80results about How to "Realize automatic discrimination" patented technology

The invention provides a real-time automatic monitoring method and a real-time automatic monitoring system for underground light non-aqueous phase liquid pollutant dispersion. The method comprises the following steps: taking a monitoring point location; obtaining a resistance variation range E after characteristic LNAPLs pollutant and characteristic LNAPLs pollution is caused at the monitoring point location; determining the length of a resistance probe; determining a distance between electrode rings; hanging one set of resistance monitoring device in a monitoring well according to an underground water flow direction; respectively injecting at least one set of resistance monitoring device at monitoring point locations upstream and downstream underground water; setting collecting parameters; transmitting monitoring data in real time; and automatically alarming when the resistance reaches the lower limit of the E. The real-time automatic monitoring method provided by the invention has the advantages of simple operation, accurate measurement and reliable operation; data can be monitored in real time and wirelessly transmitted; the diffusion process of the LNAPLs after leaking can be dynamically monitored in real time; the redistribution process of the LNAPLs can be also monitored during extracting rainfall or underground water; and the real-time automatic monitoring method provided by the invention can be widely applied to dynamic monitoring of underground pollution after the LNAPLs is leaked in large-scale petrochemical enterprises, gas stations and so on.

The invention provides a method for detecting the shielding of a camera based on automatic exposure. The method comprises the steps of acquiring a video image collected by the camera; computing the luminance value of each pixel in the image, and obtaining the average luminance value of the image; computing the luminance contrast of the image; judging whether the exposure of the image is normal by combining with the average luminance value and the luminance contrast; and judging whether the camera is shielded based on the exposure of the image. Accordingly, a device for detecting the shielding of a camera based on automatic exposure is also provided. According to the method or device provided by the invention, through computing the average luminance value and the contrast of the image, and judging whether the automatic exposure camera is shielded by combining with the average luminance value and the contrast, the automatic identification on partial shielding of the camera is achieved, and an alarm prompt is provided for a shielding phenomenon by an alarming mechanism so as to correct in time, and thereby the invalidity of monitoring pictures caused by that the camera is shielded is avoided.

The invention discloses an intelligent detection method of high-speed turnout crack damage based on vibration signal fusion. The intelligent detection method comprises the following steps: firstly, respectively mounting vibration acceleration sensors at a turnout tip, a turnout center and a turnout tail of a high-speed turnout, so as to collect vibration signals, and respectively carrying out CEEMD adaptive decomposition on the vibration signals of the three detection points, so as to obtain a limit number of IMF components; then, screening out a main IMF component by using correlation analysis, and calculating singularity entropy eigenvalue of the main IMF component; fusing singularity entropy eigenvalue of the vibration signals of a plurality of the detection points into an eigenvector, and inputting the eigenvector into LSSVM for training and testing, so as to achieve judgement of the working condition and damage type of the turnout. According to the intelligent detection method, different crack damage of the turnout can be effectively distinguished, and the method has good realtime performance and noise resistance, and provides a novel technological means for intelligent diagnosis of turnout fault.

The invention discloses a holder positive and negative side distinguishing device including a machine frame; the machine frame is provided with a holder positioning mechanism for positioning a holder, a holder distinguishing mechanism, a holder sorting mechanism, and a main control machine; a to-be-distinguished holder is transported to a holder supporting plate by a conveyor belt and then is pushed into a holder supporting plate position by a pushing device: (1) two first cylinder on the machine frame cooperate with holder positioning blocks for positioning the holder; (2) a lamp cover moves downwards until the bottom of the lamp cover makes contact with the holder supporting plate; (3) a camera takes photos to distinguishing that the holder is in a positive side or a negative side; and (4) according to the situation after judgment, an overturning direction of the holder supporting plate is controlled, and the holder is allowed to roll out from a holder outlet according to the same direction finally to enter a subsequent machining process.

