28results about How to "Easy to classify" patented technology

The utility model relates to the technical field of storage bag, disclose a kind of portable storage bag, including storage bag body, the inner side edge position of storage bag body is provided with first zipper, the inner side wall of storage bag body is provided with external closure layer, the inner side wall of storage bag body is fixedly connected with connecting layer, the inner side wall of storage bag body is fixedly connected with split layer, the outer surface of storage bag body is fixedly connected with support air bag column, the upper surface of storage bag body is fixedly connected with elastic air bag, the side of elastic air bag is provided with air inlet, the other side of elastic air bag is provided with check valve, the one end of check valve is fixedly connected with exhaust valve, the utility model is set to external closure layer, second zipper, connecting layer, third zipper and split layer, and by the setting of support air bag column, elastic air bag, air inlet, check valve and exhaust valve, effectively solve the problem that the internal article of existing storage bag is easy to be chaotic, and the structure is insufficiently supported.

The embodiment of the application relates to an algorithm and a device of a Bayesian classifier of a non-uniform background, which are applied to an underwater active sonar system, the algorithm comprising: obtaining underwater data to be measured and auxiliary data through the active sonar system; the number K of the auxiliary data is greater than 0; modeling classification of the unknown covariance matrix structure into a binary hypothesis testing problem; hypotheses of the binary hypothesis testing problem comprise H i Wherein i=0, 1, H0 is a case that the unknown covariance matrix is a complex conjugate symmetric matrix; H1 is a case that the unknown covariance matrix is a real symmetric matrix; a Bayesian model is set, the Bayesian model comprising a complex inverse Wishart random matrix and a real inverse Wishart random matrix; a classifier for distinguishing the two hypotheses is obtained by using a minimum error probability criterion.

The utility model relates to food storage fresh -keeping device technical field, concretely is a kind of food processing storage fresh -keeping device, including fresh -keeping box, the upper portion of fresh-keeping box is provided with upper storage box, the lower portion of fresh-keeping box is provided with lower storage box, the inside sliding connection of upper storage box has N type frame, the inside of N type frame is provided with storage box, the inner chamber top of lower storage box is provided with partition, the bottom of partition is provided with corrugated partition pipe, this food processing storage fresh -keeping device, by the inside sliding connection of upper storage box N type frame, N type frame is pulled out the inside of upper storage box, the storage box embedded in the inside of N type frame can be expanded to both ends under the fixed effect of sliding block, it is convenient to place and take food, the partition set in the inside of storage box can reduce the damage due to accumulation in the storage process of food, improve the convenience and fresh -keeping effect of food storage.

The utility model discloses a vehicle -mounted refrigerator with bottle body support, include: refrigerator main part has the shell and the inner container and the upper cover, refrigeration component is used for refrigerating and heating for refrigerator, bottle body support subassembly, including support main part, support main part has the bearing plate and the positioning board, bearing plate with support main part is connected, the positioning board can overturn installation, has the positioning hole for limiting bottle body. The utility model discloses beneficial effect lies in, realizes vehicle -mounted refrigerator refrigeration and heating, and has bottle body support subassembly simultaneously, guarantees that bottled article will not be damaged in the vehicle driving process, satisfies a variety of use needs.

The utility model provides a kind of fusing body resistance full detection test equipment, it is related to fusing body resistance detection field, and it includes: base mechanism, the base mechanism includes support base, unqualified collection box and tester, the top of the base mechanism is provided with sorting mechanism, the sorting mechanism includes support frame, good product collection box, slide rail frame, finger cylinder and clamping claw, the top of the base mechanism is provided with detection mechanism, and the detection mechanism includes fixed frame, rotating motor, rotating disc, positioning fixture, product body to be measured, cylinder frame, drive cylinder, test probe and inductor.The utility model, the rotation of rotating disc makes that product body to be measured can rotate to the below of test probe, so that detection staff only needs to place product body to be measured to the top of positioning fixture, reduces the work pressure of staff, while it is convenient to classify product body to be measured, and it is convenient to sort after detection of product body to be measured.

