970results about How to "Improve feature extraction" patented technology

The invention discloses a lightweight semantic segmentation network and method for power grid inspection, and the method comprises the steps: employing an encoder-decoder architecture, inputting original data into a lightweight MobileNet v2 backbone network for feature extraction at an encoder part, inputting an obtained feature map into a lightweight large-scale ASPP module, and carrying out the feature extraction; further performing multi-scale feature extraction on the high-dimensional feature map; and in the decoder part, high-dimensional features and low-dimensional features obtained by the backbone network are respectively input into a cross attention wavelet fusion module, after an obtained feature map is fused with the output of a decoder, the feature map is recovered to the size of original data through down-sampling operation by a full-dimensional dynamic convolution post-processing module, and a final prediction mask is obtained. According to the invention, by optimizing a coding and decoding structure, the receptive field of the model and the modeling capability of a narrow and long target are improved, the balance between the precision and the complexity can be effectively realized, and the method is suitable for an unmanned aerial vehicle inspection system with limited resources.

The invention discloses a karst collapse risk analysis method and system based on geological big data mining, and the method comprises the steps: collecting multi-source heterogeneous geological data, and carrying out the data fusion through employing a Tucker decomposition-based tensor decomposition algorithm; cleaning, complementing and standardizing the fused data; feature selection and construction: selecting features related to the karst collapse risk; carrying out feature extraction on the preprocessed data; based on the features extracted by the feature engineering, constructing a risk assessment model by using a graph convolutional neural network, and optimizing model performance through an ensemble learning method; carrying out risk assessment and prediction by utilizing the trained model, and carrying out visualization through a geographic information system; and real-time risk assessment is realized based on an edge computing technology, and the model is continuously optimized. According to the method and the system, complex heterogeneous data can be processed, the time change trend of geological parameters can be accurately captured, the prediction accuracy under complex geological conditions is high, and the timeliness of early warning is greatly improved.

The invention discloses a brain tumor MRI image segmentation model and method based on an improved Swin UNETR network, and belongs to the technical field of image segmentation. The brain tumor MRI image segmentation model is obtained by improving a Swin UNETR as a baseline model and comprises an encoder, a jump connection and a decoder, the decoder comprises a plurality of convolution fusion modules and an attention mechanism module, feature information of different levels is fused in the decoder by using the plurality of convolution fusion modules, and the attention mechanism module is used for identifying the feature information of the different levels of the feature information of the different levels of the feature information of the different levels of the feature information of the different levels of the feature information of the different levels of the convolution fusion modules. And the attention mechanism module is used to focus the feature information obtained by the plurality of convolutions. By adopting the scheme of the invention, the problem that the existing medical image segmentation technology based on deep learning is insufficient in brain tumor MRI image segmentation precision and is difficult to realize accurate segmentation especially in the face of different types of tumors and variability can be effectively solved, the calculation amount is small, and the image segmentation efficiency can be effectively improved.

The invention discloses a motor imagery electroencephalogram signal classification method based on multi-scale feature extraction and adaptive feature fusion, which is called EEG-MAFENet for short, and belongs to the field of brain-computer interfaces. The EEG-MAFENet is provided to promote the development of a motor imagery brain-computer interface and solve the problem of low classification accuracy in the prior art. Firstly, multi-scale time domain information and channel space information of electroencephalogram signals are extracted through a multi-scale feature extraction (MFE) module, and a primary spatial-temporal feature sequence is obtained. Next, the most valuable global and local feature information is adaptively fused through an adaptive feature fusion (AFF) module to obtain an advanced fusion time feature vector; and finally, completing a classification task by a classifier module formed based on a full connection layer. Experiments prove that the classification method provided by the invention can more effectively improve the feature extraction and classification performance of the motor imagery electroencephalogram signals.

