysisNeural learning methodsSound detectionData set
The invention provides an abnormal sound detection method and system based on a neural network of multiple receptive fields, and relates to the technical field of industrial intelligent diagnosis and quality control, and the method comprises the steps: collecting audio signal data of industrial equipment in various working states, and marking the collected audio signal data; constructing a data set based on the marked data and dividing the data set; the audio signals in the data set are preprocessed, and Mel spectrum features are extracted from the preprocessed audio signals; inputting the Mel spectrum features into an abnormal sound detection model constructed in advance for training to obtain a trained abnormal sound detection model; and inputting a to-be-detected audio signal sample into the trained abnormal sound detection model for prediction to obtain a prediction result, and classifying the industrial equipment according to the prediction result. According to the method, the multi-scale audio features can be dynamically captured, and the accuracy and robustness of abnormal sound detection are remarkably improved through rich feature representation capability.
Owner:ZHONGKE SHENGXUAN (SUZHOU) TECH CO LTD

Gas cylinder storage leakage point positioning detection method and system

The invention discloses a gas cylinder storage leakage point positioning detection method and system, and the method comprises the steps: building a gas cylinder leakage model under the conditions of no wind and wind, simulating a leakage process through a gas turbulence diffusion model, and introducing a sensor array to detect a gas leakage point; an adaptive filtering algorithm AF is introduced to carry out noise removal on data detected by the sensor array; extracting time difference information hidden in a gas sensor signal by using continuous wavelet transform (CWT) and a modulus maximum value method, judging the position of an optimal leakage point, and obtaining positioning information of a certain time point; a window function is introduced into a Kalman filtering algorithm, multiple sets of positioning information at different time points are obtained through processing of the improved Kalman filtering algorithm, interference of measurement invalid signals not in the range of the window function is filtered out, and the optimal positioning point is estimated. Compared with the prior art, the sensor array is introduced to realize leakage detection and positioning, a window function of a related time point is introduced into Kalman filtering, and the detection accuracy is effectively improved.
Owner:HUAIYIN INSTITUTE OF TECHNOLOGY

Image super-resolution reconstruction method based on compressed sensing

PendingCN119540054AHigh precisionOvercome the limitations of fixed sparsenessImage enhancementImage analysisPattern recognitionData set
The invention discloses an image super-resolution reconstruction method based on compressed sensing, and belongs to the field of image reconstruction. The problems that the updating speed of existing hardware cannot meet the requirement for signal precision, and on the premise that the hardware cost is not increased, an image with higher definition cannot be reconstructed are solved. The invention discloses an image super-resolution reconstruction method based on compressed sensing, and the method comprises the steps: designing a dictionary learning process, and obtaining high-frequency component parameters needed in image restoration; comprising the steps of constructing a training data set and performing dictionary training; wherein a corresponding low-resolution image is obtained by carrying out quality reduction and scale reduction calculation on an image so as to carry out a process of training a data set; constructing an over-complete dictionary pair by using a K-SVD algorithm, and training the dictionary; firstly, a reduced low-frequency component is obtained, and then the resolution is reconstructed. According to the method, the GPSR algorithm is utilized, the limitation of fixed sparseness of a traditional OMP algorithm is overcome, a more accurate sparse coefficient is obtained, and the precision of the reconstructed image is further improved by adopting a bilateral filtering mode.
Owner:HARBIN UNIV OF SCI & TECH

Liquid crystal polyimide aerogel with excellent shape memory performance and preparation method thereof

ActiveCN119529369AExcellent shape memory performanceThe preparation method is simple and environmentally friendlyImidePolymer science
The invention discloses liquid crystal polyimide aerogel with excellent shape memory performance and a preparation method of the liquid crystal polyimide aerogel, and belongs to the field of polymer aerogel materials. According to the liquid crystal polyimide aerogel with the excellent shape memory performance, the shape fixing rate is larger than or equal to 99%, and the shape recovery rate is larger than or equal to 97%; the heat conductivity at the room temperature is larger than or equal to 0.09 W m <-1 > K <-1 >; specifically, liquid crystal diamine, flexible diamine, a liquid crystal dianhydride monomer and a thermal curing agent are subjected to a reaction and solvent replacement, and polyamide acid dry filaments are obtained; and forming an aqueous solution from the dry polyamide acid filaments, injecting the aqueous solution into a silica gel mold, carrying out directional freezing casting on a customized freezing circulating device with a copper plate at the top, and carrying out freeze drying, thermal imidization and curing after molding, so as to obtain the liquid crystal polyimide aerogel with excellent shape memory performance.
Owner:JIANGNAN UNIV

