13results about How to "Efficient discrimination" patented technology

The embodiment of the invention discloses a landslide displacement monitoring method, system and device, and a storage medium, and aims at minimizing the total mean square error, and can adaptively find an optimal weighting factor corresponding to each piece of landslide displacement data, so that the fused data value is optimal. According to the optimization method, the error difference between the landslide displacement data can be fully considered, so that a more accurate and reliable data fusion result is obtained, and therefore, the first weighting factor of each landslide displacement data is utilized to carry out weighted fusion on all the landslide displacement data to obtain the landslide displacement deformation data of the landslide displacement monitoring point; the landslide displacement deformation data can comprehensively and effectively judge the landslide deformation stage, and the landslide deformation stage is closer to the real condition of landslide deformation.

The invention relates to an optical detection image fixed star discrimination method based on dimension transformation, and the method comprises the following steps: obtaining an optical detection image at a moment k, and determining a three-dimensional pointing position of a target in a sensor coordinate system; based on the three-dimensional pointing position of the target in the sensor coordinate system, determining the three-dimensional pointing position of the target in the orbital coordinate system; based on the three-dimensional pointing position of the target in the orbital coordinate system, determining the three-dimensional pointing position of the target in the earth inertial coordinate system; repeating the steps to obtain the three-dimensional pointing position of the target in the earth inertial coordinate system at the sequence moments of k + 1, k + 2...; and judging whether the target is a fixed star or not based on the three-dimensional pointing positions of the target in the earth inertial coordinate system at the multiple moments. According to the method, the fixed star track can be effectively judged under the condition of no fixed star catalogue.

The present invention discloses an APT detection method that integrates cloud-network-end logs and threat knowledge. The cloud-network-end fusion logs of multiple platforms are collected and processed into a four-tuple sequence in process units and time order to construct an origin graph of the cloud-network-end. Attack entities and entity relationships are extracted from network threat intelligence to form an attack causal graph, which is used in cloud-network-end APT detection. The present invention collects cloud-network-end fusion logs of multiple platforms and extracts attack causal graphs from network threat intelligence. It does not rely on manual work, thereby avoiding the formulation of a large number of rules, helping security industry personnel to more accurately and efficiently identify cloud-network-end fusion APT attacks, thereby achieving effective detection of cloud-network-end fusion APTs.

The application discloses a detection method for a QUIC protocol denial of service attack (QDA) in a backbone network. The method can realize detection of a composite denial of service attack under the QUIC protocol. The detection method comprises an offline training stage and an online detection stage. In the offline training stage, according to the attack characteristics of the composite QDA, representative one-way traffic feature data is extracted, and corresponding feature groups are constructed. These features can effectively deal with the case that there is a large amount of one-way traffic in the backbone network node. Finally, a specific machine learning method is used for training to obtain an attack detection model. The application can be used for QDA attack behavior detection and early warning in a massive traffic scene such as a backbone network, and provides a basis for maintaining the security of the network.

This invention provides a method, system, and storage medium for discriminating surface electromyography (EMG) signals in chronic low back pain based on a Fourier analysis neural network. First, a dataset is acquired, containing several EMG signals and their corresponding ground truth labels. The EMG signals are preprocessed to obtain preprocessed EMG signals. These preprocessed EMG signals are then input into a constructed Fourier analysis-based neural network to obtain a predicted label for each EMG signal. A loss function is constructed based on the ground truth labels and predicted labels, and the Fourier analysis-based neural network is trained to obtain a trained Fourier analysis-based neural network. The EMG signal to be detected is then input into the trained Fourier analysis-based neural network to obtain the corresponding predicted label. This invention, by constructing and training a Fourier analysis-based neural network, aims to extract discriminative latent periodic features, achieving automated discrimination of abnormal patterns in EMG signals of chronic low back pain, thus achieving more robust, generalizable, and efficient discrimination.

The application discloses a measurement method of a brain load positive effect threshold based on a beta high-frequency band time specificity, and is characterized in that a plurality of test tasks of different brain load grades are set from low to high, and original electroencephalogram data of a person to be measured under different brain load grades are collected; the original electroencephalogram data are preprocessed, average band power index values of the person to be measured in a beta high-frequency band under different brain load grades are calculated according to the preprocessed electroencephalogram data; and the minimum time segment serial number of a statistical significant difference between double conditions of different brain load grades is calculated t i and a time segment serial number difference between adjacent brain load grades Adjt i , obtaining Adjt i a brain load grade corresponding to the minimum time i , i *is the brain load positive effect grade threshold of the person to be measured; and the method has the advantages that the beta high-frequency band time specificity change in electroencephalogram signal data is analyzed, and the brain load positive effect threshold of the person to be measured is objectively and accurately evaluated.

The invention discloses a method and a system for identifying construction procedures in a tunnel based on visual information, and relates to the technical field of tunnel construction, and the method comprises the following steps: obtaining a real-time monitoring video in a tunnel construction environment; extracting feature points and descriptors of the feature points in each frame of the video by using an SIFT feature detection algorithm, matching the feature points in adjacent frames by using a feature matching algorithm, determining matching points in each frame, and judging and identifying key frames in the monitoring video according to the number of the matching points of every two adjacent frames and the change of displacement; and inputting the key frame into a target detection model, extracting multi-scale features through a plurality of multi-scale attention modules sequentially arranged in a backbone network, performing up-down sampling and splicing fusion on the multi-scale features through a reinforced feature fusion network, and finally extracting and outputting target information through a network detection head. And identifying the current tunnel construction process based on the target information. According to the method, different tunnel construction procedures can be efficiently and accurately identified, and smooth tunnel construction is guaranteed.