The invention discloses a 6D pose estimation method and system based on an RGBD camera, and relates to the technical field of computer vision processing. Calibrating camera parameters based on the calibrated camera imaging; reconstructing a three-dimensional model and constructing a data set; segmenting the target object and calculating the actual size based on the three-dimensional model and the data set; establishing a 6D pose estimation network model based on the three-dimensional model and the data set; and performing pose estimation based on the actual size. The method has the advantages that the accuracy of 6D pose estimation is remarkably improved by combining the RGB image, the depth image and the point cloud data, object region segmentation is performed by using the SAM model, the actual size of the object is visually calculated by combining the depth information and the point cloud data, the precision of size estimation is improved, and the method is suitable for popularization and application. The introduction of depth information enables the pose estimation system to have better adaptability to objects with different sizes, shapes and surface features, and improves the robustness and accuracy of the system.

The invention discloses a rotary machine fault diagnosis method based on residual contraction convolution and an attention mechanism, and belongs to the field of rotary machine fault diagnosis. According to the method, firstly, a soft threshold filtering activation function is designed based on Meta-ACON, a soft threshold filtering module is developed around the activation function, and the module can adaptively learn a filtering threshold and a forward dynamic weighting coefficient, so that irrelevant features are effectively filtered, and the expression ability of effective features is enhanced. And then, in combination with multi-channel and multi-scale residual convolution, a multi-channel residual shrinkage convolution unit is constructed to realize multi-scale and multi-channel effective feature extraction. Besides, according to the method, a high-order statistical modeling idea is utilized, a GSoP attention mechanism is introduced behind a multi-channel residual shrinkage convolution unit, and the ability of extracting discriminative features is enhanced through second-order statistical information of a high-level channel feature map. Test results show that the method has good fault identification capability and generalization capability.

The invention provides a high-resolution remote sensing image semantic segmentation method based on multi-scale depth supervision, and the method comprises the following steps: 1, obtaining an ISPRS Potsdam data set and a Vaihingen data set; 2, building a multi-scale deep supervision network MDSNet, wherein the multi-scale deep supervision network MDSNet comprises a spatial branch network, a semantic branch network and a multi-scale feature aggregation part; each of the spatial branch, the semantic branch network and the multi-scale feature aggregation part is of a four-layer structure; step 3, inputting the training set into a multi-scale deep supervised network MDSNet for training, calculating a loss function and performing back propagation, updating network parameters, and obtaining a trained parameter model; and step 4, inputting the test set into the trained parameter model, and outputting a segmentation image of the target image. According to the method provided by the invention, the feature extraction capability is improved, and meanwhile, the semantic segmentation accuracy, the training efficiency and the stability are improved.

The invention provides a time sequence prediction method and system for a KAN enabling graph attention parallel loop network, and relates to the technical field of time sequence prediction, and the method comprises the steps: obtaining time sequence input data; inputting the time sequence input data into a feature extraction layer based on fast Fourier transform and discrete wavelet transform, and preprocessing the time sequence input data by the feature extraction layer to obtain spliced feature extraction data; inputting the spliced feature extraction data into a feature aggregation layer of graph attention based on KAN, and preprocessing the spliced feature extraction data by the feature aggregation layer to obtain feature aggregation data; and inputting the feature aggregation data into a time sequence prediction model of the parallel gating circulation unit network based on KAN and attention, and outputting time sequence prediction data by the time sequence prediction model. According to the method, the calculation complexity of time sequence prediction is reduced, and the time sequence prediction efficiency and accuracy are improved.