Remote sensing target detection method and system based on multi-scale feature fusion, and medium

The invention belongs to the technical field of computer vision, and relates to a remote sensing target detection method and system based on multi-scale feature fusion and a medium, and the method comprises the steps: obtaining a remote sensing image, inputting the remote sensing image into a bottom-to-top feature extraction backbone network, and obtaining a multi-level feature map Ck; inputting the Ck into a multidirectional feature strengthening network to obtain a multi-scale feature map Yk; inputting the Yk into a feature extraction backbone network from top to bottom to obtain a multi-scale fusion feature map Pk; inputting the Pk into a target detection network for target detection to obtain a detection result; the multidirectional feature enhancement network performs all-directional frequency domain information enhancement on each level of feature map, performs feature enhancement of different scales on frequency domain enhancement features, and performs bidirectional space coding and feature weight optimization on H and W dimensions by using a coordinate attention mechanism. Multiple information of frequency domains and space domains of low dimensions and high latitudes of the feature layer is comprehensively concerned, the detection capability of small targets is effectively improved, and the capability of coping with complex backgrounds is enhanced.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

Sea surface floating small target detection method based on multi-modal image fusion

The invention relates to a sea surface floating small target detection method based on multi-modal image fusion. Comprising the following steps: acquiring a plurality of image pairs, preprocessing the acquired image pairs, and constructing an image sample data set by using the preprocessed image pairs; enhancing the contrast and detail information of the infrared polarization image by using an optimization algorithm; extracting low-pass sub-band and band-pass directional sub-band information of different modal images by using multi-scale transformation, determining fusion weights of the two parts according to the three-channel simplified pulse coupling neural network, and finally obtaining a final fusion result by using inverse transformation of the multi-scale transformation; and inputting the three-mode fusion image into a fixed detection network to obtain a target detection result. According to the method, the relative effect of the infrared polarization, the infrared intensity and the visible light image is effectively measured, the contrast and detail information of the infrared polarization image are effectively enhanced, the fusion feature is enhanced, and better fusion image quality and target detection performance are achieved.
Owner:FUZHOU UNIV

InSAR (Interferometric Synthetic Aperture Radar) watershed slope displacement monitoring method and device

The invention relates to the technical field of slope monitoring, in particular to an InSAR (Interferometric Synthetic Aperture Radar) watershed slope displacement monitoring method and device. According to the technical scheme, the method comprises the following steps: S1, acquiring multi-temporal SAR image data of a watershed area through a satellite; s2, preprocessing the SAR image data to obtain an initial interferogram; s3, de-noising processing is carried out on the interferogram; s4, eliminating terrain errors; s5, real displacement information is obtained; s6, carrying out phase calibration; s7, carrying out continuous space-time displacement monitoring on the side slope of the monitoring area; s8, constructing a displacement prediction model based on machine learning to predict the future displacement trend of the slope; s9, forming a displacement risk assessment report in combination with the slope displacement space-time distribution diagram and historical data; and S10, according to the result, sending early warning information to a related disaster prevention and reduction mechanism. According to the invention, the precision, real-time performance and effectiveness of slope displacement monitoring are comprehensively improved, and powerful technical support is provided for early warning and protection of geological disasters.
Owner:GUILIN UNIVERSITY OF TECHNOLOGY

Non-paired cross-modal medical image conversion method based on contour guide path regularization

The invention belongs to the technical field of medical images and deep learning, and particularly relates to a non-paired cross-modal medical image conversion method based on contour guide path regularization, which comprises the following specific steps: constructing a conversion path from a source domain to a target domain; a path regularization method is applied to smooth path transition and ensure the stability in the image conversion process; according to a path regularization technology based on contour guidance, the model can be focused on a key anatomical region, the structural consistency and accuracy of the generated CT image are further improved, and the training time can be remarkably shortened by reducing additional training steps. The invention aims to simplify the unsupervised training normal form and overcome the defects of the traditional GAN-based single-step generation method in the aspect of capturing the complex anatomical structure of the medical image. While the training time and the sampling time are considered, the problem that the quality of the generated medical image is limited is overcome.
Owner:GUANGDONG UNIV OF TECH

An adaptive image defogging and enhancement method and system in a mine dust and fog environment

