99results about How to "Reduce computation" patented technology

The invention belongs to the technical field of oil and gas pipeline safety monitoring and unmanned aerial vehicle image analysis, and discloses a two-stage oil and gas pipeline occupation pressure monitoring method and system based on unmanned aerial vehicle space perception and deep learning fusion, and the method comprises the steps: obtaining image data and space data of a to-be-monitored region, combining the image data and the space data with a DEM, and obtaining a DEM image; realizing projection of geographic coordinates of the pipeline to image pixel coordinates to obtain a plane image of the pipeline; constructing and deploying a lightweight target detection network DRF-YOLOv8n, performing reasoning on a plane image of the pipeline based on the lightweight target detection network DRF-YOLOv8n, outputting a candidate bounding box, and performing screening and key frame judgment; a Promp-guide SAM pixel level segmentation module is used to realize pixel level segmentation of the key frame, and a candidate pixel mask is generated; and judging whether occupation occurs or not through the intersection of the pipeline projection line and the candidate pixel mask, and outputting alarm information with geographic coordinates.

The application relates to a website weak password detection method and system, a storage medium and an intelligent terminal, relates to the field of electronic information security technology, and comprises the following steps: acquiring login website information, user information and password information; determining website importance information; determining possible protection awareness strength information; determining theoretical dictionary level information; matching weak passwords contained in the dictionary corresponding to the theoretical dictionary level information and the password information; if the matching is successful, outputting weak password information; and if the matching is unsuccessful, outputting normal password information. The application improves the problem that if very basic passwords are still simulated to log in during the exhaustive process, unnecessary scanning is caused, and the efficiency is low. The application has the effects of determining the possibility of weak password use of a user by understanding the protection awareness of the user and the importance of the website to the user, reducing the operation strength of the system, and improving the weak password detection efficiency.

The application discloses an airborne radar downburst optimized detection method and belongs to the technical field of radar signal processing. An airborne Doppler radar is used to emit a radar waveform to a possible wind shear dangerous area in front of a route, and wind speed estimation in a transmission direction is obtained by analyzing radar echo IQ data. The application sets a pitch angle of radar scanning according to a take-off or landing stage of an airplane. After receiving the IQ signal, frequency domain correction is performed, data is preprocessed according to spectrum width information, and a wild value area is eliminated. After the preprocessing is completed, a frequency domain clutter suppression algorithm is used to suppress a ground clutter area near zero frequency, and then wind shear wind speed estimation is performed. The application proposes a novel vertical wind speed estimation model, establishes a connection between a downburst core area range and a vertical wind speed gradient, forms a detection and sensing capability for a three-dimensional area of the downburst, and proposes an airplane scanning strategy for optimizing the model.

This invention relates to a practical laser spot localization method for array detectors based on radial basis function (RBF) neural networks, belonging to the field of detector technology. It utilizes RBF neural networks to achieve high-precision localization of laser spots. The RBF neural network is designed and trained, using the training set as the input layer and the output layer as the predicted spot position. The parameters and number of neurons are continuously trained and adjusted. The test set is input into the trained neural network to predict the spot position, and the effectiveness of the RBF neural network is tested. This invention has fewer parameters, lower computational load, higher accuracy, and is applicable to n*n multi-module array detectors. Furthermore, the experimental acquisition process in this application uses a quadrant detector, making the operation relatively simple.

The invention provides a battery power consumption lightweight analysis device, a lightweight log generation device, a battery power consumption lightweight analysis method, a lightweight log generation method, electronic equipment and a computer program product. The battery power consumption lightweight analysis device comprises: a system configuration analysis module for analyzing a system configuration file of a target electronic device to obtain system attribute feature data; the battery statistical data acquisition and analysis module acquires a battery statistical data packet and analyzes the battery statistical data packet to determine battery characteristic data; the lightweight log generation module carries out packaging processing on the system attribute characteristic data and the battery characteristic data to generate a lightweight log file; the log data reading and analyzing module reads the lightweight log file and analyzes the lightweight log file to extract attribute feature data and battery feature data; and the battery power consumption analysis module is used for analyzing the battery power consumption of the target electronic equipment according to the attribute characteristic data and the battery characteristic data to generate a power consumption analysis result.