The invention relates to the technical field of light source control, and discloses a multispectral LED detection light source synchronous control method, device and equipment and a storage medium. The method comprises the following steps: carrying out double-interlaced spectral data sampling on a light source, and carrying out baseline extraction and compensation correction on preset three-dimensional spectral parameters; mapping of multi-dimensional mixed domain time sequence characteristics and decomposition of working characteristic tensors are carried out on the corrected calibration spectrum data, and modulation of recursive phases and group delay compensation calculation are carried out; performing crosstalk compensation of a multi-stage LED light source channel and suppression reconstruction of a multi-harmonic component on the light source phase compensation vector subjected to compensation calculation to generate light source correction control data; and performing light source driving signal synthesis on the light source correction control data to obtain a light source driving waveform sequence, and performing light source channel distribution and synchronous adjustment on the light source driving waveform sequence to generate a light source synchronous control result of the target multispectral LED detection light source. According to the invention, high-precision synchronous control of the multispectral LED detection light source is realized.

A strip steel surface defect area dimension measurement method relates to a defect dimension measurement method, and comprises the following steps: calibrating a camera coordinate system, and shooting a strip steel surface image by using a camera; identifying defects on the surface of the strip steel in the image by using a defect identification model, dividing a defective area in the image, measuring the size of the defective area, and calculating the actual size of the defective area according to the coordinate conversion relation; the defect identification model is based on YOLOv8s, a linear focusing attention mechanism is fused into a C2F module of a backbone network to form a C2F-FLA module, global features of an image are concerned, and more defect related features are captured; a bidirectional weighted feature pyramid network structure is adopted in the Neck network part, and a wavelet convolution module is utilized to respond to low-frequency information in the image; and performing down-sampling by adopting a Haar wavelet down-sampling module. The feature extraction capability is improved while the model parameter quantity, the large calculation amount and the calculation complexity are reduced, so that the detection result is more accurate.

The invention provides a target tracking method and system based on a grouped attention feature extraction network, and the method comprises the steps: obtaining a template and search features based on a template and a search image through employing the grouped attention feature extraction network; learning local areas of the template target image features and the search image features, and aggregating to obtain target information; splicing the template with the search features to obtain the input of the self-attention encoder in the asynchronous interaction stage; the fused features are combined with a prediction head for calculation, a classification regression result is obtained, and a target is predicted according to the classification regression result; training the target tracking model in combination with the large-scale data set to obtain a trained target tracking model, and performing accurate target tracking on the target by using the trained target tracking model; according to the multi-scale feature extraction network based on Transform, sequential local and global information interaction in the windows and between the windows is promoted, and the optimal balance between performance and speed is achieved.

The invention discloses an image denoising method based on double-branch feature fusion, and the method specifically comprises the steps: 1, carrying out the collection of an image, and carrying out the preprocessing of the collected image; 2, a denoising network with double-branch feature fusion is built, the network comprises a generation network and a discrimination network, and authenticity discrimination is carried out on a denoised image generated by the generation network and a real image through the discrimination network; step 3, adversarial training is carried out on the generative network and the discriminant network to obtain a trained dual-branch feature fusion denoising network; and step 4, performing denoising processing on the image by using the trained generative network. The method is remarkably superior to a traditional denoising method in denoising quality, model generalization ability and operation time, different types of noise can be effectively removed, and more image details and structure information are reserved.

The invention provides an electric arc additive manufacturing macroscopic surface defect detection method based on YOLOv8 improvement, and the method comprises the steps: 1, obtaining a picture of a weld bead surface defect, and making a weld bead defect data set; 2, dividing the weld bead surface defect image data set into a training set, a verification set and a test set according to a set proportion; step 3, constructing a neural network model based on the improved YOLOv8; 4, training the weld bead surface defect detection model by using the training set and the verification set to obtain an optimal weld bead surface defect detection model; and 5, inputting the test set into the optimal weld bead surface defect detection model, and outputting a test result. The weld bead macroscopic surface defects are automatically detected based on the improved YOLOv8 model, and the accuracy and precision of weld bead macroscopic surface defect detection are greatly improved.