The present invention belongs to the field of image defogging technology, and proposes an adaptive image defogging enhancement method and system in a dusty and foggy environment in a mine. First, an image processing model including an image grading preprocessing module, an adaptive processing module and an optimization module is established; the image processed by the image grading preprocessing module is input into the adaptive processing module; the image processed by the adaptive processing module is further optimized by the image optimization module; and the optimized image is output. In practical application, the present invention can accurately, efficiently and stably defog images in a dusty and foggy environment in a mine, significantly improving the image quality and the ability to identify safety hazards in coal mines.
Owner:CHINA UNIV OF MINING & TECH

Low-light image enhancement method, system and device and storage medium

The invention discloses a low-light image enhancement method, system and device, and a storage medium, relates to the technical field of computer vision, and solves the problems of non-uniform illumination distribution and difficult denoising in an image restoration process in an existing illumination mapping technology. Respectively obtaining a low-frequency image of the low-illumination image and a high-frequency image of the low-illumination image based on the low-illumination image; obtaining a low-frequency image of the enhanced low-illumination image based on the low-frequency image of the low-illumination image; obtaining a high-frequency image of the enhanced low-illumination image based on the high-frequency image of the low-illumination image; obtaining an enhanced image based on the low-frequency image of the enhanced low-illumination image and the high-frequency image of the low-illumination image; designing a loss function based on the enhanced image, and training a frequency decomposition network model, a low-frequency path illumination adjustment network model and a high-frequency path illumination adjustment network model; and performing the operation on the low-illumination image to obtain a final enhanced image.
Owner:JILIN UNIVERSITY

Self-adaptive contrast enhancement method and system for medical image

The invention discloses a self-adaptive contrast enhancement method and system for a medical image, and relates to the technical field of medical image processing, and the system comprises the following components: S1, image input and preprocessing, S2, quantum heuristic parameter search initialization, S3, a search process based on quantum state superposition and entanglement characteristics, S4, optimal parameter combination determination, and S5, optimal contrast enhancement. And enhancing the image contrast. According to the method, noise in the medical image is effectively removed through self-adaptive weighted median filtering and linear normalization preprocessing operation, the gray value range of the image is normalized, a high-quality input image is provided for subsequent contrast enhancement operation, meanwhile, a quantum heuristic parameter search algorithm is utilized, and the image quality is improved. According to the system, the optimal contrast enhancement parameter combination can be automatically searched and determined, the self-adaptive enhancement of the contrast of the medical image is realized, the definition and detail expressive force of the image are improved, and a doctor can more accurately diagnose diseases.
Owner:CHANGZHOU NO 2 PEOPLES HOSPITAL

PCB surface defect detection method and system based on deep learning

The invention belongs to the field of target detection, and provides a PCB surface defect detection algorithm and system based on deep learning, and the algorithm comprises an image collection module, a defect labeling module, a detection model training module, a defect detection module, and a defect data management module. A YOLOv8 target detection network is improved, and the target detection network is improved for the problems that manual visual inspection is low in efficiency, traditional machine detection is low in precision and high in false detection rate, and a deep learning model is difficult to deploy on an industrial production line. A lightweight convolution model is used for replacing an original convolution structure, so that the parameter calculation amount of the model and the deployment difficulty of the model in an industrial production line are reduced; a GAM attention mechanism is introduced into an original target detection network, and the PCB defect detection capability of the detection network is improved; and the positioning loss is optimized, and the detection precision of the target network is further improved by introducing an Inner-CIoU mode. Finally, through testing, the average precision mAP (at) 0.5 of the method for detecting the defects of the PCB is 99.2%.
Owner:HARBIN UNIV OF SCI & TECH

3D medical image registration method and device for contrast learning enhancement of double attention converters

The invention discloses a 3D medical image registration method and device for contrast learning enhancement of double attention converters. The registration method comprises the steps that a 3D medical image data set is acquired and preprocessed; then down-sampling the fixed image and the floating image to create an image pyramid; then, a registration network based on a converter is established, finally, a loss function value between a fixed image and a registered image is calculated by using a loss function, and a network model is optimized through back propagation, so that a trained neural network model is obtained. According to the invention, by combining a self-attention mechanism and a cross attention mechanism, local and global context information can be comprehensively captured, and the modeling capability of image details and an overall structure is improved. And measuring the distance of the corresponding feature points between the images by using a contrast loss function. Meanwhile, the similarity between the images is further calculated by adopting a normalized total gradient and a modal independent neighborhood descriptor loss function, so that the accuracy and precision of medical image registration are improved.
Owner:武汉工商学院