The application discloses a handshake system, method and device of a three-dimensional convolutional neural network, computer equipment and a medium. The method comprises the following steps: obtaining to-be-transported data, and writing the to-be-transported data into a cache as a cache data block in a preset size; writing the size of the cache data block into a register group according to a preset storage mode to obtain data block record information; when a data operation request is received, obtaining required data information from the data operation request, wherein the required data information comprises three-dimensional coordinates; determining whether the cache data block contains data corresponding to the required data information based on the three-dimensional coordinates and the data block record information; and if the cache data block contains the data corresponding to the required data information, generating a response signal indicating that there is valid data in the cache data block. The handshake efficiency is improved by using the application.

The invention relates to a three-dimensional reconstruction method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps of obtaining a structured light stripe image; the structured light stripe image comprises a first view and a second view; performing phase decomposition on the structured light stripes in the first view and the second view to obtain a first phase diagram and a second phase diagram; based on a depth and parallax mapping relation of the structured light stripe image, determining a parallax search range corresponding to a first target pixel point in the first phase image in the second phase image, and performing matching according to the parallax search range to obtain a second target pixel point in the second phase image; disparity maps of the first phase diagram and the second phase diagram are obtained according to the first target pixel point and the second target pixel point; and performing three-dimensional reconstruction on the structured light stripe image based on the disparity map to obtain point cloud data. By adopting the method, the three-dimensional reconstruction efficiency can be improved.

The invention relates to a DPU-based fog cloud collaborative Internet of Things intrusion detection system and method. The system comprises an edge node used for receiving target data sent by Internet of Things equipment, encoding the target data through a pre-trained stacked auto-encoder to obtain a plurality of encoding features, determining a prediction capability index of each encoding feature, and determining a target encoding feature according to each prediction capability index; the cloud server is used for receiving the target coding features sent by the edge node, inputting the target coding features into a pre-trained Internet of Things intrusion identification model, extracting target global features, target local features and target time features associated with the target coding features and Internet of Things intrusion through different branch networks in the Internet of Things intrusion identification model, and sending the extracted target global features, target local features and target time features to the edge node; and outputting an Internet of Things intrusion detection result according to the target global feature, the target local feature and the target time feature. By adopting the system, Internet of Things intrusion detection with low resource occupation, high precision and high efficiency can be realized.

The application discloses a kind of SF6 circuit breaker fault prediction method and system considering the effect of breaking arc, collect breaking arc data and SF6 gas decomposition product content after breaking, in the process of operation and after fault, establish circuit breaker historical breaking data set, build cumulative arc energy-SF6 gas decomposition product content relationship model, obtain the SF6 gas decomposition product content corresponding to different breaking times, breaking current;Establish SF6 gas decomposition product content historical operation data set;Build SF6 gas decomposition product content prediction model based on LSTM, obtain the prediction result of SF6 gas decomposition product content;Establish SF6 gas decomposition product content historical fault data set, build SF6 circuit breaker fault diagnosis model;The prediction result of SF6 gas decomposition product is as the input of SF6 circuit breaker fault diagnosis model, obtains fault diagnosis result, realizes the fault prediction of SF6 circuit breaker.

The embodiment of the application relates to the radar technical field, and discloses a signal processing method, an electronic device and a storage medium. The signal processing method comprises the following steps: transmitting a detection signal in a Doppler multiplexing (DDM) mode through a plurality of transmission channels; processing a return signal to obtain distance-Doppler two-dimensional data; calculating a likelihood probability of a Doppler spectrum under different aliasing combination conditions of transmission channel orders of at least two targets according to the distance-Doppler two-dimensional data; and determining Doppler spectra of the targets according to an aliasing combination corresponding to a maximum value of the likelihood probability. At least, it is beneficial to correctly restore signals before Doppler spectrum aliasing in the presence of at least two targets.