The invention belongs to the technical field of image processing, and discloses a transformer fault detection method based on an improved YOLOv8 model. Aiming at the problems of inaccuracy in feature extraction, low precision, poor real-time detection capability and the like existing in fault detection of a traditional transformer fault detection method, the invention establishes a transformer fault detection model based on an improved YOLOV8 model architecture, a self-calibration convolution defogging module is introduced, a generative adversarial network (GAN) and a multi-source data fusion network structure are added, a BiFPN structure is used, an ASPP module is replaced by an ASPFCSPC module, and a YOLOV8 model architecture is improved, so that the feature extraction capability of a transformer fault detection model is enhanced, the calculation complexity of the model is reduced, and the fault detection efficiency of the transformer is improved. The transformer fault detection precision, accuracy and real-time detection capability are improved, and the stability and adaptability of the model are enhanced.

The invention relates to the field of medical care informatics, in particular to an ophthalmic disease prediction model training method based on artificial intelligence. Comprising the following steps which are executed in sequence: S1, acquiring a data set of ophthalmic diseases, and performing disease labeling on each image in the data set; s2, performing data expansion on the labeled data set to obtain a training data set of ophthalmic diseases; s3, performing feature extraction on the training data set of the ophthalmic disease by using a neural network algorithm based on dynamic weight redistribution as a feature extraction model to obtain extraction features of the ophthalmic disease; s4, using a self-encoding neural network based on feature refinement as a feature dimensionality reduction model to perform dimensionality reduction on the extracted features, and obtaining initial low-dimensional features of the ophthalmic disease after dimensionality reduction; s5, inputting the initial low-dimensional features of the ophthalmic diseases after dimension reduction into a classifier model for training, and obtaining a classification result; s6, classifying the data set by using the trained model; features of ophthalmic disease data are effectively captured.

The invention discloses a self-supervised skeleton behavior recognition method, system and device based on action semantic guidance and a medium, and the method comprises the steps: firstly, obtaining a to-be-trained skeleton video sample, and carrying out the preprocessing of the to-be-trained skeleton video sample; secondly, generating a semantic prototype of each type of action in the preprocessed human body skeleton data set through an action semantic generation module, namely generating a corresponding action text semantic feature; besides, the preprocessed human body skeleton data set is input into a feature enhancement module (FAM), skeleton data enhancement is achieved, and skeleton features corresponding to actions are generated; and finally, fusing action text semantic features and skeleton features of corresponding actions by using a multi-modal fusion module, assisting the model in learning more distinctive and representative action representations from the semantic level, improving the fuzzy action recognition precision, and through a comparative learning method, improving the fuzzy action recognition accuracy. A self-supervision mode aims to guarantee model recognition performance and reduce dependence on large-scale labeled data at the same time, and skeleton-based human body action classification is achieved; the system, the device and the medium are based on a self-supervised skeleton behavior recognition method guided by action semantics, human body actions are recognized and classified, and the system, the device and the medium have the advantages of being high in accuracy and good in robustness.

The invention discloses a rainy day target detection method based on paired image data generation and knowledge distillation, and belongs to the technical field of automatic driving target detection. The method comprises the following steps: acquiring a real image of a road scene and a semantic segmentation mask; inputting the semantic segmentation mask and the weather condition text information into a trained improved diffusion model to obtain a road scene generation image under a rainy day condition; introducing a CoT module and an ODConv module into the target detection model for optimization, and then improving the target detection model by adopting a dynamic weighted aggregation loss function to obtain an optimized target detection model; performing knowledge distillation to obtain a rainy day target detection enhancement model; and performing target detection by using the rainy day target detection enhancement model. The method provided by the invention is excellent in performance in a rainy day scene, has good generalization ability, can provide higher-precision target detection in various complex scenes, and effectively improves the detection accuracy and stability.