Transform-based marine organism population anomaly monitoring method and system

The invention relates to the technical field of marine monitoring, in particular to a Transform-based marine organism population anomaly monitoring method and system. The method comprises the following steps: acquiring marine organism population data of different monitoring points, including remote sensing satellite image data, time sequence measurement data of an underwater sensor network and high-resolution image data acquired by an unmanned aerial vehicle; preprocessing the acquired marine organism population data; performing feature screening on the marine organism data based on a gate control mechanism; carrying out anomaly detection on the screened features on the basis of a Transform model; and outputting an anomaly detection result and feeding back the anomaly detection result to the Transform model for retraining. According to the self-adaptive feature screening method based on the gate control mechanism, the features most relevant to the marine organism population change can be automatically recognized and screened out, and therefore interference of data redundancy and noise is reduced.
Owner:YANTAI UNIV

A sleep staging method based on deep learning

The present invention relates to the technical field of sleep diagnosis, and in particular to a sleep staging method based on deep learning, the method comprising: collecting a number of target physiological signals, preprocessing each target physiological signal to obtain a number of standard physiological signals; synchronously aligning and integrating each standard physiological signal in turn to obtain a standard sleep data set; determining the actual correlation between each physiological signal to be integrated in the standard sleep data set, and determining an actual signal combination strategy according to the actual correlation; generating an actual sleep data set according to the actual signal combination strategy, and extracting features from the actual sleep data set to generate actual feature parameters; inputting the actual feature parameters into a deep learning model to determine the actual sleep state; analyzing the actual sleep state, and performing corresponding sleep intervention according to the analysis results. The present invention realizes low-cost non-intrusive sleep monitoring, thereby improving people's physical health.
Owner:BEIJING SLIP TECHNOLOGY CO LTD

Clarified defatted coconut milk, preparation method thereof and application of clarified defatted coconut milk in tea beverage

The invention provides clarified defatted coconut milk, a preparation method thereof and application of the clarified defatted coconut milk in tea beverages. The preparation method of the degreased coconut milk comprises the following steps: S1, crushing and squeezing coconut meat to obtain coconut milk; s2, the coconut milk is filtered and centrifuged, a grease layer and an insoluble substance layer are removed, and degreased coconut milk is obtained; s3, sequentially adding pectinase and protease into the degreased coconut milk for enzymolysis, and filtering to obtain filtered pulp; and S4, adding a compound enzyme preparation into the filtered slurry, carrying out ultrasonic treatment, enzyme deactivation and filtration, centrifuging the filtrate, taking a clear juice layer, and filtering to obtain the clarified degreased coconut milk. According to the method, coconut pulp is taken from coconuts with proper fruit age and is juiced to prepare the coconut milk, and then proper filtration, clarification and centrifugal extraction conditions are combined, so that carbohydrates in the coconut milk can be better reserved, and substances such as protein and pectin in the coconut milk can be effectively removed or degraded, so that the degreased coconut milk becomes clear, and the quality of the coconut milk is improved. And turbidity or precipitation can be effectively avoided after the defatted coconut milk tea beverage is compounded.
Owner:COCONUT RES INST OF CHINESE ACAD OF TROPICAL AGRI SCI +2

Digital human generation method and device

PendingCN119131209Akeep identitiesReduce identity information differencesCharacter and pattern recognitionAnimationImaging processingRadiology
The invention discloses a digital human generation method and device, and belongs to the technical field of image processing. The digital human generation method comprises the following steps: processing a collected face image corresponding to a target object to obtain a face feature image corresponding to the target object and at least one mouth mask image; processing the acquired audio data corresponding to the target object to obtain mouth shape information corresponding to the target object; and processing the facial feature image, the at least one mouth mask image and the mouth shape information to obtain a digital portrait corresponding to the target object. According to the digital human generation method, the identity information difference between the generated image and the original image is reduced, the training model is prevented from being finely adjusted again on a new reference image, and the generalization ability and universality of the model are improved.
Owner:GRG BANKING EQUIPMENT CO LTD

Method for preparing NaP molecular sieve / activated carbon composite material from coal gangue

The invention discloses a method for preparing a NaP molecular sieve / activated carbon composite material from coal gangue, belongs to the technical field of coal-series solid waste recycling, and aims to solve the technical problem that the alkali resistance and wear resistance of a coal gangue-based molecular sieve need to be further improved in the prior art. The method comprises the following steps: adding a crystallized molecular sieve into a vapor deposition device, carrying out vapor deposition to obtain a carbonized molecular sieve, and carrying out hot isostatic pressing to obtain the carbon-coated molecular sieve. And carrying out vapor deposition, hot isostatic pressing and oxidation liquid treatment, and finally cross-linking through a hydrophilic modifier to obtain the wear-resistant, alkali-resistant and hydrophilic composite molecular sieve.
Owner:HUAINAN NORMAL UNIV +1