The invention discloses a steel frame structure damage diagnosis method based on a one-dimensional depth separable convolutional neural network, and the method comprises the steps: constructing a neural network model for steel frame structure damage diagnosis through employing a one-dimensional depth separable convolutional kernel, carrying out the preprocessing of a to-be-diagnosed steel frame vibration response signal, dividing a training set and a test set, and carrying out the detection of the to-be-diagnosed steel frame vibration response signal. Inputting into a steel frame structure damage diagnosis model to determine a current damage type; according to the scheme, a neural network structure layer comprises a trunk part, three groups of one-dimensional depth separable convolution blocks, two independent depth separable convolution layers and three maximum pooling layers, the depth separable convolution blocks extract local features of input data, the maximum pooling layers carry out pooling operation on local areas of the input data, and the maximum pooling layers carry out pooling operation on the local areas of the input data. And finally inputting to a full connection layer to finish classification. According to the method, the damage type of the steel frame structure can be effectively identified, the model parameter quantity is small, the training cost is saved, the model accuracy is guaranteed through one-dimensional data training, and meanwhile important features can be prevented from being lost through time-frequency conversion.

A sound signal processing method performed by a sound signal processing device includes: an acquisition step of acquiring a first sound signal indicating a first sound and including first attribute information for specifying an attribute of the first sound signal, and a second sound signal indicating a second sound and including second attribute information for specifying an attribute of the second sound signal; the second sound signal includes second attribute information for determining an attribute of the second sound signal. A determination step for determining whether or not to merge the acquired first sound signal and the acquired second sound signal on the basis of an index corresponding to a first arrival direction in which the first sound reaches a listening position, which is the position of the listener, and a second arrival direction in which the second sound reaches the listening position; a merging step for generating a merged sound signal in which the first sound signal and the second sound signal are merged when it is determined to merge the first sound signal and the second sound signal; and a reproduction step of outputting an output signal based on the generated combined sound signal.

The application provides an improved sparse learning out-of-grid direction-of-arrival estimation method based on variational Bayesian. The method is characterized in that: the original data received by a hydrophone array is preprocessed, a real value transformation is used to convert a vectorized covariance matrix signal in a complex number field to a real number field, and the idea of variational sparse Bayesian learning and grid evolution is combined to make the grid evolve from an initial uniform grid to a non-uniform grid adaptively in an iteration process. The evolution process includes grid updating and grid fission. The evolved grid points are gradually close to the real source position through the alternately iterative grid updating process and grid fission process. Compared with the traditional compressed sensing method, the method has higher DOA estimation accuracy, reduces the operation complexity, optimizes the operation efficiency, improves the resolution capacity of the source, and has higher application value in actual engineering, especially in the case of few snapshots and low signal-to-noise ratio.

This invention discloses a method and system for constructing a vectorized feature set of an indoor scene point cloud model. The method includes acquiring point cloud data, visible light images from different directions, and depth maps for the indoor scene; performing target object segmentation on the point cloud data based on semantics, and image segmentation on the visible light images and depth maps based on edges; extracting 3D features from enclosed regions of the point cloud, visible light images, and depth maps based on the segmentation results; and then fusing them to generate a vectorized representation of the spatial 3D visual data of the indoor scene. This invention is simple to implement and can achieve a complete expression of multi-dimensional scene data based on point cloud, visible light, and depth map data. Due to the use of a graph structure for data vectorization and fusion, the computational load can be effectively reduced, enabling efficient and low-cost construction of a vectorized feature set for an indoor scene point cloud model.