The invention belongs to the technical field of medical image processing, and particularly relates to a double-encoder semi-supervised heart MRI double-ventricle segmentation method based on improved SAM. A teacher model and a student model which are used both comprise double encoders, prompt encoders and mask decoders, each double encoder comprises a ViT branch and a CNN-Mama branch which are parallel, parameter fine tuning of the ViT branch is realized through a position adapter and a feature adapter, the CNN-Mama branch mainly comprises a convolution fusion Mama block and a second-order dimension attention block, and the CNN-Mama branch is used for realizing fine tuning of parameters of the CNN-Mama block. The convolution fusion Mama block introduces an enhanced convolution block on the basis of a VSS block, and the second-order dimension attention block uses a covariance matrix to mine second-order statistical information of data; carrying out weighted summation on the output features of the ViT branch and the CNN-Mama branch to obtain the output features of the double encoders; the output features of the double-encoder and the prompt encoder are input into a mask decoder to be decoded, and a segmentation result is obtained; and performing joint training on the teacher model and the student model based on interpolation consistency regularization semi-supervision, and using the trained student model as a segmentation model for double-ventricle segmentation of heart MRI. The segmentation model has strong feature extraction capability, can effectively extract the edge information of the ventricles and the boundary information of the double-ventricle connection region, and improves the segmentation precision.

The invention belongs to the technical field of Internet of Vehicles and communication security, and relates to a vehicle intrusion detection system and method based on a Hybrid EfficentNet model, and the system is obtained by training the Hybrid EfficentNet model. Wherein the Hybrid OfficientNet model comprises an initial convolutional layer, a backbone network, a self-adaptive global pooling layer, a Flaten layer, a discarding layer, a hidden layer and a classification layer; the backbone network is composed of one Fusion MBConv module and 15 MBConv modules, and each MBConv module comprises an expansion layer, a depth separable convolution module, a channel and space attention mechanism and a projection layer; residual connection is used in the MBConv module; the system has excellent capability of detecting abnormal traffic, realizes high accuracy, high-efficiency processing capability and high response speed, meets the requirement of convenient operation in a vehicle environment, and effectively solves the technical problems that an existing intrusion detection model is complex and inconvenient and cannot be directly deployed in a resource-limited vehicle-mounted network environment.

The invention relates to the technical field of American ginseng quality evaluation, and discloses an American ginseng quality evaluation method based on multi-task deep learning, a computer readable storage medium and a computer program product. The method comprises the following steps: collecting near infrared spectrum data of each American ginseng sample; determining the ginsenoside content of each American ginseng sample to obtain label data of each American ginseng sample; according to the collected near infrared spectrum data and the obtained label data, constructing and training an American ginseng quality detection model; according to the constructed American ginseng quality detection model, producing area tracing and ginsenoside content prediction are carried out on American ginseng to be evaluated. According to the method, the improved multi-task deep neural network is constructed, so that the feature extraction capability is improved, nondestructive origin traceability and total ginsenoside content prediction of American ginseng can be realized at the same time, the detection accuracy and efficiency are improved, and the defects of a traditional method in the aspects of real-time performance and nondestructive detection are overcome.

The invention relates to the technical field of SAR (Synthetic Aperture Radar) image detection, in particular to a lightweight SAR image aircraft target detection method, which comprises the following steps of: S0, acquiring an SAR image; the method comprises the steps that S1, a Mosaic data enhancement mode is adopted in an input layer, and an SAR image is preprocessed; s2, inputting the picture subjected to data enhancement into the improved backbone network, and outputting a feature map; s3, improving the YOLOv8n model to obtain a lightweight detection model; the method specifically comprises the following steps: improving a neck network of a YOLOv8n model to obtain an improved neck network; a decoupling detection head of the YOLOv8n model is improved, and an improved decoupling detection head is obtained; pruning is carried out through a DepGragh model, and a deployable light-weight detection model is obtained; s4, training the lightweight detection model to obtain a final target detection model for deployment; s5, inputting the output feature map obtained in the step S2 into an improved neck network, and S6, inputting the output obtained in the step S5 into an improved decoupling detection head.