Optical fiber ribbon optical cable with forming part

The invention belongs to the technical field of optical cables, and discloses an optical fiber ribbon optical cable with forming parts, the optical fiber ribbon optical cable is provided with a cable core and an outer sheath located outside the cable core, the number of optical fiber ribbons in the cable core can be changed by changing the number of the forming parts and the size of ribbon containing channels formed between the adjacent forming parts, and the optical fiber ribbons are located in the ribbon containing channels. Each optical fiber ribbon in each ribbon accommodating channel is of an integrated structure, and each optical fiber ribbon in each ribbon accommodating channel is provided with at least three bends; the outer sheath wraps the cable core. The invention has the main beneficial technical effects of simple structure, easy manufacture, easy storage, more saved manufacturing space, higher space utilization rate, less non-fiber material consumption and low cost, and is more suitable for large-core optical fiber ribbons.
Owner:CHANGSHU YUTONG PHOTOELECTRIC TECH CO LTD

Three-dimensional scene representation model construction method based on scene adaptive local voxel multi-plane image

The invention belongs to the field of three-dimensional reconstruction, and particularly relates to a three-dimensional scene representation model construction method based on scene adaptive local voxel multi-plane images, which comprises the following steps: dividing a three-dimensional scene to be represented into voxel grids, and constructing a global multi-plane image MPI of each voxel; determining intersection points of different light rays and each voxel MPI under each known view angle, encoding the three-dimensional position coordinates of each intersection point corresponding to each light ray into a multi-dimensional position vector, and encoding the direction of each intersection point corresponding to each light ray into a multi-dimensional direction vector; a multi-dimensional position vector set and a multi-dimensional direction vector set corresponding to the light and a color label of a pixel corresponding to the light in the collected scene image form a training sample; constructing a three-dimensional scene representation network, wherein the three-dimensional scene representation network comprises a position multi-layer MLP network, a direction multi-layer MLP network and a color calculation unit of each voxel; and training by adopting the training sample to obtain a three-dimensional scene representation model. The detail features of the local area can be effectively captured.
Owner:HUAZHONG UNIV OF SCI & TECH

Human face video compression method and system based on key frame prediction

The invention relates to the technical field of video signal compression, and particularly discloses a face video compression method and system based on key frame prediction, and the method comprises the steps: face video quality analysis, face video key frame extraction, face video compression, and face compressed video decompression and reconstruction. The method accurately evaluates the quality of a face initial video, deeply knows details and changes of extracted frame contents by analyzing feature data of each extracted frame, provides an accurate basis for subsequent key frame extraction, ensures that the extracted key frame can accurately reflect the changes in the video, and improves the accuracy of key frame extraction. Different compression codes are adopted for key frames and feature frames in the face video, compression resources are utilized more efficiently, meanwhile, the quality and definition of the video are guaranteed, and whether the final quality of the compressed face video reaches the expectation or not is judged by analyzing the feature data of the second face video, so that the face video quality is improved. And the quality and the effect of face video compression are ensured to meet expected requirements.
Owner:HUNAN UNIV

Electrified dust removal device

The invention relates to the technical field of electrified dust removal of electrical equipment, in particular to an electrified dust removal device which comprises a telescopic rod, a cleaning motor, a mounting shell, a cleaning head, a mandrel and a cleaning brush, and the right side of the mounting shell and the upper portion of the telescopic rod are detachably and fixedly mounted together. The electric equipment cleaning device is reasonable and compact in structure, by arranging the sun gear and the planet gear, when the cleaning motor works, the multiple rotating shafts can be driven to rotate through the mandrel, so that the multiple cleaning brushes are driven to rotate, when the cleaning brushes rotate, bristles can be driven to rotate on the surface of the electric equipment, and when the brushes rotate, cleaning of dirt on the surface of the electric equipment can be accelerated; by arranging the telescopic rod, the height of the mounting shell can be adjusted, so that electrical equipment with different heights can be cleaned, the operation is more convenient, the right side of the mounting shell and the upper part of the telescopic rod are detachably and fixedly mounted together, the mounting shell and the telescopic rod are convenient to disassemble, and the practicability is high. And after disassembly, the mounting shell and the telescopic rod are convenient to transport and store.
Owner:ALTAY POWER SUPPLY CO OF STATE GRID XINJIANG ELECTRIC POWER CO

