216results about How to "Reduce false alarm rate" patented technology

The invention discloses a vehicle collision-proof early warning device, which is mainly and technically characterized in that the vehicle collision-proof early warning device comprises an image acquisition module, an image processing module, a millimeter microwave radar speed and distance measuring device, an ultrasonic sensor, a speed sensor, a brake sensor, an accelerator sensor, a steering sensor, a single-chip microcomputer and an early warning device, wherein the image acquisition module is used for acquiring information at the front of a vehicle and transmitting the information to the image processing module, the image processing module processes the information and then transmits the information to the single-chip microcomputer, the millimeter microwave radar speed and distance measuring device simultaneously transmits measured relative distance and relative speed between an ahead vehicle and the vehicle on which the vehicle collision-proof early warning device is installed to the single-chip microcomputer, the ultrasonic sensor transmits the measured speed of the vehicle on which the vehicle collision-proof early warning device is installed, the single-chip microcomputer processes the image information and the distance and speed data and then judges whether the vehicle is in a warning critical state or not, and finally a result processed by the single-chip microcomputer is displayed through the early warning device. The vehicle collision-proof early warning device has the advantages that the false warning rate of the early warning device is effectively reduced, the accidents such as rear-end collision are reduced and the transport efficiency is improved.

The invention discloses a method and device for automatically processing a malicious code sample. The method includes the steps of obtaining the malicious code sample, extracting static characteristics from the obtained malicious code sample, matching the static characteristics of the malicious code sample with known static characteristics in a static characteristic library, continuing to extract dynamic behavior characteristics from the malicious code sample if it cannot be judged that the obtained malicious code sample is a malicious code according to the static characteristic library, matching the dynamic behavior characteristics of the malicious code sample with known dynamic behavior characteristics in a dynamic behavior characteristic library, and conducting false alarm feedback if it can be judged that the obtained malicious code sample is not a malicious code according to the dynamic behavior characteristic library so as to prompt that the malicious code sample is a non-malicious code. The malicious code sample can be accurately recognized, and the false alarm rate of the malicious code is lowered.

The invention relates to a multi-sensor fused comprehensive car alarm system which comprises a multi-sensor fusion module, a central processing module, a wireless communication module, a GPS (global position system) positioning module and an alarm device, wherein the multi-sensor fusion module is connected with the central processing module; the central processing module is connected with the alarm device and the GPS positioning module, and carries out interaction with a mobile phone of a car owner by the wireless communication module; the multi-sensor fusion module adopted by the invention comprises a microwave Doppler sensor, an acceleration sensor, a pyroelectric infrared sensors and a Hall switch element, so that the multi-sensor fusion module can carry out all-round monitoring on carinformation and cope with various theft means, thereby improving the security and reliability of the system, and reducing the rate of error alarming; the alarm device comprises an antiriot spray device, so that the car owner can remotely control a thief and subdue a thief on site; the wireless communication module supports video transmission, picture transfer and short message supervisioning; andmeanwhile, the car alarm system also has the GPS positioning module, thereby facilitating the position search of a car.

A system and method for dual-band detection of incoming threats using an initial passive detection system and a selectively activated active detection system. Advantages include improved threat detection accuracy, reduction of false alarms, and reduced radiation emission from the active system, thereby making the active system more difficult to detect and reducing irradiation levels of users and bystanders. Variations include systems employing passive optical or electro-optical detectors, systems employing active RADAR or LADAR detectors, and systems connected to alarm signal/threat mitigation systems. The system may be configured for use on ground vehicles or small watercraft. Variations of the system may be specifically configured to detect incoming munitions launches.

The invention relates to pipeline leakage positioning system and method based on collaborative detection with negative pressure wave and sound wave, which belong to the technical field of fault diagnosis of a petroleum transmission pipeline. Leakage points in a pipeline are detected based on a conventional negative pressure wave technology, an upper computer and a pressure detection lower computer are arranged in two dispatching rooms at the front and rear ends of a long petroleum transmission pipeline and lower computers for sound wave signal detection are arranged on the long petroleum transmission pipeline at an interval of 0.5-1km. The invention has the advantages that by using the collaborative detection with the negative pressure wave and the sound wave, the invention can position the leakage points of the pipeline precisely and lower the false report rate effectively; because a GPRS technology is adopted for communication, the invention is free from paving lines, has low cost and is convenient to move. The positioning system benefits the avoidance of large data volume transmission and the improvement of the communication efficiency. The positioning method used by the upper computer is simple and easy to use.