The utility model discloses a kind of cleaning and wrong shelf detection equipment for library, it is related to library equipment maintenance technical field, comprising: equipment main body, mounting arm, dust suction member, dust suction fan, pipeline, detection device, mounting arm is located in equipment main body;Dust suction member is located in mounting arm and is communicated with dust suction fan by pipeline, dust suction fan is located in equipment main body;Detection device is equipped on equipment main body and / or mounting arm, detection device is configured as the label information of book can be identified and sends label information to host computer. Thus, by setting dust suction member and dust suction fan, dust on bookshelf and book surface can be cleaned, cleaning efficiency is high, cleaning effect is better, and, by setting detection device, books placed in wrong bookshelf can be detected while cleaning bookshelf and book surface dust, so that staff can classify books placed in wrong bookshelf, save manpower and time cost, improve the maintenance level of library.

The present application relates to a hyperspectral band selection method based on KL divergence weighted sparse self-representation, belonging to the technical field of hyperspectral image processing. The method obtains a hyperspectral image cube, converts it into a two-dimensional spectral matrix; uses PCA and superpixel segmentation algorithm to segment the image into multiple regions; constructs a neighborhood graph of the spectral matrix, calculates a similarity graph and performs a normalization operation; measures the information quantity difference between bands based on KL divergence, calculates the dissimilarity matrix of the spectral matrix; constructs a weighted sparse self-representation model that fuses spatial-spectral information, solves the coefficient matrix; arranges the L2 norm of the coefficient matrix row vector in descending order, selects the bands corresponding to the first M row numbers as the final band subset. The method fuses the spatial-spectral information of the hyperspectral image, reduces the similarity of the selected bands based on KL divergence, effectively improves the accuracy of band selection, and is beneficial to subsequent classification.

This invention belongs to the field of data processing and artificial intelligence technology, specifically a method and system for automatic classification of medical device patents based on deep learning. The method includes the following steps: acquiring textual data of medical device patents and performing preprocessing; introducing medical device classification standards as an external knowledge base, and calibrating the labels of the preprocessed text data to form a standard dataset; for categories with insufficient sample numbers in the standard dataset, generating synthetic data based on the patent text using a generative model; constructing a deep learning classification model that integrates global and local semantic features; training the deep learning classification model using the enhanced dataset, and inputting the medical device patent text to be classified into the trained model for classification prediction, outputting its corresponding medical device classification standard category. This invention solves the problems of mismatch between existing patent classification systems and industry regulatory standards, and low classification accuracy.

This utility model provides an industrial robot cable package structure, relating to the field of industrial robot technology. It includes a cable package body, on which adhesive hooks are fixedly installed, and adhesive strips that abut the hooks are also fixedly installed. A classification and fixing mechanism is provided on the cable package body. This utility model categorizes cables by attaching each cable to a fixing frame and marking two points of the cable with the same color at the same position on two fixing frames. The housing is moved inside the cable package body, bringing the threaded block into contact with the threaded sleeve. Rotating the housing secures it. A connecting frame moves the insert out of the movable slot, allowing a spring to use its own elasticity to move the movable block out of the slot, causing it to abut against the corresponding cable, thus supporting the cable and preventing tangling, facilitating subsequent maintenance.

The application provides an access device and an access system with the same. The access device comprises: a plurality of files; each file is provided with a fourth positioning part; a plurality of file boxes, each of which contains at least one file; each file box is provided with a third positioning part; an access cabinet, which has a plurality of access spaces for placing file boxes; a moving mechanism, at least part of which is movably arranged, and the moving mechanism comprises a first positioning part and a second positioning part; wherein the first positioning part and the third positioning part cooperate with each other and drive the file box to move through the moving mechanism, so as to take out the file box from the access space or place the file box in the access space; or the second positioning part and the fourth positioning part cooperate with each other and drive the corresponding file to move through the moving mechanism, so as to take out the file from the corresponding file box in the access space. The application solves the problem of inconvenient file management in the prior art.