The invention discloses a black vehicle-oriented real-time point cloud fusion detection method, system and device, and a medium. The method comprises the following steps: step 1, obtaining point cloud data through a laser radar; 2, scanning the point cloud data, marking the continuous point cloud blocks with the reflection intensity of 0 as unlicensed car identifiers, and storing the unlicensed car identifiers in point cloud attribute positions; step 3, performing dual-thread compensation, performing clustering operation on the point cloud data set with the unlicensed car identifier by using DBSCAN, and performing imaging processing on a clustering operation result by using Point Pill to obtain a three-dimensional image model; forcibly outputting an unlicensed vehicle detection frame image according to an image processing result or improving the confidence coefficient as an unlicensed vehicle compensation confidence coefficient, and preferentially adopting a model result; when the two conditions do not belong to the above two conditions, performing weighted fusion on the operation result; and step 4, fusing operation results in the step 3 into a complete three-dimensional frame image. Therefore, the problem of missing detection is solved, and the detection rate and detection efficiency are improved.

Embodiments of the present application disclose a cache control method, device, equipment and medium. The method comprises: obtaining a target access heat of cached data; controlling a storage address of the cached data to be added to a target array according to the target access heat; and updating the cached data in a cache area according to the storage address in the target array. The above scheme introduces the target access heat, updates the storage address in the target array, and then updates the cached data in the cache area in stages, thereby avoiding the influence of cache area avalanche on the database and reducing the access pressure of the database. Meanwhile, the above scheme reduces unnecessary occupation of cache resources, improves the use efficiency of cache resources, and is convenient to operate and has small computational complexity.

This invention provides a method and system for directional sound generation from a loudspeaker based on offline simulation and feedforward control, relating to the field of electroacoustic technology. The method includes: based on the feedforward compensation database, in the online implementation phase, performing frequency decomposition on the input audio signal using a digital signal processor and querying the database to obtain compensation parameters corresponding to each frequency component, thereby obtaining a real-time inverse compensation signal; using the real-time inverse compensation signal, synchronously driving the main loudspeaker and auxiliary loudspeaker, causing the compensated sound wave radiated by the auxiliary loudspeaker to destructively interfere with the sound wave behind the main loudspeaker, forming an acoustic quiet zone. This invention combines offline simulation to construct a precise database with online feedforward query for real-time compensation, ensuring accuracy and efficiency throughout the entire process from data foundation to acoustic quiet zone formation.

The present application relates to a tea bag sorting method based on an improved YOLO model and a parallel robot, which integrates an improved lightweight YOLOv11 model and an optimal suction decision algorithm, cooperates with a double-robot execution end, realizes the full-chain intelligent upgrading of tea bag sorting from "perception" to "decision" to "execution", uses the improved deep learning model, greatly reduces the calculation amount while maintaining high precision, ensures the real-time matching with the high-speed production line, adopts the cooperative work of the double robot based on the optimal suction decision, makes the system throughput increase by more than 70% compared with the single robot system, not only can quickly output the tea bag position, but also innovatively provides the "confidence" quantitative index and direction judgment, realizes the leap from simple "detection" to fine "quality evaluation and classification", and has high economic value.

A kind of identification detection method for coking furnace shutter valve based on inspection robot, specific steps are as follows: step one: increase the auxiliary image mark of image recognition on shutter valve;Step two: collect image data, carry out model training, establish coking furnace shutter valve dataset;Step three: collect shutter valve image data by inspection robot, judge whether state is normal, if yes, continue to check next shutter valve image, otherwise, send abnormal data to control center.The present application first adds two kinds of auxiliary marks to shutter valve, according to the detection requirement, corresponding installation is carried out on switch cover and telescopic rod, which is convenient for detection and recognition algorithm implementation.Then, track inspection robot is used for inspection and shooting, and the data shot is transmitted to detection and recognition module.The detection and recognition algorithm is optimized and improved on the basis of yolov3, the backbone network is replaced by MobileNet V2 to reduce network depth and improve speed.

The application discloses a recognition method, device, equipment and medium, the method comprises: obtaining an image to be recognized containing at least a target object; target detection is performed on the image to be recognized to obtain a first image; wherein the first image comprises an image containing the target object segmented from the image to be recognized; direction detection is performed on the first image to obtain direction information of the first image; the first image is corrected based on the direction information to obtain a second image; the second image is recognized to obtain a recognition result of the target object.

