40results about How to "Reduce false positives and false negatives" patented technology

The invention discloses an intelligent video analysis-based forest smoke and fire detection method and a special system thereof. The method includes the following steps: the real-time video images of the current detected scene, which are captured by monitoring cameras in a forest region, are transmitted to a video analysis server via a network; a smoke and fire detection module of the video analysis server detects and splits the moving foreground of the multiple frames of video images to generate a foreground list; according to colors, textures and motion features, the split moving foregrounds are preprocessed, and foreground areas without smoke and fire are filtered; on the basis of an intensive optical flow method, the continuous frames are dynamically analyzed, and dynamic motion features are extracted; and a support vector machine-based classifier classifies the features and outputs a detection result. The method and the special system thereof adopt the multi-video analysis technology to extract the unique features of smoke and fire, reduce false alarms, alarm failure and other circumstances, can analyze the video images of the day-night cameras in all weathers, can automatically detect smoke and naked fire in the daytime, and also can be used for detecting naked fire at night.

The invention provides a turnout abnormity early warning and fault diagnosis method. The method comprises the following steps of (1) based on the historical data of a turnout to be tested, selecting astandard motion curve; (2) using a clustering algorithm to segment the motion curve to be tested; if correct segmentation can not be performed, turning to a step (3); otherwise, turning to a step (4); (3) comparing a power curve in the motion curve to be tested with the power curve of all fault samples, searching the fault sample which is most similar to the power curve, and affirming that the motion curve to be tested has a fault corresponding to the found fault sample; (4) for each segment of the motion curve to be tested, extracting a first kind of characteristics; and (5) comparing with the standard motion curve and determining whether the motion curve to be tested is abnormal or normal. Compared with an existing segmentation method, the method has universality; simultaneously, compared with the prior art, the selected standard curve is scientific and reliability is high; in addition, adaptive diagnosis is realized, a demand to a person is less and efficiency is high.

The invention provides an intelligent video perimeter rail system and a control method thereof, belonging to the technical field of video monitoring. The intelligent video perimeter rail system comprises front-end subsystems, a middle transmission subsystem, and a back-end center monitoring subsystem. Each front-end subsystem consists of an automatic tracer and a plurality of fixed video alarms; the middle transmission subsystem consists of a wireless router, a switchboard and a transmission cable; the back-end center monitoring subsystem comprises a center server, a monitoring host and a client; the front-end subsystems are connected in a wired, wireless or combined wired and wireless transmission manner; and the front-end subsystems are connected with the back-end center monitoring subsystem in a wired, wireless or combined wired and wireless transmission manner. Intelligent analyzing modules are arranged in the video alarms and the automatic tracer in an embedded mode, thereby reducing a report negative rate and a false positive rate in a traditional solution. The intelligent video perimeter rail system records and alarms after an abnormal target enters a prevention area through intelligent analysis, so that the network bandwidth is effectively reduced; the cost is reduced; and the technical defects and problems in automatic alarming, tracing, recording triggered by alarming, real-time transmission long-distance monitoring can be solved.

The invention relates to a complex attack identification technology for a wireless intrusion detection system. The complex attack identification technology for a wireless intrusion detection system comprises the following steps: step 1, data acquisition; step 2, single step attack identification; step 3, complex attack identification; step 4, information display interface. The beneficial effect of the invention comprises: 1) the hybird architecture is applied to the identification of the single step attack, the identification accuracy of the single step attack is enhanced, and the false positives and false negatives are reduced; 2) the evaluation mechanism is used for identifying the complex attack, and the simple yes or no is not used for judging the single step attack behaviors, so that the attack sequence is generated, therefore, the distortion of the final result due to the information distortion among the modules can be effectively reduced; 3) an algorithm for identifying the complex attacks and predicting the final intention of the attackers is designed, which has better prediction results for complex attack intention with obvious characteristic behaviors.