Two-stage three-level high-voltage battery charging and discharging device and control method

The invention relates to a two-stage three-level high-voltage battery charging and discharging device and a control method. The two-stage three-level high-voltage battery charging and discharging device comprises an ACDC circuit connected with a power grid and used for converting alternating current into direct current; the DCDC circuit comprises a first capacitor, a second capacitor, a third capacitor and a first flying capacitor control circuit; the anode of the first capacitor is connected to the first flying capacitor control circuit through a first connection point, and the cathode of the first capacitor is connected with the anode of the second capacitor; the negative electrode of the second capacitor is connected to the first flying capacitor control circuit through a third connection point; the positive electrode of the third capacitor is connected to the first flying capacitor control circuit through a fourth connection point, and the negative electrode of the third capacitor is connected to the first flying capacitor control circuit through a fifth connection point; the first flying capacitor control circuit is connected to the ACDC circuit through a second connection point; and the LC filter circuit is connected with the DCDC circuit. The system stability and reliability are enhanced, and the test requirements of the high-voltage battery are met.
Owner:SUZHOU XINNENG XIANFENG TESTING TECH CO LTD +2

Optical image segmentation method based on ResUnit + + model

The invention discloses an optical image segmentation method based on a ResUnit + + model, and the method comprises the following steps: image preprocessing: marking a cloud layer label and a snow layer label on image data, and dividing the image data into a three-channel data set which is formed by stacking a background mask, a cloud layer mask and a snow layer mask; feature extraction: dividing the preprocessed data set into a training set and a test set, and inputting the training set into a ResUnit + + model for feature extraction so as to identify feature information corresponding to a background, a cloud layer or a snow layer in the image data; target segmentation and classification: taking the extracted feature information as an index of a final segmentation and classification task of a cloud layer and a snow layer, and outputting a segmented image of the image by a ResUnit + + model based on the extracted feature information so as to divide each pixel point in the image into a background, the cloud layer or the snow layer; in the feature fusion process, high-level semantic information can be better reserved, and meanwhile, a more accurate segmentation result is obtained in combination with low-level detail information.
Owner:CHINA AERO GEOPHYSICAL SURVEY & REMOTE SENSING CENT FOR LAND & RESOURCES

Efficient artificial breeding and intermediate cultivation method for holothuria leucospilota

The invention discloses an efficient artificial breeding and intermediate cultivation method for holothuria leucospilota. The holothuria leucospilota with mature gonad is firstly dried in the shade and then transferred into seawater, sodium sulfite is added to make water oxygen deficit, and spawning induction is stimulated. According to the physiological habits of the holothuria leucospilota, hypoxia suffocation stimulation is carried out on the holothuria leucospilota by using sodium sulfite, and a very good spawning induction effect is achieved. Compared with an existing traditional tropical sea cucumber artificial spawning induction method, such as a running water stimulation method and a shade drying method, with the highest spawning induction rate of 13.3-18.0%, the effect of sea cucumber spawning induction is more obvious (the spawning induction rate reaches up to 75-81%) by means of the method. The sodium sulfite has no obvious toxicity to the parent sea cucumber and is low in price; a traditional spawning induction method must be operated at night, but the method has no requirement for spawning induction time, has the advantages of being high in operability, easy to popularize and capable of reducing parent parameter quantity and providing stable fertilized egg supply for large-scale sea cucumber breeding.
Owner:SOUTH CHINA AGRICULTURAL UNIVERSITY +1

Distortion image correction method based on fast multi-scale residual network

The invention discloses a distorted image correction method based on a fast multi-scale residual network. Relates to the field of image geometric distortion correction, in particular to the technical field of distortion image correction based on a fast multi-scale residual network. Comprising the following steps: S1, constructing a data set, and generating a paired data set of a distorted image and a standard image and a deformation displacement field file of a corresponding number of the distorted image by using a geometric distortion model; s2, a distorted image correction model based on a fast multi-scale residual network is constructed, the model is divided into a backbone network and a tributary network, the backbone network is used for recovering image features, and the tributary network is used for outputting image classification features; s3, training the distorted image correction model based on the fast multi-scale residual network by using the pairwise data set of the distorted image and the standard image and the deformation displacement field file; and S4, processing the to-be-corrected image by using the trained distortion image correction model based on the fast multi-scale residual network to obtain the corrected image and the corresponding classification thereof.
Owner:JILIN UNIVERSITY +1