The present invention is a signal processing method to significantly improve the detection and identification of source emissions. More particularly, the present invention offers a processing method to reduce the false alarm rate of systems which remotely detect and identify the presence of electronic devices through an analysis of a received spectrum the devices' unintended emissions. The invention identifies candidate emission elements and determines their validity based on a frequency and phase association with other emissions present in the received spectrum. The invention compares the measured phase and frequency data of the emissions with a software solution of the theoretically or empirically derived closed-form expression which governs the phase and frequency distribution of the emissions within the source. Verification of this relationship serves to dramatically increase the confidence of the detection.

The invention discloses a method and a device for dynamic threshold anomaly traffic detection of DDOS (distributed denial of service) attack. The method includes: acquiring historical data of anomaly traffic detection, and generating a traffic threshold above a tolerable line and a traffic threshold below the tolerable line of anomaly traffic detection according to the historical data of anomaly traffic detection; comparing a traffic value obtained by detection to the traffic threshold above the tolerable line and the traffic threshold below the tolerable line of anomaly traffic detection; if the traffic value is larger than the traffic threshold above the tolerable line or smaller than the traffic threshold below the tolerable line, sending out an anomaly traffic detection alarm. By the method and the device for dynamic threshold anomaly traffic detection of the DDOS attack, the traffic threshold above the tolerable line and the traffic threshold below the tolerable line can be generated effectively by utilization of the historical data, the defect of high difficulty in threshold determination in the process of anomaly traffic detection of the DDOS attack n the prior art is made up, accuracy of DDOS anomaly traffic detection is improved significantly through a statistics-based threshold determination method, and false alarm rate of anomaly traffic detection alarm is lowered.

The invention relates to a method to monitor and eliminate general unknown viruses, which comprises: inducing virus basic behaviors composted of a plurality of behavior factors through analysis and research on large amount of computer virus, determining specificity of each behavior factor and each basic virus behavior and forming a specificity-adjusting virus behaviors database which is used to judge whether a program belongs to a virus; simultaneously forming an adjustable normal program false-virus behaviors database according to false-virus behaviors generated on normal programs, wherein the database is used to judge whether a program belongs to a normal program; sorting and processing the viruses which are detected by the above method. The invention can improve the specificity of unknown virus and reduce the error report rate of behavior analysis, ensures the adjusting properties of virus behaviors specificity, increases quantity and range of unknown viruses detecting, and has distinctive advantages compared with prior computer virus monitor technique.

The invention provides a double-channel real-time bioaerosol monitoring method and a device thereof. The method is characterized by comprising the following steps: (1) concentrating atmospheric aerosol particles; (2) carrying out double-channel ultraviolet induced fluorescence detection; (3) data-processing; (4) cleaning the surface of an enrichment board of the particles and regenerating; and (5) repeatedly carrying out a new round of loop detection. The device comprises an aerosol-particles concentration unit, a double-channel ultraviolet induced fluorescence detection unit, a particle enrichment board regeneration mechanism, a rotary table and a control system. The invention can measure the density and the change of bio-particles such as bacteria, spores, viruses and the like in an ambient atmosphere in real time so as to achieve an early warning function on bioaerosol content in the atmosphere.

An apparatus comprising a first detection device, a second detection device and a processing circuit. The first detection device may be configured to generate a first signal in response to a first type of input. The second detection device may be configured to generate a second signal in response to a second type of input. The processing circuit may be configured to (i) determine whether the first signal is a known type of signal, (ii) determine whether the second signal is a known type of signal and (iii) generate a warning signal in response to the first signal and the second signal.

The invention discloses an intelligent fire monitoring and evacuation indicating system and method. The system comprises a local monitoring system which is mounted in a monitored building for fire monitoring and evacuation indication. The local monitoring system comprises a monitoring center and node devices. The node devices comprise environment information detection apparatuses mounted at positions prone to fire in the monitored building and detection indication apparatuses mounted in corridors of floors or/and staircases. Communication among the node devices and between the node devices and the monitoring center is performed through wireless transceiving modules. The intelligent fire monitoring evacuation indicating system and method have the advantages of being rapid in reaction, accurate in information, labor saving, safe, reliable and the like; and the cost is low, the practicality is high, and the efficiency of fire monitoring and crowd evacuation can be improved greatly.

The invention provides an sql (Structured query language) injection detection method and device, and belongs to the technical field of network security. The method comprises the following steps: extracting http (hypertext transport protocol) parameters from an http request; combining each http parameter with each predefined dynamic sql template into an sql sentence; performing grammatical analysis on each sql sentence respectively, determining that the http parameter does not lead to an sql injection attack if the sql sentence is inconsistent with an sql grammatical rule, and outputting a marked character string which corresponds to the sql sentence if the sql sentence is consistent with an sql grammatical rule; matching the marked character string with a marked character string set having an sql injection attach characteristic, and determining that the http request is an sql injection attack if matching is successful. According to the method and the device, the execution efficiency of sql injection detection can be increased, and the false alarm rate of sql injection detection is lowered.