The application discloses a method for detecting memory access overflow and a device for detecting memory access overflow. The method comprises the following steps of preprocessing source codes and analyzing a lexical structure, a grammatical structure and a semantic structure, and generating a control flow graph, a data dependence graph and an abstract syntax tree; establishing various detection models to different detection scenes, analyzing the control flow graph, the data dependence graph and the abstract syntax tree according to the various detection models, performing expression iteration, traversing state space, and outputting detection results; according to detection results, judging whether access overflow is present in source codes, and performing feedback output to detected memory access overflow. Therefore, the various detection models are defined by reviewing the control flow graph of the source codes, the data dependence graph and the abstract syntax tree, and expression iteration and state machine analysis are realized, the memory access overflow in the source codes can be fully excavated, the space and time cost for traversing the state space can be reduced, and underreporting and misinformation of the memory access overflow are reduced.

A process of rapid and highly accurate analysis of spectral data, includes both a linear scanning (LINSCAN) method and an advanced peak detection method for pattern recognition. One or both of the methods are used to support the detection and identification of chemical, biological, radiation, nuclear and explosive materials. The spectra of various targets can be analyzed by the two spectral analysis methods. These two methods can be combined for dual confirmation, greater accuracy, and to reduced false positives and false negatives, relative to what can be accomplished by either alone.

The invention relates to the wireless network field, and discloses a method and a device for recognizing a pseudo WiFi network by a wireless terminal. The method comprises the following steps: whether four handshake packets appear in the current network communication data packet is monitored; if yes, at least one following step or a step combination is executed: A, whether the handshake information in the four handshake packets is consistent with the handshake information used when the WiFi network is accessed last time, if not consistent, the current WiFi network is determined to have a risk of belonging to a pseudo WiFi network; and B, after the WiFi network is accessed, a first route arriving at a specified IP address is acquired, whether the first route is consistent with a second route arriving at the specified IP address when the WiFi network is accessed last time is judged, and if not consistent, the current WiFi network is determined to have a risk of belonging to a pseudo WiFi network. A more precise recognition method is given to the pseudo WiFi network, and error judgment and missed judgment can be reduced.

The invention discloses a malfunction study and determination optimization method based on a power distribution main station. The method includes the steps of (1) when a sharp drop occurs, distinguishing that the sharp drop is caused by an accident or manual operation; (2) recognizing the state of automation equipment to conduct trip point analysis; (3) additionally conducting filtering disposal on emergency power cut information; (4) verifying malfunctions of an adopted system; (5) making protective signals of non-trip-out equipment on a feeder line subjected to the trip point analysis for processing; (6) recognizing a distribution transformer of an internet port of a power plant and distribution transformers of multiple power sources, and ignoring the distribution transformers above during the trip point analysis; (7) recognizing the power cut causes of the distribution transformers and the electrification state of the distribution transformers; (8) conducting analysis and generatinga prewarning distribution transformer list. By adopting the malfunction study and determination optimization method, the accuracy of study and determination is improved, error reports and report omissions can be reduced, and the working efficiency of dispatching and repairing distribution networks is greatly improved.

The invention discloses a service data anomaly detection method based on time sequence classification, which comprises the following steps: s1, extracting offline service data, classifying the offlineservice data according to a time sequence, and generating a sample library comprising different types of time sequences; s2, associating different types of time sequences in the sample library with different time sequence anomaly detection algorithms; s3, acquiring online service data, and classifying the online service data according to the time sequence based on the classification of the time sequence in the sample library; and S4, carrying out anomaly detection on the classified online time series according to an association relationship between time series classification and a time seriesanomaly detection algorithm. The method automatically classifies and identifies different types of time sequences, automatically selects parameters or algorithms to perform time sequence anomaly detection, automatically identifies the types of the time sequences when processing large-scale time sequence anomaly detection, reduces false alarms and missing alarms of alarms, and effectively saves the labor cost.