The invention discloses a super-pixel high-frequency enhanced lightweight image super-resolution reconstruction method, device and equipment and a medium. Performing shallow feature extraction based on a convolutional network on an input image, grouping adjacent pixels into superpixels through local clustering, and clustering the superpixels according to similarity to obtain a local area; an intra-pixel attention module and an inter-pixel attention module are introduced, and the two attention mechanisms cooperate with each other in local and global feature extraction. Meanwhile, high-frequency prior information is introduced by using a high-frequency enhancement branch to capture high-frequency details, and the quality of a reconstructed image is improved. According to the method, pixels with similar characteristics are aggregated to form superpixels, and local and global information interaction is realized through an in-pixel attention module and an inter-pixel attention module respectively. Meanwhile, high-frequency prior information is added into the network by using a high-frequency enhancement module, so that the feature extraction capability of the network is effectively improved. A large number of experiments on a standard data set verify that the method has a better reconstruction effect.

The invention discloses a short-term photovoltaic power generation power prediction method adaptive to weather change, which comprises the following steps of: firstly, acquiring meteorological data around a photovoltaic panel and power data of a photovoltaic module, preprocessing the data, and dividing a training set and a test set according to a proportion; according to the method, a Bi LSTM-Attention model based on ISSA optimization is constructed, photovoltaic power generation power prediction is realized by capturing time sequence characteristics and historical information of input data, meanwhile, a mean square error between a predicted value and a true value is calculated to continuously fit the predicted value, prediction model parameters can be updated, model self-learning is realized, and the prediction efficiency is improved. And the accuracy of the short-term photovoltaic power prediction value of the prediction model is further improved. According to the method, the short-term prediction precision of the photovoltaic power generation power under different weather conditions is improved, so that a photovoltaic grid-connected system is helped to respond to the control parameter design of the power grid stability requirement, and the support capability of photovoltaic power generation on the power grid is improved.

The invention relates to the technical field of composite material damage detection, and particularly discloses a composite material damage intelligent detection and modeling method based on ultrasonic point cloud and deep learning, and the method comprises the steps: S1, data collection, S2, ultrasonic data processing, S3, simulation model construction, S4, PVT-RCNN model processing, S5, RI CP processing, and S6, damage model generation. According to the method, an ultrasonic point cloud data acquisition platform and a data processing method are designed, a sufficient composite laminate ultrasonic point cloud data set with obvious features is constructed, a PVT-RCNN model which is based on PV-RCNN and is improved by Swi n Transformer is designed, the feature extraction capability of the model on point cloud is improved, three-dimensional space information of damage is intelligently obtained, and the accuracy of the damage is improved. An RI CP registration algorithm and a damage model generation method are designed, a finite element model with damage is directly generated according to damage point cloud, and the material performance can be obtained through calculation.

The invention discloses an insulator defect detection method based on feature extraction and cross-scale fusion, and belongs to the technical field of insulator defect detection methods. According to the method, a defect detection model is obtained through training to find defects existing in a detected insulator image, and the method comprises the following steps: S1, inputting an image data set and a corresponding label data set; s2, designing a lightweight fast multi-scale extraction module to complete feature extraction; s3, designing an efficient feature fusion module to complete feature fusion; s4, selecting image features from the output of the cross-weighted bidirectional feature pyramid network through intersection-to-parallel ratio perception query selection; and finally, obtaining a confidence coefficient score and a bounding box of the target through a solution terminal, and further finishing decoding. The method effectively solves the problems that an existing detection method is insufficient in feature extraction capacity in the feature extraction stage and large in model parameter in the feature fusion stage.