The invention belongs to the field of voice recognition, and particularly relates to a phone-call recording access failure reason recognizing method. The method comprises the following steps: markingaccess failure reasons by signals; if reasons cannot be obtained by signal classification, extracting an audio fingerprint characteristic sequence from to-be-recognized phone-call recording, and searching from an audio fingerprint database by the sequence; if matched fingerprints can be found out, marking the access failure reason for to-be-recognized phone-call according to access failure reasonlabels in fingerprint key values; and if the matched fingerprints cannot be found out, recognizing audio contents into text contents by automatic voice recognition, classifying in an access failure document classifying model by a text classifying method on the basis of the contents, and marking the to-be-recognized phone-call recording by access failure reason classifying results obtained by classification. The method can recognize recording files in an offline manner, streaming phone-call voice can also be recognized, the universality is high, and the phone-call recording access failure reason recognizing method is suitable for different application scenarios of the call center.

The invention relates to an overall-modal extraction method. The overall-modal extraction method includes 1, establishing a beam element model of an overall structure, 2, reading node information and modal order of an overall finite element model and acquiring inherent frequency of modal at different order of the overall finite element model and feature vector related to mass matrix normalization, 3, extracting nodes according to node information, selecting the extracted nodes as major nodes of core pole and assistors, and nodes corresponding to branches of engines and effective load branches, 4, appointing normalized node number, 5, judging main vibration direction, 6, acquiring feature vector and generalized mass normalized according to appointed nodes, and 7, outputting modal results. With the overall-modal extraction method, efficiency of modal extraction is greatly improved and risks caused by failure from manual modal extraction are reduced.

The invention belongs to the field of automatic monitoring, particularly relates to a method for distinguishing and processing abnormal automatized monitoring data. The method includes: subdividing abnormal data in the monitoring data into accidental data, mutation data and slowly-varying data, wherein the accidental data are caused by abnormal construction factors and can be replaced by the average value of the monitoring data once the accidental data are determined; the mutation data and the slowly-varying data are caused by normal construction factors in projects, and high attention should be paid and corresponding processing measures should be taken once the mutation data and the slowly-varying data are determined. The method for distinguishing and processing the abnormal automatized monitoring data has the advantages that compared with ordinary data collection systems, an automatized data collection system adopted in the method can effectively distinguish, classify and correspondingly process abnormal data, automatized monitoring and reasonable construction guiding are realized truly.

The invention provides an algorithm and a system for judging household change relationship of a station area based on power consumption information and geographical position. The method calculates thedaily line loss rate of the station area based on the power transformer of the station area and the power consumption data of the user, then calculates the average line loss rate, processes the average line loss rate data by first-order difference, and accumulatively sums up the data to obtain the fluctuation of line loss, and classifies the station area according to the line loss rate and the fluctuation of line loss of the station area. The abnormal users are identified according to the influence of line loss fluctuation by users, and the user-change relationship is adjusted based on spatial proximity relationship. On the basis of user power consumption information and spatial information, the invention analyzes and excavates the potential correlation relationship among power consumption data, GIS data and machine account data, and analyzes and calculates the reasonable household change relationship with the goal of minimizing the line loss in the station area, thereby improving theaccuracy of line loss calculation in the station area.

The invention provides a twelve-lead standard electrocardiogram acute myocardial infarction intelligent determining system based on artificial intelligence. The twelve-lead standard electrocardiogramacute myocardial infarction intelligent determining system comprises a data acquisition system and a cloud platform data storage system. The data acquisition system is in communication connection withthe cloud platform data storage system. The cloud platform data storage system is connected with a data modeling analysis system. The data modeling analysis system is connected with a data display system. The data acquisition system comprises an electrocardiogram acquisition system, an angiography device, a clinical testing observing table, and time between chest ache occurring and electrocardiogram acquiring. The twelve-lead standard electrocardiogram acute myocardial infarction intelligent determining system depends on a big data cloud platform and establishes an artificial intelligence algorithm based on characteristic identification acute myocardial infarction (fixed property, fixed position and fixed time) of the twelve-lead standard electrocardiogram, thereby establishing a whole myocardial infarction automatic determining intelligent system platform, quickly realizing acute myocardial infarction automatic determining, and improving diagnosis efficiency.

This invention discloses an automatic searching and judging method of key information in the business cards, which comprises the following steps: to divide the characters string of the business cards got to get one row of string area or pass area; then to order the number of the character parameters and pass area number and to extract the first some areas; to judge the key font information area through judging whether the adjacent areas qualify the positions or captions.