A kind of defect detection method of magnetic material, original binary image is obtained by image acquisition and preprocessing;Extract the compactness, bite mark sharpness and four-corner asymmetry of original contour before closing operation, constitute Raw Mask Trident feature, retain the tiny defects covered by morphological operation;At the same time, 16-dimensional macro shape features are extracted by closing operation, combined with 10-dimensional microscopic texture features extracted by uniform pattern LBP, a multi-feature fusion system is constructed;The training of multi-feature fusion system is completed by using weighted XGBoost classifier, and the decision boundary is optimized by dynamic threshold search, and finally the defect is judged.The present application realizes anti-interference detection by the feature extraction method of Raw Mask Trident, balances the precision and recall rate by combining sample weighting and dynamic threshold, can realize high-precision, low-false alarm, interpretable automatic detection of edge collapse, bite mark, internal crack and color point, and realizes quality control in the production process of magnetic components.

The present application belongs to the technical field of road facility equipment layout in traffic engineering, more particularly, relates to a method for monitoring the traffic state of a basic road section of an expressway. The method comprises the following steps: firstly, randomly setting detectors at different intervals on the expressway to obtain original vehicle trajectory data; then, pre-processing the original data; and then, calculating the running speed, speed difference, headway, headway time and vehicle density of each vehicle at each time in the detection interval based on the original data. Based on the fuzzy C-means clustering traffic state clustering analysis, the interval traffic data collected in 1 min is clustered and divided into traffic states, and each category is given a label to provide a training data set for the state discrimination algorithm. The traffic state discrimination model is trained, the precision rate is selected as the machine learning evaluation index, and the traffic state discrimination model based on the random forest is selected. The determination of the detector layout interval based on the traffic state transition.

The invention discloses a mutual authentication and switching authentication method between a radio frequency source and equipment in backscatter communication. According to the method, the problems of mutual authentication between equipment and a radio frequency source in a backscatter communication system, switching authentication of long-distance mobile equipment and robustness in an authentication process are solved. The method effectively improves the safety of a backscatter communication system, is suitable for authentication of equipment and a radio frequency source in static and dynamic scenes, and particularly provides comprehensive safety guarantee for switching authentication requirements in a long-distance mobile scene. With the continuous development of the Internet of Things technology, the method provides powerful security support for various backscatter communication application scenes, and ensures that the communication process always keeps high security and reliability in various complex and dynamic environments.

The present invention relates to the field of pollution source tracking, and discloses a method, device and storage medium for tracking groundwater pollution sources based on machine learning. The method includes the following steps: obtaining groundwater historical data and environmental data; extracting environmental features from environmental data, and extracting pollutant features from historical data; configuring a fusion model for groundwater pollution tracking; the fusion model is obtained by fusing multiple machine learning models; obtaining the correlation between environmental features and pollutant features, and obtaining weight factors based on the correlation, and using the weight factors as additional features input to the fusion model; deploying the fusion model, and completing pollution source tracking using the fusion model. The present invention introduces the mutual fusion of machine learning algorithms and supervised learning algorithms, and substitutes the improved multi-dimensional empirical formula, realizes efficient identification and tracking of pollution sources through autonomous model learning, introduces edge computing, performs real-time data processing and analysis, reduces data transmission delays, and improves the real-time nature of identification.

The invention discloses an emotional disorder detection method based on a spatio-temporal dynamic dependency modeling graph enhanced network, which comprises the following steps of: performing segmentation processing on electroencephalogram signals, constructing a functional connection relationship between electroencephalogram channels through a phase locking value, and mapping the functional connection relationship into a graph structure to extract spatial connection characteristics between the channels; in combination with time sequence feature coding and graph convolution operation, fragment-level feature representation containing dynamic information in the channels and synchronization information between the channels is obtained; a dynamic dependency integration mechanism is introduced, and a local time sequence dependency relationship between adjacent electroencephalogram fragments is modeled; and performing global time sequence modeling on the multi-fragment features by using a sequence modeling structure based on an attention mechanism, extracting long-range dependency features and completing emotional disorder state discrimination. According to the method, the spatial connection characteristics and multi-scale time sequence dynamic information of the electroencephalogram signals can be modeled at the same time, the accuracy, stability and generalization ability of emotional disorder recognition are improved, and the method is suitable for application scenes such as auxiliary diagnosis of emotional disorders and mental health monitoring.

According to the scanning identification method and system of the graduate employment questionnaire, paper test paper is converted into image data by using an image scanning technology, the filled test paper is further analyzed by using a deep learning method of machine vision, the validity of the test paper is judged, the test paper which does not meet the standard is rapidly screened out, and the accuracy of the test paper is improved. A lot of time is saved for later survey and analysis, and the efficiency of questionnaire survey and analysis is improved.