The utility model relates to door and window processing technical field, proposed door and window equipment drilling and milling machine head, including head body, longitudinal adjusting mechanism and transverse adjusting mechanism, still include collecting mechanism, collecting mechanism includes collecting box, and collecting box is connected with elevating system, and the fixed connection of collecting box is limited in frame, and the collecting cylinder is provided on the limited frame, and the filter screen is provided on the collecting cylinder, and the fan is installed on the collecting cylinder, and the collecting box is hinged with the apron through a plurality of hinges, and the apron is connected with the collecting box through a plurality of hasp locks, and the top of apron is connected with the collecting pipe, and the connecting block is fixedly provided on the collecting pipe, and the electric push rod is installed on the head body, and the output end of electric push rod is connected with the connecting block, this device can effectively solve the debris and dust pollution when processing, and the suction of fan will be efficiently sucked, and the collecting pipe enters the collecting box, and the filter screen in the collecting cylinder realizes gas-solid separation, and the debris and dust are intercepted, and the classified processing is convenient, and the environmental pollution is reduced.

This invention discloses a long-tail image classification method based on Siamese networks with enhanced feature diversity. The method constructs an asymmetric Siamese network: one branch processes the original image, and the other branch processes a mixed sample composed of multiple images, with both branches sharing a backbone network. For the mixed sample branch, a feature diversity enhancement module is designed. Through a fusion process constrained by a lower triangular mask matrix, it guides the rich feature information of the head category to the tail category, thereby enhancing the feature representation of the tail category, while maintaining the feature map spatial structure unchanged throughout the process. The network training employs a hybrid loss function, combining classification loss, contrastive loss, and dynamic gradient balancing loss for collaborative optimization. After training, only the original image branch is used for inference, without adding additional computational burden. This invention effectively alleviates the classification imbalance problem under long-tail data distribution and significantly improves the recognition accuracy of the tail category.

The application discloses a coarse-to-fine double-flexible competition hybrid collaborative non-negative representation classification method, relates to the technical field of image classification, and comprises the following steps: obtaining a test sample and an original training sample set of n samples of C categories; performing normalization processing on the samples in the original training sample set; in a coarse representation stage, solving optimal representation coefficients of the test sample on the original training sample set by using double-competition collaborative representation; calculating the contribution weights of each category according to the optimal representation coefficients, and screening out a fine training subset highly related to the test sample; in a fine representation stage, adopting a double-flexible competition non-negative representation model on the fine training subset, and solving optimal representation coefficients of the test sample on the subset by using an alternating direction multiplier method; calculating representation residuals of each category of the test sample in the original training sample set; and classifying according to the residual of each category, so that the test sample is identified as the category corresponding to the minimum residual, thereby achieving better classification effectiveness and discriminability.

The application discloses a method, device and equipment for optimizing a power grid safety domain based on a denoising diffusion model. The method comprises: screening condition samples close to a safety boundary from power injection samples of a direct current power system in a closed-loop learning manner based on solvability of a power flow equation; inputting the condition samples into a denoising diffusion model, and converting random noise into candidate samples close to the safety boundary under the condition sample constraint through the denoising diffusion model; verifying the candidate samples based on solvability of the power flow equation; and updating the safety boundary of the power grid safety domain of the direct current power system based on the candidate samples that pass the verification. The method of the application realizes efficient characterization and fast convergence of a static voltage safety domain of a direct current system under the premise of ensuring accuracy, solves problems of a traditional method in terms of calculation complexity, boundary conservativeness, sample utilization and the like, and can provide efficient and reliable technical support for safety evaluation and operation boundary prediction of the direct current system.