The invention discloses a system and method for front vehicle collision early warning based on a visual sense. The system comprises a vehicle front scene obtaining module, an effective area calculation module, a vehicle detection module, a vehicle positioning module and a collision early warning module. The vehicle front scene obtaining module is used for recording the real-time road condition scene in the front of the vehicle; the effective area calculation module is used for calculating projection positions of detection areas a bicycle road ahead of the vehicle and a left lane and a right lane which are adjacent to the bicycle road in an image; the vehicle detection module is used for detecting vehicles inside the effective areas and determining a vehicle candidate area; the vehicle positioning module is used for obtaining the confidence coefficient of the vehicle inside the vehicle candidate area and an accurate position of the lower rim of the front vehicle; and the collision early warning module is used for publishing early warnings of different levels according to the collision times. The system and the method are applicable to various server weathers, provide vehicle collision time, set the early warning level and timely and accurately make early warning prompt.

The invention discloses a radar video linkage based intelligent boundary security system which comprises a radar, a numerical control linkage ball device, a server, a switchboard, a workstation, a monitoring video storage device, a monitoring video display device and an alarming device, wherein the server is connected with the radar through an RS485 bus; intelligent boundary security system software is installed on the server; the control on the three-dimensional linkage of the radar and the numerical control linkage ball device is the core of the software; the numerical control linkage ball device, the server, the workstation, the monitoring video storage device, the monitoring video display device and the alarming device are connected with a same local area network through cables; the numerical control linkage ball device is connected with the local area network through a photo-electric converter and an optical fiber. The radar video linkage based intelligent boundary security system has the beneficial effects of active detection, intelligent identification, locating and tracking, low false report, short construction period and high adjusting speed.

The invention relates to a correlative detecting method of network intrusion events, which pertains to the network security technical field. In the correlative detecting method of the invention, an intrusion detecting system reports the intrusion warning information and stores all warning informations respectively into four database tables. Beginning from the attack of the fourth database table, each warning item in the fourth database table is searched and matched previous hacker attack behavior in the third database table which stores backdoor; if the searching and matching are successful, new searching and matching are carried out; the warning items in the third database table are searched and matched in the second database table which stores promotion authority; the process is analogized until warning items in the first database table are matched, and the results concerning successful matching are stored into the event database table after being associated; beginning from the fourth database table when all warning items are finished, the searching and matching are carried out from the third database table, the process is analogized till the first database table are matched. The correlative detecting method of the invention reduces false warning rate of intrusion event detecting system and improves the accuracy rate of reporting security event.

The invention relates to a breakout prediction method for continuous casting, belongs to the technical field of metallurgical continuous casting, and in particular relates to a breakout prediction method in plate blank continuous casting process. In terms of temperature monitoring-based breakout prediction algorithm, the self-adapting genetic algorithm is introduced into the BP (Back Propagation) neural network to achieve automatic optimization of the network structure and improve identification accuracy of a temperature monitoring model. In terms of friction monitoring-based breakout prediction algorithm, a friction monitoring model based on logical decision and neural network is established. Based on the prediction results of the two models, the method achieves combination of temperature monitoring accuracy and friction monitoring sensitivity, provides the prediction mechanism based on temperature monitoring in addition with friction monitoring and couples the temperature monitoring model to the friction monitoring model, so as to ensure the accuracy of final prediction and effectively reduce the false prediction rate.

The invention discloses an urban road video monitoring system, which includes a front-end data acquisition subsystem, a network transmission subsystem and a central monitoring subsystem which are connected in sequence. According to the invention, by use of the front-end data acquisition subsystem, the network transmission subsystem and the central monitoring subsystem, a reliable monitoring system of low false alarm rate, powerful function and simple structure is achieved, the functions of red light running, speed detection, high-definition photographing, license plate recognition and the like can be achieved, unified management of urban road and reasonable dispatch of vehicles are facilitated, and the urban road video monitoring system has good economic and social benefits. The invention can be widely applied to various road video monitoring systems.