The invention relates to an automatic identification method of plant leaves damage symptom, which comprises the following steps: 1)obtaining target leaves image; 2)processing the target leaves image; 3)analyzing image differential and area standard colour value; 4)performing quadratic function regression analysis and regression parameter calculation according to the area image standard colour value; 5)calculating threshold parameter by using a variance analysis method; and 6) using optimization parameter for discriminant analysis of the plant leaves damage symptom. By using image measurement and computer identification technology, automatic discrimination of the plant leaves V type and inverted V type damage symptom can be realized, the damage mechanism is determined through the damage symptom, so that automatic identification method has important guidance effect for accurately and quantificationally the plant leaves V type and inverted V type damage symptom as well as damage source; and the automatic identification method is in favor of solution of nondeterminacy and controversy of the plant leaves damage symptom during scientific research and production, and has important theory and practical meaning for automatically determining calamity and disease such as plant leaves meteorology.

The present invention relates to the seismic data quality monitoring field, in particular to an abnormal seismic trace automatic discriminating method and device. The method comprises the steps of obtaining the sampling point truth-value of a seismic trace; dividing the seismic trace into the time window data which is continuous and has the equal width from the zero hour, and according to the sampling point truth value, calculating the zero-crossing times of each time window data; summing the zero-crossing times of each time window data to obtain the total zero-crossing times of the seismic trace; comparing the total zero-crossing times with a first preset value; when the total zero-crossing times is less than the first preset value, determining that the seismic trace is an abnormal seismic trace of a dead trace type. According to the embodiment of the present invention, according to the total zero-crossing times characteristic of the abnormal seismic trace of the dead trace type, by just gathering and comparing the zero-crossing times simply during the discrimination process, and being compared with the prior art, the method of the embodiment of the present invention is simple, is high in operation efficiency, and satisfies the demand of the mass seismic data quality control.

The invention relates to a method for identifying a settlement pattern during shield tunnel construction. The method comprises the following steps of: 1, acquiring ground settlement monitoring data of a shield tunnel and storing the data into a database; 2, analyzing and identifying a transverse settlement curve pattern and a longitudinal settlement curve pattern of the tunnel and determining the pattern of the tunnel; and 3, analyzing to obtain construction parameters associated with the pattern of the tunnel, so that guidance and basis are provided for construction optimization. Compared with the prior art, the method has the advantages that the current situation that monitoring data and construction parameters can only be identified manually is changed; and automatic identification of the curve pattern, automatic generation of a curve analysis result and fitting of relevancy of various parameters are realized, so that the guidance and basis are provided for construction optimization.

The invention provides a constructor HSE integrated management system. The system comprises a personnel basic information real-time acquisition subsystem used for acquiring personnel basic informationand transmitting the acquired real-time information to a hydroelectric big database; a personnel construction equipment use permission automatic judgment subsystem used for construction equipment operation management; a personnel faint state automatic alarm subsystem used for automatically judging whether the personnel is in a faint state or not and giving an alarm; a personnel dangerous activityarea automatic recognition subsystem used for monitoring whether personnel enter a dangerous area or not and giving an early warning prompt; an import and export personnel information automatic recognition subsystem used for identifying and verifying import and export personnel information; and a safety personnel automatic deployment subsystem used for monitoring the distribution condition of thesafety personnel and carrying out safety management. Real-time information acquired by the personnel basic information real-time acquisition subsystem is used as a part of data basis for real-time analysis, so that the functions of the subsystems are realized. According to the system, full-automatic HSE comprehensive management of engineering construction field personnel is realized.