The utility model relates to water quality detection technical field discloses a kind of online offline multipurpose water quality detection device, including detection cylinder, water inlet switching component, water outlet switching component and electrode;Detection cylinder is equipped with detection cavity and the water inlet and water outlet being communicated with detection cavity;Water inlet switching group includes online water inlet pipe, offline water inlet pipe, water inlet tee valve and water inlet switch valve, and online water inlet pipe and offline water inlet pipe are connected with water inlet by water inlet tee valve, and water inlet switch valve is located between water inlet tee valve and water inlet;Water outlet switching component includes online water outlet pipe, offline water outlet pipe, water outlet tee valve and water outlet switch valve, and online water outlet pipe and offline water outlet pipe are connected with water outlet by water outlet tee valve, and water outlet switch valve is located between water outlet tee valve and water outlet;Electrode is located on detection cylinder, and its measurement part is located in detection cavity.The utility model can solve the problem of how to switch online and offline water quality detection.

This invention relates to a hyperspectral band selection method based on spectral-spatial hypergraph regularized sparse self-representation. To address the insufficient utilization of band adjacency relationships in traditional sparse self-representation models, this method first integrates spectral similarity and spatial proximity to design a comprehensive index for evaluating band correlation. It explores and evaluates adjacency relationships between bands from multiple perspectives, constructing a hypergraph to generate a more complete inter-spectral local structure. Second, based on the aforementioned local band constraints, it supplements this with global band characteristics to jointly guide the optimization process of the sparse self-representation band selection model, obtaining the optimal low-dimensional representation of high-dimensional data and achieving band selection. This method overcomes the limitations of using a single similarity metric for decision-making and fully explores the multi-faceted adjacency relationships between bands, effectively reducing band redundancy, improving the quality of band subsets, and contributing to improved analytical accuracy in subsequent tasks.

The utility model provides a portable feed sampling device belongs to feed sampling equipment technical field, including the first sampling port of a plurality of settings of sampling outer tube, the second sampling port of a plurality of settings of sampling inner tube, first sampling port and second sampling port interval setting, the lower end detachable of sampling outer tube is provided with the wall breaking cone, the bottom fixed setting of sampling inner tube is provided with the limiting ring, the upper portion of limiting ring is provided with the limiting assembly, and the top of limiting assembly is provided with the connecting part, the utility model discloses through the rotation sampling inner tube to make second sampling port can with first sampling port uninterrupted coincidence, and then take the feed of different layer depth to sampling inner tube, again through the inner wall of sampling outer tube to the second sampling port closure, when pouring material makes first sampling port with second sampling port opposite, and then through the grip lever to draw out the material taking inner tube, and then can make the sample of the top of sampler first discharge, to discharge the feed of different layer depth in turn, it is convenient to classify the feed, and also the maintenance of the parts of sampler is convenient.

An automated scoring system for medial temporal lobe atrophy (MTA) based on YOLO deep learning includes: a coronal slice selection module, an improved YOLOv8 model construction module, an optimal coronal slice automatic selection module, and a medial temporal lobe atrophy detection and automatic scoring module. This invention provides an automated MTA scoring technique that balances accuracy, versatility, and ease of use. By using a lightweight deep learning YOLO object detection framework and combining it with an appropriate loss function, computational overhead is reduced while maintaining algorithm performance. Furthermore, a user-friendly front-end interface improves the efficiency and accuracy of clinical MTA scoring. This invention provides an efficient, accurate, and consistent automated MTA scoring system with significant clinical application value and promising prospects, and is expected to play an important role in the early diagnosis and monitoring of neurodegenerative diseases such as Alzheimer's disease.

The application provides a kind of based on improved Yolov5 water surface target detection method and device, it includes the following steps: 1) image acquisition, after data enhancement constructs water surface target set;2) using Gaussian filter to shallow feature map does smoothing operation and realizes image preprocessing;3) embedding multi-angle attention mechanism in Yolov5 network, so that the feature extracted is richer, improve the resolution decline, local information loss problem caused by ordinary method;4) feature scale conversion migration fusion Yolov5 network carries out feature processing;5) using Distance-IoU loss method loss calculation, finally, the three kinds of losses are weighted calculation, and then the network is optimized, and more accurate results are obtained.The method combines computer vision technology and detection technology and other fields, improves the efficiency and accuracy of water surface target detection, avoids the problems such as low detection accuracy, slow recognition speed and unsatisfactory multi-target detection caused by existing methods.