The invention discloses an intelligent home intrusion detection method based on deep learning, and relates to an online system, in particular to a method combining a fuzzy neural network and deep learning and used for judging whether intrusion behaviors exist in the network or not. According to the method, deep learning and the fuzzy neural network are organically combined, and the problems that an existing intelligent home intrusion detection technology is difficult to process a large amount of high-dimensional data, the false positive rate is high, the false negative rate is high, and the detection rate is low are solved. By means of an offline system, running parameters of the online system are determined, and the online system conducts real-time intrusion detection. Compared with the prior art, the method is an active monitoring model aiming at intelligent home network attack behaviors, and has the advantages of being high in detection rate, low in false negative rate and the false positive rate, high in real-time performance and the like.

The invention provides a DNS covert channel detection method based on a neural network. The deep learning is applied to the technical field of DNS covert channel detection, the accuracy of the existing DNS covert channel detection method is improved, the false alarm rate of the existing DNS covert channel detection method is reduced, numeric features are combined with character features for use toshorten the model training time and to improve the accuracy rate, a suspicious domain name is queried by whois, whether the domain name is legal is automatically judged according to rules, and the domain name that does not conform to a neural network prediction result is used for retraining the neural network to form a loop so as to continuously improve the neural network.

The embodiment of the invention discloses a locomotive fire prevention and control system, which comprises a plurality of fire detection subsystems arranged in different fire prevention areas, a control subsystem and a fire extinguishing execution subsystem, wherein the fire detection subsystem is provided with a plurality of different types of fire detection sensors according to detection requirements corresponding to the fire prevention area; the control subsystem communicates with the fire detection subsystem, a locomotive microcomputer network control system, a locomotive vehicle-mounted protection system and a ground locomotive remote diagnosis system, the control subsystem can carry out logic judgment on the fire state of the fire prevention area where each fire detection subsystem is located in real time based on the detection data collected by each fire detection subsystem, and corresponding processing actions are executed and displayed according to corresponding judgment results. By means of the locomotive fire prevention and control system, complete and closed-loop fire protection can be formed for the whole locomotive from fire detection, control judgment to fire extinguishing execution, and the safety of railway transportation is fully guaranteed.

The invention relates to an anti-theft device by utilizing infrared detection and directions to trigger alarms, which can particularly effectively achieve the aim of sending off an alarm if illegal and no alarm if legal. A plurality of infrared emitters distributed in rectangular frames or non-rectangular frames and a plurality of corresponding infrared receivers form an infrared barrier detector net, and a front infrared receiving head and a rear infrared receiving head in each infrared receiver are respectively a triggering infrared receiving head and a non-triggering infrared receiving head; when a person or an object enters from the direction of the triggering infrared receiving head, an infrared triggering head triggers the infrared barrier detector net to send off the alarm, and when the person or the object enters from the direction of the non-triggering infrared receiving head, the infrared triggering head does not trigger the infrared barrier detector net to send off the alarm.

The invention relates to an alarm method and device and belongs to the Internet technical field. The method comprises the following steps that: a monitoring video is obtained; whether a moving object exists in the monitoring video is detected; when a moving object exists in the monitoring video, whether the moving object is a specified object is judged; and when the moving target is the specified object, alarm information can be prohibited from being sent to a terminal. According to the alarm method of the invention, when the moving target is the specified object, the alarm information can be prohibited from being sent to the terminal, and therefore, false alarms caused by the movement of the specified object can be avoided, and a false alarm rate can be reduced, and alarm accuracy can be improved.

The invention discloses an old people stumbling detection method based on a handset, comprising: collecting the information of the triaxial accelerometer and the gyroscope sensor of a handset to obtain an acceleration and an angular acceleration; employing a wavelet filtering method to filter the collected acceleration and angular acceleration to obtain a post-filter acceleration and angular acceleration; constructing time domain characteristic vectors of the post-filter acceleration and angular acceleration, and employing an improved SVM (Support Vector Machine) learning classification algorithm to perform a first level filtering on the characteristic vector of the stumbling acceleration of the triaxial accelerometer; obtaining a second level stumbling determination result through the characteristic vector of the stumbling angular acceleration of the gyroscope sensor; and finally determining whether a person stumbles and completing stumbling identification. The old people stumbling detection method is used for detecting stumbling, and has the characteristics of high accuracy, short detection time, and low cost; the method can be applied only through taking a handset, and is convenient to use.

The invention relates to the field of burglary prevention for street lamp cables and discloses an anti-theft alarm method for a street lamp cable with a self-checking function. An anti-theft front end is arranged in a substation control cabinet; the anti-theft front end provides a weak direct current for the street lamp cable at a time interval when a street lamp is not opened and switches the weak direct current into an alternating current for supplying power to the street lamp at a time interval when the street lamp is opened; the anti-theft front end detects the weak direct current and the alternating current on the street lamp cable to judge whether voltage and current are abnormal, determine types of abnormal states of the voltage and the current and transmit abnormal information of the front end simultaneously; and an anti-theft tail end is arranged in a terminal pole of the street lamp at the tail of the cable, and the anti-theft tail end judges whether a line is abnormal by detecting whether the voltage is present on the street lamp cable, transmits abnormal information of the tail end, and determines a fault result of the street lamp cable according to the front-end abnormal information of the anti-theft front end and tail-end abnormal information of the anti-theft tail end. The invention is applicable to anti-theft alarm for the street lamp cable.