The invention provides a night infrared detection method and system based on an improved lightweight YOLOv8 model, and relates to the technical field of deep learning and target detection. The method comprises the following steps: firstly, acquiring an infrared image data set, preprocessing the infrared image data set, and dividing the preprocessed infrared image data set into a training set, a verification set and a test set; an improved lightweight YOLOv8 model is constructed; according to the improved lightweight YOLOv8 model, on the basis of a YOLOv8 model, a DWR module used for improving the multi-scale and feature extraction capability of an infrared image, a CGAFusion module used for improving the feature fusion capability of the image, and an LSDECDDetect module are added. Training the improved lightweight YOLOv8 model by using the training set, and verifying the effectiveness of the improved lightweight YOLOv8 model by using the verification set in the training process to obtain a trained improved lightweight YOLOv8 model; and inputting the test set into the trained improved lightweight YOLOv8 model to obtain an infrared detection result, thereby improving the vehicle detection accuracy under the condition of poor visibility at night.

The invention provides a sick juvenile fish identification method based on an improved YOLOv10 model, and the method comprises the steps: collecting a fish image in a water area as a to-be-detected image, and carrying out the preprocessing of the to-be-detected image, and obtaining an image data set; the method comprises the following steps: constructing a YOLOv10-UDFishNet model; the image data set is input into the YOLOv10-UFFishNet model, the YOLOv10-UFFishNet model is trained through the image data set, and the trained YOLOv10-UFFishNet model is obtained; and inputting a fish image in a water area to be detected into the trained YOLOv10-UFFishNet model to obtain a fish detection result and position information of a fish individual target. According to the method, the sick juvenile fish in the turbid underwater environment can be accurately and efficiently identified.

The invention belongs to the field of time sequence anomaly detection, and discloses a time sequence anomaly detection method and system based on frequency enhanced convolution Transform, and the method comprises the steps: carrying out the sequence-level preprocessing of an input sequence, including normalization and block generation; encoding the input sequence by adopting a multi-head self-attention mechanism; extracting interrelationships in the sequence by a full convolution module having a plurality of stacked 1D convolution layers of different kernel sizes; s2 and S3 are repeated, and multi-layer coding is carried out on the sequence; converting the encoded embedding into a frequency signal by using a Fourier transform method; the KL divergence and the absolute error are respectively used for calculating the consistency loss of the time domain and the frequency domain, calculating the point-by-point anomaly score, and identifying whether the point is abnormal or not. Wide experiments on five common data sets show that the method is superior to an existing method in the aspect of abnormal recognition.

The invention relates to the technical field of quality management, in particular to a real-time quality management method and system for a PVC pipe, and the method can be applied to quality management of the PVC pipe, especially real-time defect detection of a high-speed production line. The method comprises the following steps: firstly, acquiring a multi-modal data set of the PVC pipe in real time and preprocessing the multi-modal data set to obtain preprocessed data; secondly, constructing a deep learning model, extracting multi-scale features of the preprocessed data by using depth separable convolution and asymmetric convolution, and reducing parameters of the deep learning model; thirdly, fusing the extracted multi-scale features by using multi-modal fusion, and carrying out feature dimension reduction in combination with global mean pooling; then, training a deep learning model and evaluating the detection performance thereof; and finally, detecting and alarming the surface defects of the PVC pipe in real time. According to the method, the surface defects of the PVC pipe are detected based on the deep learning model, and the detection accuracy is ensured while the real-time performance of detection is improved.

The invention discloses an older puerpera risk assessment early warning method and system, and relates to the technical field of data processing, and the method comprises the steps: 1, collecting the physiological index data of a target puerpera at each time point; 2, performing time-frequency domain analysis on the health index vector of each time point by using wavelet transform to obtain a fused feature of each time point; step 3, taking the fused features of each time point as input, inputting the fused features into a bidirectional LSTM network, and outputting a forward hidden state and a backward hidden state corresponding to each time point by the bidirectional LSTM network; 4, taking the backward hidden state with the abnormal score exceeding a set threshold value as a forward abnormal hidden state; and step 5, if the final risk probability exceeds a risk judgment threshold value, giving risk early warning. The accuracy, the real-time performance and the dynamic performance of the health risk assessment of the older puerpera are remarkably improved.