The application relates to a blanking and screening device and a detection equipment. The blanking and screening device comprises a conveying assembly, a film coating assembly, a marking assembly, a material distributing assembly and a material placing assembly. The conveying assembly is used for conveying workpieces. The film coating assembly is used for coating the top surface and the bottom surface of the workpieces. The film coating assembly comprises a first film coating component and a second film coating component. The first film coating component is located on the top side of the conveying assembly, and the second film coating component is located on the bottom side of the conveying assembly. The marking assembly comprises a first support, a driver and a marking piece. The first support is arranged above or below the conveying assembly. The driver is arranged on the first support. The driver is connected with the marking piece. The driver drives the longitudinal movement of the marking piece. The material distributing assembly is arranged on the conveying assembly. The material distributing assembly is used for picking up the workpieces. The material placing assembly comprises a first placing component and a second placing component. The conveying assembly can convey the workpieces to the first placing component. The material distributing assembly can pick up the workpieces conveyed by the conveying assembly and transfer the workpieces to the second placing component.

The present application provides a kind of grape growth adaptive bird and insect prevention bag based on auxetic support structure and its preparation method, including bag body, bag body includes elastic outer bag, and the auxetic support structure of setting in elastic outer bag, auxetic support structure includes tubular porous cell auxetic material, the bottom periphery of bag body is fixed with flexible air-permeable material sheet, and the top is formed into bundle mouth structure;With the growth of grape, auxetic support structure is subjected to longitudinal tension, and lateral expansion will occur, adapt to grape longitudinal and transverse growth, when grape ripens, bag expansion, approximate cluster shape, intuitive reaction grape mature cluster size, facilitate the classification of cluster size;And grape fruit and elastic outer bag between setting auxetic support structure as support, can prevent elastic outer bag and fruit adhere, ensure ventilation, set up air-permeable material in the bottom of bag, can ensure ventilation effect;Use elastic outer bag as the outermost layer of grape bagging, so that grape bagging has waterproof property, and has the function of preventing insects and birds.

The application discloses a multi-purpose stacking machine, which comprises a machine body, a moving mechanism arranged at the lower end of the machine body and capable of freely moving the machine body on a plane, a lifting frame arranged above the machine body and vertically moving along the machine body, two groups of opposite clamping arms arranged on the lifting frame, a driving element arranged on the lifting frame and capable of driving the clamping arms to extend or retract along the length direction, a clamping element arranged on the clamping arms and used for clamping goods, an RFID scanning unit arranged on the clamping arms and capable of scanning and identifying registered goods information when the clamping element clamps the goods, a supporting table arranged on the machine body and below the lifting frame and used for lifting the goods, an avoiding opening adapted to the supporting table and formed in the middle of the lifting frame, two object carrying plates arranged in the inside of the avoiding opening of the lifting frame and slidably connected with the lifting frame in the direction of approaching or moving away from each other, and a pushing element arranged on the lifting frame and used for driving the object carrying plates to move. The application has the effects of increasing the use function and improving the stacking stability of the goods.

The application discloses a traffic sign category determination method and device, a vehicle and an electronic device, and relates to the technical field of vehicles. The method comprises the following steps: acquiring a to-be-tested image, wherein the to-be-tested image is an image comprising a traffic sign; performing gray scale processing on the to-be-tested image to obtain a position image, wherein the position image is used for representing the position of the traffic sign in the to-be-tested image; and determining a recognition result according to the to-be-tested image and the position image, wherein the recognition result is used for representing the category of the traffic sign. The application solves the technical problem that, in the prior art, the category of a traffic sign is identified by using a traditional computer graphics method, a machine learning method and other deep learning algorithms, so that the neural network model is complex, the recognition rate is relatively low, and the classification effect is not ideal.