The present invention belongs to the technical field of image registration, and in particular relates to a method for optical and synthetic aperture radar image registration. In the invention, the registration is based on the use of line characteristics and the space constraint criterion; through the extraction of the corresponding line characteristics and the solution to the corresponding intersection points, the point characteristics can be obtained indirectly. The intersection points are subjected to characteristic vector description, and the initially matched point pairs are screened through the space constraint criterion to finally complete the image registration. Compared with the commonly used registration method, the image point pair registration rate is raised to some extent, which facilitates the subsequent image fusion.

The invention discloses a network attack detection method based on a spectral clustering algorithm. An NS2 network simulator is used to realize effective simulation of a denial of service (DoS) attack stream, and a DoS attack model is established. Experimental data is acquired and verified. Algebraic transformation is applied, and a spectral clustering algorithm based on a transition probability matrix is designed, and then a channel of a suspicious concealed type network attack is screened out. The network attack detection method is advantageous in that a system is designed, tested, and verified, and then the conclusion and the suggestion of the Internet attack detection method are acquired; compared with the prior art, by adopting the abnormality detection method of the spectral clustering algorithm, the false alarm rate is low, and the detection rate is high.

The embodiment of the invention discloses an intelligent electrocardio monitoring method and device. The method includes the steps that a probe is utilized to collect human electrocardiosignals; the signals are converted to digital signals through a digital analog converter, and primary denoising processing is conducted on the signals; signal segments are classified through a trained deep learning structure model; different processing methods are adopted according to different signals, and the false detection risk is minimized. According to the intelligent electrocardio monitoring method and device, through a signal collection mode with less interference to a patient and a signal classification means, the heart rate can be detected more accurately, the false alarm rate of clinical equipment is lowered, the rehabilitation process of the patient can be better guaranteed, the work efficiency of the medical staff can be further improved, the workload of the medical staff is reduced, and the sensibility to the alarm of the intelligent electrocardio monitoring device is improved.

A method, image processing system, and computer-readable recording medium for item defect inspection are provided. The method is as follows. A test image and a reference image of a test item are received. A test block and a corresponding reference block are obtained from the test image and the reference image to generate a test block image and a reference block image. The test block image and the reference block image are respectively partitioned into multiple sub-blocks. All interfering sub-blocks are identified and filtered out from the test block image and the reference block image, and a shift calibration parameter is obtained accordingly. The test block in the test image is calibrated based on the shift calibration parameter to generate a calibrated test block image. The calibrated test block image and the reference block image are compared to obtain defect information of the test item corresponding to the test block.

The invention provides an improved building fire-fighting facility alarm signal analyzing and processing method. The method includes the steps of 1 inputting basic information of building fire-fighting facilities, 2 judging operating states of the alarm fire-fighting facilities, 3 shielding alarm signals at non-working time, 4 determining fire or monitoring alarm information of the fire-fighting facilities, 5 determining fault alarm information of the fire-fighting facilities, and 6 outputting actual state information. According to the method, the false alarm rate of the building fire-fighting facilities is lowered, interference to fire-fighting management from unrelated alarm information is reduced, the use efficiency of a fire-fighting networking system is improved, a monitoring result is made to be objective and credible, and objective data for further evaluating system effectiveness of the networking fire-fighting facilities are provided. The method has practical value and application prospects in the application field of fire-fighting engineering technologies.

The invention aims to provide a method and a device for detecting a loophole of a source code. The method of the invention comprises the steps as follows: determining one or a plurality of sensitive nodes in an abstract syntax tree based on the abstract syntax tree corresponding to the source code; finding dependent nodes possessing dependency relationship with parameter information corresponding to initial nodes based on the abstract syntax tree so as to use the dependent nodes as child nodes of the initial nodes; using the dependent nodes as new initial nodes when the dependent nodes do not correspond to user input information, and repeating the above steps. The advantage of the solution of the invention is that the method and the device can be used for establishing path-sensitive information based on the dependency relationship of the data via analysis based on manners such as assignment operations so as to reduce false alarm rate of the loophole, and improving integrity of the obtained sensitive path, and comprehensiveness of detecting the loophole.