The invention provides a method and device for detecting a living body in a vehicle based on a millimeter wave radar. The method comprises the steps of detecting an object in the vehicle through a sensing module on the millimeter wave radar, employing an IQ complex signal for demodulation, reducing the false alarm and missing alarm to the greatest extent, and rapidly recognizing a sleeping baby; tracking the object according to the detection information of the to-be-detected object, filtering out clutters such as mechanical vibration effectively, and distinguishing a living body and a non-living body effectively; if the object is a living body, using a microcontroller module on the millimeter wave radar for controlling a vehicle body alarm assembly to give an alarm, and sending remote alarm information to a mobile phone of a driver through a 5G communication module. Whether a living body exists or not is judged by detecting micro displacement and sign information, and the specific position of the living body is determined through DOA frequency spectrum estimation.

The invention discloses a railway unattended crossing automatic sound-light alarm system and method. The alarm system comprises a proximity sensor unit arranged on up and down railways to detect trainpassing information and send detection signals to a signal receiving unit, the signal receiving unit receiving the signals of the proximity sensor unit and sending the signals to a control unit afterrecognizing whether the signals are true or false, an alarm control unit controlling a crossing signal machine and a sound-light alarm device, a fault detection unit detecting the working conditionsof the proximity sensor unit, the signal receiving unit and / or the control unit in real time and sending abnormal signals to the control unit and / or uploading the abnormal signals to a far-end upper computer, a video monitoring unit conducting video monitoring, storing a crossing site and uploading monitoring signals, and a signal receiving and sending unit sending the abnormal signals of the fault detection unit and / or the video signals of the video monitoring unit, receiving remote control signals and sending the remote control signals to the control unit. The alarm method includes the stepsof train proximity detection, abnormity detection, video monitoring and alarm control.

The invention discloses a power equipment partial discharge identification method, system and equipment and a storage medium. The method comprises the following steps: acquiring sampling data of power equipment within a certain time by adopting a broadband high-speed sampling mode; the sampling data are converted into a PRPD spectrogram, and signal feature extraction and classification are carried out on the PRPD spectrogram; statistical features in the signal features extracted from each type of PRPD spectrogram are substituted into a pre-trained partial discharge identification model corresponding to the type, and whether partial discharge exists in the power equipment or not and the partial discharge type to which the power equipment belongs are obtained; the partial discharge identification model is obtained by training the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong by using a neural network. Broadband high-speed sampling is adopted to collect sampling data, complete partial discharge waveforms are obtained, signal feature extraction and classification are carried out on the PRPD spectrogram, and the pre-trained partial discharge identification model is adopted to carry out partial discharge identification, so that the equipment partial discharge monitoring precision is improved, and false alarms and missing alarms are reduced.

The invention discloses a heat source detection method based on an infrared thermal imager, which comprises a gradient division step and a heat source screening step, wherein the gradient division step includes the sub-steps of acquiring infrared raw bare data in the monitoring scene, solving the mean difference of coordinates of each pixel, establishing a mean difference two-dimensional array, and sequentially searching a pixel set located in each gradient on the basis of the mean difference from the center to the edge according to an infrared picture; and the heat source screening step includes the sub-steps of partitioning the pixel set belonging to the corresponding gradient, searching pixel blocks distributed according to the gradient of the infrared picture from the center outward from the first gradient, and screening out pixel blocks which are judged as fire sources / high temperature sources prone to starting a fire. The invention further discloses a heat source detection systembased on the infrared thermal imager and a storage medium. The heat source detection method and system based on the infrared thermal imager and the storage medium can identify a heat source more accurately, thereby reducing false alarms and missed alarms.

The invention relates to a cloud computing network monitoring system based on an end user environment, which comprises a host network, a protection module and a network monitoring server, a security authentication module is arranged between a router and terminal equipment, the router is connected with a proxy gateway module and an information storage module, the information storage module is respectively connected with the router and a cloud computing database, and the modules realize mutual data interaction through a cloud API. Compared with the prior art, the system has the advantages that the monitoring effect is perfect, the monitoring efficiency is high, the network environment of the terminal equipment is better, and the like.