The utility model discloses a kind of cloth and linen cleaning intelligent spin-drying device, it is related to cloth and linen cleaning technical field, including rotary cylinder, sleeve is equipped in the outside of rotary cylinder, vertical cylinder is equipped in the inside of rotary cylinder, the bottom inner wall of rotary cylinder is equipped with annular frame body, annular frame body includes circular part, circular part is located in the inside of vertical cylinder, and the outer wall of circular part and the inner wall of vertical cylinder are mutually contacted, the wall of circular part is equipped with multiple filter holes, in this device, let rotary cylinder drive vertical cylinder to carry out synchronous rotation simultaneously in rotation process, realize the cloth and linen of two kinds of body shape to carry out quick classification spin-drying, improve cloth and linen sorting efficiency.

The utility model discloses a crystal oscillator leakproofness testing arrangement belongs to testing arrangement technical field, including fixed mounting panel, the top center of fixed mounting panel is provided with the air deflector, and the outside of air deflector is provided with fixed establishment, and the top all -around of fixed mounting panel is provided with detection seal cover, and the outside of fixed mounting panel is provided with the accommodation mechanism, and the bottom of fixed mounting panel is provided with the air outlet detection box, and the top of air outlet detection box is provided with the exhaust pipe, the utility model is provided with fixed establishment, and the side of shell is provided with a plurality of holes, can let the connecting column from these holes pass, when crystal oscillator removes to the air deflector top through the feeding mechanism, and crystal oscillator will touch trapezoidal clamp plate, and simultaneously through connecting column to push back the slide plate to the spring, and compress the spring, two trapezoidal clamp plate is constrained to spring at this moment, and the oscillator is fixed, and through trapezoidal clamp plate clamps crystal oscillator, prevents the influence detection result.

The utility model relates to lye recovery device technical field, and disclose the methanol to olefin waste lye recovery device with pretreatment structure, including base, the top fixed embedding of base has the motor, the output shaft outer wall of motor fixedly has the connecting disc, the outer wall of connecting disc is fixedly connected with two connecting plates symmetry, and the top of two connecting plates all is fixedly connected with mounting panel, the top of mounting panel penetrates respectively and is inserted with first storage jar and second storage jar. The utility model uses simple, the centrifugal force that second storage jar was driven to carry out circular motion by motor can produce when solid in lye is deposited in liquid bottom pretreatment, and the water vapor that the heated lye in second storage jar produces can be transported to first storage jar by fan, the lye in second storage jar can be completely evaporated and form solid, thereby let solid and liquid separate storage go to the subsequent classification processing of convenience.

The utility model relates to ball lens processing technical field, and disclose a kind of high saturation magnetic induction intensity magnetic force ball grinding round equipment, including support column, the inside fixedly connected with bottom plate of support column, the one end fixedly connected with top plate, frame and valve of support column.This utility model structure is reasonable, and the magnetic force ball to be detected is placed in lower disc, is placed in hole, and the flange plate is buckled to lower disc by rotating valve, first motor rotation drives rotating seat to make flange plate rotate and grind round the magnetic force ball of lower disc, after equipment polishes magnetic force ball into qualified size, magnetic force ball drops into collecting device through hole, magnetic force ball and the residue after polishing drop into first drawer, and magnetic force ball and residue are classified by filter screen, filter screen needs regular replacement by staff, prevent blockage, residue drops into second drawer by filter screen, and residue in second drawer is discharged by handle, prevent blockage equipment and cannot better classify magnetic force ball and residue.

This invention discloses a method for classifier parameter identification based on Extreme Learning Machine (ELM). Step 1: Divide the classification dataset into training and test datasets. Step 2: Construct an ELM model and use a modified whale optimization algorithm to obtain the optimal initialization parameters for the multi-novel ELM. Step 3: Train the ELM model online using the training dataset, evaluate the model using the structural risk loss function, identify and update the output weights, and complete the classification training of the obtained dataset to refine the ELM model parameters. Step 4: Input the test dataset to be classified into the multi-novel ELM model trained online in Step 3, and identify the category of the test dataset online. If new data is input, repeat Step 3 to classify the newly input classification dataset. This invention improves the classification accuracy of the ELM model.