The invention discloses a three-way orthogonal coil receiving and signal merging method and a three-way orthogonal coil receiving and signal merging system for a rotary permanent magnet mechanical antenna, and belongs to the field of magnetic induction communication. The method comprises the following steps: performing analog conditioning and analog-to-digital conversion on received magnetic induction signals in each space direction, performing Hilbert transform on digital signals in each space direction, extracting instantaneous phase information of each channel, and obtaining complex signals in each space direction; and performing amplitude and phase coherence degree estimation on the complex signals in each spatial direction, determining a weight coefficient of each spatial direction based on the coherence degree between the complex signals in each spatial direction, and combining the complex signals in each spatial direction based on the weight coefficient of each spatial direction to obtain a combined signal. The signal amplitude of a receiving end signal is effectively improved, the signal-to-noise ratio is improved, and the direction robustness and the communication reliability of the system in a complex environment are enhanced.

The utility model relates to a gesture recognition light control system, which comprises a high-definition camera, an embedded artificial intelligence experiment box and a lamp panel suite, the high-definition camera and the lamp panel suite are respectively connected with the embedded artificial intelligence experiment box; the embedded artificial intelligence experiment box comprises a power supply, an MIPS processor, a neural network acceleration rod, a radiator, a mouse, a keyboard and a display, wherein the MIPS processor is electrically connected with the power supply, the neural network acceleration rod, the radiator, the mouse, the keyboard and the display. Gesture images are collected through a high-definition camera, gesture recognition is carried out through a trained neural network model in an embedded artificial intelligence experiment box, gestures can be accurately recognized, the lamp panel suite is controlled to be turned on, and real-time control is achieved. According to the utility model, image acquisition and image identification are completed through local equipment, calculation can be carried out without the help of a network or a cloud server, the privacy of a user can be better protected, and the intelligent use experience of the user is greatly improved.

The application discloses a method for suppressing range ambiguity clutter of an end-on array airborne early warning radar based on subarray synthesis. Firstly, echo signals are subarray synthesized along horizontal rows of an end-on array to form one column of spatial channel data, and then the column of spatial channel data is subarray synthesized along horizontal columns to form one row of spatial channel data, so as to form a cross-shaped horizontal spatial synthesis channel data. Then, the obtained horizontal spatial synthesis channel data is subjected to clutter suppression processing by using a STAP method, so as to effectively filter out the clutter of the end-on array airborne early warning radar. Finally, the data subjected to the STAP processing is subjected to constant false alarm rate (CFAR) detection, so as to realize detection of a moving target. The scheme is suitable for clutter suppression processing of the end-on array airborne early warning radar and has the characteristics of less training sample requirement and low operation complexity.

The application discloses a filter and a synchronization signal processing method. The filter comprises a first filter set and a second filter set; input ends of the first filter set and the second filter set are connected to a signal sampling component; output ends of the first filter set and the second filter set are connected to a filter output end; the signal sampling component is used for collecting a double-sampling-point communication signal and transmitting the communication signal to the first filter set and the second filter set; the first filter set is used for receiving a first communication signal and filtering the first communication signal to obtain a first filtered signal; the second filter set is used for receiving a second communication signal and filtering the second communication signal to obtain a second filtered signal; and the filter output end is used for coupling the first filtered signal and the second filtered signal to obtain a final filtered signal. According to the embodiment of the application, the downsampling is realized while the filtering precision is ensured, and the operation amount is reduced.

The application discloses a device identification method and device, apparatus and storage medium, and belongs to the technical field of computers. The method comprises: obtaining device feature information of a to-be-identified device; performing feature embedding processing on the device feature information to generate overall feature information of the to-be-identified device; determining cross feature information of the to-be-identified device based on the overall feature information; and determining that the to-be-identified device is an Internet service device in the case that the overall feature information and the cross feature information meet a target condition. In the technical scheme of the application, high-dimensional discrete device features are embedded into low-dimensional dense overall features, the device features are more efficiently represented and the amount of calculation is reduced under the premise of ensuring the completeness of the feature information, then the cross features between the features in the overall features are determined, the feature information is fully mined, and then the type of the to-be-identified device is determined according to the overall features and the cross features, so that the efficiency of identifying the Internet service device is improved, and the identification accuracy is also improved.