The invention discloses a positioning automatic discrimination system of acute myocardial infarction based on CNN neural network, and relates to the technical field of myocardial infarction positioning discrimination. The positioning automatic discrimination system comprises a data acquisition system, a cloud platform data storage system, a positioning discrimination analysis system and a data display system. A wearable ECG monitor is connected with to-be-discriminated personnel to record and generate a 12 lead primitive electrocardiogram; an electrocardiogram acquisition system acquires 12 lead primitive electrocardiogram data including wave amplitude of P wave, wave amplitude of a QRS wave group, wave amplitude of a ST section and wave amplitude of T wave; and the positioning discrimination analysis system utilizes a positioning discrimination model acquired based on CNN neural network training to carry out convolution calculation to acquire discrimination intermediate data, and thediscrimination intermediate data are mapped by a sigmoid function to acquire discrimination result data, so that positioning discrimination of occurrence positions of the acute myocardial infarction of the to-be-discriminated personnel is made. The positioning automatic discrimination system of the acute myocardial infarction based on the CNN neural network makes accurate positioning discrimination of the occurrence positions of the acute myocardial infarction of the to-be-discriminated personnel.

The invention discloses a license number recognition method. The license number recognition method comprises the following steps: S10, judging the category of a license plate; S20, selecting the corresponding license plate recognition scheme which is a 92-type license plate recognition scheme or a new-energy license plate recognition scheme according to the judged license plate category; S30, recognizing the license number according to the selected license plate recognition scheme, wherein S10 comprises the following steps: emitting projection light to a license plate by a light source; receiving reflected light of the license plate for the projection light; analyzing spectral power distribution of the reflected light; judging the category of the license plate according to distribution oflight waves of all frequency bands in the spectral power distribution. The license number recognition method is simple and high in reliability, and 92-type license plates and new-energy license platescan be judged and recognized automatically.

The invention discloses an airport low-slow small target prevention and control method based on photoelectric image automatic identification. The method comprises the following steps of obtaining a plurality of images of the low-slow small targets from a network, and carrying out the on-site actual measurement on, and collecting the plurality of images of the low-slow small targets; performing thefeature extraction on each image, classifying the target types corresponding to the images according to the extracted image features, and obtaining a low-slow small target classifier; preprocessing the to-be-prevented and controlled target image, extracting the image features, and then inputting the features into the classifier to obtain the type of the to-be-prevented and controlled target; anddetermining the position of the to-be-prevented and controlled target, and driving the corresponding prevention and control equipment to work in combination with the type of the to-be-prevented and controlled target. According to the method, the multiple prevention and control strategies of the airport area are achieved by classifying different targets, the optimal combination of different low-slow-speed small prevention and control devices is achieved, the influence of the low-slow small targets on the airspace of the airport is greatly reduced, the defects of a traditional manual monitoringairspace and the limitation of a single control strategy are overcome, and the prevention and control efficiency and precision of the low-slow small targets are improved.

The invention discloses a license plate identification system. The system comprises a type discriminating module for determining the type of a license plate, a scheme confirmation module for selectinga corresponding license plate identification scheme according to the license plate type determined by the type discriminating module, and a license plate identification module for identifying the license plate according to the license plate identification scheme selected by the scheme confirmation module. The type discriminating module determines the type of the license plate by determining frequency information or frequency change information of reflected light for white light of the license plate or determining the length-width ratio of the license plate. The scheme confirmation module selects to implement identification according to traditional or new-energy license plate by determining the type. According to the selected scheme, the license plate identification module segments, identifies and combines a license plate image. Thus, both traditional and new-energy license plates are discriminated and identified automatically, the method is simple, the computational complexity is low,an identification result is highly accurate, and the intelligent effect is good.

A mast verticality adjustment control method and device provided by the invention relate to the technical field of rotary drilling rigs. The mast verticality adjustment control method comprises the following steps: receiving an effective verticality adjustment signal; detecting the current angle value of a mast; calibrating the angle of the mast; selecting current output parameters from a preset current parameter curve according to the current angle value of the mast; and controlling the output current value according to the current output parameters, and adjusting the angle of the mast untilthe angle of the mast reaches zero. According to the mast verticality adjustment control method, the requirement parameters in the use of equipment can be automatically identified, the use parameterscan be automatically modified in order to meet the use conditions, and the use parameters of the mast can be automatically discriminated, so that the separate configuration of control parameters for each piece of equipment and the change of parameters after long-time use are avoided. The mast verticality adjustment control device provided by the invention has a function of self-adjusting the verticality adjustment parameter curve of the mast of a rotary drilling rig, and has strong adaptability.