The invention discloses an early warning judgment method and an early warning judgment device for bonding bleed-out, a single row of temperature measuring points are arranged on a crystallizer along the width direction of a casting section, and the method comprises the following steps: acquiring temperature values of the temperature measuring points on a time sequence to form a two-dimensional temperature matrix; carrying out mode processing on the temperature values of the two-dimensional temperature matrix to form a mode matrix, wherein the mode processing refers to converting the temperature value of the temperature rising starting point into a, converting the highest temperature into b and converting other temperature values into c for the temperature value of each temperature measuring point if the temperature is in a rising-to-falling mode; for the non-rising-to-falling mode, converting the temperature into a; and identifying data of the mode matrix to obtain an equivalent feature contour, judging whether a triangular early warning block is formed or not, and taking whether the triangular early warning block is expanded in the transverse direction and / or the longitudinal direction or not as a judgment basis for bonding or not. According to the invention, the defect that early warning depends on multiple rows of thermocouples is solved, the maintenance cost is reduced, andthe phenomena of missing report and false report caused by thermocouple abnormality are reduced.

The invention provides an integrated detection device for fire,and belongs to the technical field of fire detection. The integrated detection device comprises: a sampling unit,wherein the sampling unit is used for sucking air from a detected area as air to be detected; an analysis unit,wherein the analysis unit is used for detecting the air to be detected to generate a characteristic value; an on-line monitoring and alarming system,wherein the on-line monitoring and alarming system comprises a control unit and an alarming unit,wherein the control unit is used for controlling the sampling unitto draw air to be detected from the area; the alarming unit is used for starting to alarm; a control host,wherein the control host is connected with the analysis unit and is used for judging whether the area is on fire through the analysis unit,and sending an alarm instruction to the on-line monitoring and alarming system to enable the alarming unit to start when judging that the area is on fire.

The invention provides a memory leak detection method and device, and the method comprises the steps: traversing classes contained in a project, and recording classes possibly having memory leak risksand member variables having risks in the classes, so as to form a risk class diagram; traversing all functions contained in the project, and constructing a linked list between function calls according to a call relationship between the functions to form a function linked list; and detecting memory leakage existing in the project based on the risk class diagram and the function linked list. By using the method of the invention, missing report and false report in memory leak detection can be reduced, and the quality and efficiency of scanning and auditing memory leak are improved.

The invention discloses a video image color cast detection device and method, which are used for detecting histogram information entropy in an image HSV color space. If the statistical information entropy is greater than a certain threshold, a color cast image is outputted, and otherwise, a normal image is outputted. The device comprises a first detection module used for converting an image ofa RGB color channel into an HSV channel, grouping the converted HSV channel, calculating an H-S histogram, comparing the information entropy of an H-S histogram with a specified threshold, and preliminarily verifying whether the image is subjected to color cast or not. A second detection module is used for , comparing the proportion of the combination with the maximum number of pixels with a specified threshold value after preliminary verification; and / or comparing the ratio of the information entropy to the absolute value with a specified threshold, carrying out secondary verification on thepreliminarily verified color cast image, to acquire the absolute value according to an absolute value formula by the position of the combination with the largest number of recorded pixels and the position of the mean value of sub-image blocks after record segmentation in the H-S histogram.

The invention provides a real-time obstacle detection method based on vehicle-mounted equipment, and solves the technical problem that the accuracy of judging the position of an obstacle in an environment in a tunnel with a structure circulating and lacking characteristics is low. The invention provides a real-time obstacle detection method based on vehicle-mounted equipment. The method comprises the steps that laser point cloud data and camera image data of a track advancing area are acquired; wherein the camera image data is a multi-frame constraint; fusing the laser point cloud data and the camera image data to obtain fused point cloud data; drawing a point cloud map according to the fused point cloud data; marking a safety clearance on the point cloud map, wherein the safety clearance is the contour of a vehicle safety driving space in the track driving area; acquiring a first position of the vehicle, wherein the first position is a real-time position of the vehicle and a position coordinate corresponding to the attitude information in the point cloud map; marking the first position on the point cloud map to obtain a target map; and filtering the point cloud data outside the safety limit in the target map.