The application provides a robot free grabbing control method fusing target detection and a robot. The method comprises the following steps: when an instruction for grabbing a target object is received, a temporary parking point for the robot to approach the target object is generated; when it is identified that the target object is not within the grabbing range of the robot at the temporary parking point, the position of the robot is adjusted based on the position difference between the target object and the grabbing range of the robot, so that the target object is located within the grabbing range of the robot after the position adjustment; when the target object is within the grabbing range of the robot, at least one group of candidate grabbing postures is determined based on the point cloud data of the target object in a depth image; a candidate grabbing posture that can be reached by the mechanical claw of the robot and does not collide is selected as a target grabbing posture; and the robot is controlled to grab the target object in the target grabbing posture. In this way, the robot can accurately and quickly grab the target object.

A safety production task-driven system includes a remote processing unit, an edge processing unit, a personal task monitoring unit, and an alert unit. The remote processing unit receives data from the edge signal receiving unit, analyzes and processes the data, and also issues task information. The edge processing unit receives data from the personal task monitoring unit within its corresponding area and transmits it to the remote processing unit. The personal task monitoring unit receives tasks and monitors the status and location of personnel during task processing in real time. The alert unit receives instructions from the remote processing unit and issues alerts. This invention can monitor the heart rate, blood pressure, body temperature, and other physical conditions of personnel performing tasks in real time during production tasks, and can also locate personnel in real time, thereby ensuring both production efficiency and safety.

The application provides a kind of based on SM2's anti double authentication signature method and equipment.The method includes: step 1, system initialization, input security parameter 1 n , output public parameter set PP;Step 2, input public parameter set PP generates key;Step 3, input public parameter set PP, the message m=(alpha, p) to be signed and private key sk generates signature s;Step 4, input public parameter set PP, the message m=(alpha, p) to be verified, signature s and public key vk carries out signature verification;Wherein, alpha is the address part in the message m to be verified;P is the content load in the message m to be verified.The application adopts anti double authentication signature, and is implemented and tested, is conducive to promoting the generalization process of cryptographic algorithm, and signature scheme meets unforgeable security, and the operation amount involved in implementation is small, and efficiency is higher, has greater application potential.

The application discloses a kind of optical module health degree calculation method, on the basis of comprehensively considering the characteristics of digitalized relay protection device and existing data, based on the method of analog combination, the health degree calculation scheme of layered fusion based on " threshold determination based on + analog combination trend analysis as core" is proposed, according to the monitoring data of module read by DDM technology of protection device, the health degree of optical module is calculated at the level of relay protection device, then based on the modeling of the analog combination mode of Arrhenius model+linear degradation model, the influence of interference factors on optical module degradation is comprehensively considered, the calculation of optical module health degree is optimized, the application of the application considers the change of influence parameter and real-time and predictability, effectively reduces the amount of operation, is suitable for the application of relay protection device environment development environment.

The application discloses a kind of improved factor graph vehicle positioning method.The application includes the following steps: first, the motion information data collected by vehicle sensor is obtained;Then, according to the collected motion information data, a factor graph model is constructed;Then, according to the factor graph model, combined with the collected motion information data, the GNSS measurement value is detected by using the multi-condition analysis fault detection method, and the fault GNSS measurement information is obtained;Then, after the fault GNSS measurement information is isolated by deleting the corresponding factor node, the GNSS measurement information added to the factor graph model is updated;Finally, based on the current factor graph model, all navigation state quantities in the sliding window are optimized and solved, and the navigation information of the vehicle carrier is obtained, so as to realize the fusion positioning of the vehicle carrier.The application effectively suppresses the influence of abnormal GNSS measurement signal on navigation accuracy and effectively reduces the system false alarm rate.