The invention relates to a complex attack identification technology for a wireless intrusion detection system. The complex attack identification technology for a wireless intrusion detection system comprises the following steps: step 1, data acquisition; step 2, single step attack identification; step 3, complex attack identification; step 4, information display interface. The beneficial effect of the invention comprises: 1) the hybird architecture is applied to the identification of the single step attack, the identification accuracy of the single step attack is enhanced, and the false positives and false negatives are reduced; 2) the evaluation mechanism is used for identifying the complex attack, and the simple yes or no is not used for judging the single step attack behaviors, so that the attack sequence is generated, therefore, the distortion of the final result due to the information distortion among the modules can be effectively reduced; 3) an algorithm for identifying the complex attacks and predicting the final intention of the attackers is designed, which has better prediction results for complex attack intention with obvious characteristic behaviors.

The invention discloses an intelligent pipeline capable of being monitoring remotely. The intelligent pipeline capable of being monitoring remotely comprises an intelligent pipeline carrier device control terminal, a handheld type data acquisition and communication end and an emergency platform monitoring end, wherein the handheld type data acquisition and communication end is used for data acquiring of the intelligent pipeline carrier device control terminal, and the emergency platform monitoring end is used for data processing, storing and distributing. When an anomaly or accident occurs, the monitoring end is acted as an emergency platform, charges for the intelligent pipeline carrier device control terminal and the handheld type data acquisition and communication ends, and behavioral intervention is carried out on the state of the intelligent pipeline carrier device control terminal and a holder of the handheld type data acquisition and communication end. According to the intelligent pipeline capable of being monitoring remotely, the effectiveness of a judgment is improved, the situations of misstatement and underreporting are reduced, so that the safety condition of the intelligent pipeline carrier device control terminal is effectively monitored.

A method and device for detecting chromosomal variation, the device comprises a sample-to-be-tested sequencing unit for sequencing a sample to be tested containing nucleic acid to obtain a sequencing result consisting of several sequencing data; a sample-to-be-tested correction unit for A normal data set is used to correct the sample to be tested; a segmentation unit is used to segment the corrected sequencing result to obtain several data fragments; and a detection unit is used to detect whether each data fragment is a copy number variation fragment. The application reduces the probability of missed detection of chromosomal abnormalities, reduces false positives and false negatives, has higher detection accuracy for chromosomal non-integrity and chromosomal copy number variation, and can detect smaller fragments under the condition of low fetal depth Chromosomal copy number variation.

The invention discloses an analysis method for targets entering a substation environment. The method comprises the following steps: an infrared sensing network is established around the substation, and the infrared sensing network surrounds the edge of a set area around the substation; constructing an infrared human behavior database, wherein a plurality of infrared human behavior templates are stored in the infrared human behavior database; The infrared sensing network performs infrared detection on the target entering the set area in real time, tracks the behavior of the target, and generates an infrared video image; and the infrared sensing network performs infrared detection on the target entering the set area in real time. Gaussian mixture model and background subtraction algorithm are adopted to extract the behavior spatio-temporal contour of the object in the infrared video image, and the behavior spatio-temporal contour of the object is enhanced in contrast. The behavior space-time contour of the target is compared with a plurality of infrared human behavior templates stored in the infrared human behavior database one by one, and similarity analysis is performed on the behavior space-time contour of the target and the infrared human behavior templates.

The invention discloses an intelligent pipeline capable of being monitoring remotely. The intelligent pipeline capable of being monitoring remotely comprises an intelligent pipeline carrier device control terminal, a handheld type data acquisition and communication end and an emergency platform monitoring end, wherein the handheld type data acquisition and communication end is used for data acquiring of the intelligent pipeline carrier device control terminal, and the emergency platform monitoring end is used for data processing, storing and distributing. When an anomaly or accident occurs, the monitoring end is acted as an emergency platform, charges for the intelligent pipeline carrier device control terminal and the handheld type data acquisition and communication ends, and behavioral intervention is carried out on the state of the intelligent pipeline carrier device control terminal and a holder of the handheld type data acquisition and communication end. According to the intelligent pipeline capable of being monitoring remotely, the effectiveness of a judgment is improved, the situations of misstatement and underreporting are reduced, so that the safety condition of the intelligent pipeline carrier device control terminal is effectively monitored.

The invention discloses an intelligent video analysis-based forest smoke and fire detection method and a special system thereof. The method includes the following steps: the real-time video images of the current detected scene, which are captured by monitoring cameras in a forest region, are transmitted to a video analysis server via a network; a smoke and fire detection module of the video analysis server detects and splits the moving foreground of the multiple frames of video images to generate a foreground list; according to colors, textures and motion features, the split moving foregrounds are preprocessed, and foreground areas without smoke and fire are filtered; on the basis of an intensive optical flow method, the continuous frames are dynamically analyzed, and dynamic motion features are extracted; and a support vector machine-based classifier classifies the features and outputs a detection result. The method and the special system thereof adopt the multi-video analysis technology to extract the unique features of smoke and fire, reduce false alarms, alarm failure and other circumstances, can analyze the video images of the day-night cameras in all weathers, can automatically detect smoke and naked fire in the daytime, and also can be used for detecting naked fire at night.

The invention provides a method and a system for detecting a reusable anomaly of a time sequence of business data, and the method comprises the steps: 1, obtaining business historical data, and detecting the business data according to the stability, periodicity, correlation and time granularity of the data; 2, matching a proper time sequence algorithm for the service data according to the detection result to form a mapping relation; 3, storing the mapping relation between the time sequence and the algorithm and parameters in a sample library; 4, acquiring real-time data of the same service, predicting the service data according to the mapping algorithm and the optimal parameter in the sample library in the step 3, and determining a prediction interval according to a prediction result; and step 5, comparing the relationship between the actual value and the prediction interval of the same time period, and giving an alarm when the prediction range is continuously exceeded for more than 10 times. The dynamic threshold value can be monitored in real time, excessive human participation is not needed, false alarm and missing alarm of alarm are reduced, and the labor cost is effectively saved.

The invention relates to a primer probe group, a kit and a detection method for PCR detection of a potato broom top virus, and belongs to the technical field of molecular biology. In order to solve the problem that the PMTV cannot be accurately and rapidly detected by an existing detection method, the invention provides a primer probe group for PCR detection of the potato broom top virus. The primer probe group comprises a primer probe group-PMTV2 and a primer probe group-PMTV3; the invention also provides a detection kit comprising the primer probe group and a detection method. According to the invention, cDNA obtained through RNA reverse transcription of a sample to be detected is taken as a template, and primers in a primer probe group are adopted to carry out duplex PCR reaction; the obtained PCR reaction product is hybridized with a chip connected with the probes in the primer probe group, developing is conducted, and if the corresponding position of the chip is bluish violet, it is judged that the chip is positive. According to the invention, the defect that naked virus RNA cannot be detected by ELISA is avoided, and false positive and false negative conditions of an RT-PCR method are reduced.

The invention relates to the field of wireless networks, and discloses a method and a device for a wireless terminal to identify a pseudo-WiFi network. In the present invention, the following steps are included: monitoring whether there is a four-way handshake packet in the current network communication data packet; if there is a four-way handshake packet, at least perform one of the following steps or a combination thereof: Step A, judge the four-way handshake packet Whether the handshake information in is consistent with the handshake information used when accessing the WiFi network last time, if not, it is determined that the current WiFi network has the risk of belonging to a pseudo WiFi network; step B, after accessing the WiFi network, obtain the specified IP The first route of the address is to determine whether the first route is consistent with the second route to the specified IP address when accessing the WiFi network last time, and if not, it is determined that the current WiFi network has a risk of belonging to a pseudo WiFi network. A more accurate identification method is given for pseudo-WiFi networks to reduce misjudgments and missed judgments.