The invention discloses a qualitative automatic discrimination system of acute myocardial infarction based on CNN neural network, and relates to the technical field of myocardial infarction qualitative discrimination. The qualitative automatic discrimination system comprises a data acquisition system, a cloud platform data storage system, a qualitative discrimination analysis system and a data display system. A wearable ECG monitor is connected with to-be-discriminated personnel to record and generate a 12 lead primitive electrocardiogram; an electrocardiogram acquisition system acquires 12 lead primitive electrocardiogram data including wave amplitude of P wave, wave amplitude of a QRS wave group, wave amplitude of a ST section and wave amplitude of T wave; and the qualitative discrimination analysis system utilizes a qualitative discrimination model acquired based on CNN neural network training to carry out convolution calculation to acquire discrimination intermediate data, and thediscrimination intermediate data are mapped by a sigmoid function to acquire discrimination result data, so that qualitative discrimination of whether the to-be-discriminated personnel suffering fromthe acute myocardial infarction or not is made. The qualitative automatic discrimination system of the acute myocardial infarction based on the CNN neural network makes accurate qualitative discrimination of whether the to-be-discriminated personnel suffering from the acute myocardial infarction or not.

The invention discloses an intelligent insect-expelling drug-throwing machine for multiple canine animals based on multiple identification. The intelligent insect-expelling drug-throwing machine comprises an FRID label, a drug-throwing machine shell, an RFID identification system, an infrared identification system, an iris identification or nasal line identification system, a drug bait storage system, a door control system, a light control system, a drug-throwing system, a camera system, a positioning system, a reminding alarm system, a control and communication system, a power supply system and a remote management platform, wherein the RFID identification system, the infrared identification system, the iris identification or nasal line identification system, the drug bait storage system,the door control system, the light control system, the drug-throwing system, the camera system, the positioning system, the reminding alarm system, the control and communication system, the power supply system and the remote management platform are arranged in the drug-throwing machine shell; the FRID label is arranged in or out of a target canine animal body; the FRID label is in communication connection with the RFID identification system; the drug bait storage system, the door control system and the drug-throwing system complete the processes of drug bait storage, drug-throwing starting andstopping, drug bait counting and weighing and the like under the command of the remote management platform and the control and communication system. In addition, the invention further discloses an implementation method of the drug-throwing machine. According to the intelligent insect-expelling drug-throwing machine and the implementation method thereof, the multiple canine animals can be distinguished simultaneously, accurate distinguishing is achieved, especially when the multiple canine animals appear simultaneously at night, related identification information is obtained, and meanwhile, accurate drug-throwing and insect-expelling are conducted respectively.

The invention provides an acquisition method and system for a steady-state time period of multi-element sensing parameters. The acquisition method for the steady-state time period of the multi-elementsensing parameters comprises the steps that sensing parameter data needing to be measured by a sensing monitoring system is acquired; according to monitoring index requirements of the sensing monitoring system, at least one key sensing parameter is screened out of the sensing parameter data; according to at least one key sensing parameter, sensing parameter vectors corresponding to multiple monitoring time points are acquired; according to a preset monitoring requirement, an initial monitoring time point meeting the preset monitoring requirement is determined; and according to a preset duration requirement, a preset distance threshold, the sensing parameter vectors and the initial monitoring time point, the steady-state time period meeting the preset distance threshold is determined. Through the acquisition method and system, the steady-state time period of the multi-element sensing parameters can be accurately acquired; monitoring time can be effectively saved, and resource waste canbe avoided; and automatic acquisition is realized, and the requirement on monitoring personnel is lowered.

The invention relates to a linear motor fault detection method and system based on big data analysis, and belongs to the technical field of motor fault diagnosis. The method comprises the following steps: obtaining control data of a linear motor to be detected, and corresponding operation data and / or state data; judging whether the operation data and / or the state data are / is abnormal or not, if yes, recording the operation data and / or the state data as abnormal data of a fault to be judged, wherein one piece of abnormal data corresponds to one piece of control data; and according to the abnormal data of the to-be-judged fault and a preset first graph data structure, judging a fault type corresponding to the abnormal data of the to-be-judged fault and a fault occurrence probability. According to the invention, the invention achieves the automatic discrimination of the fault type corresponding to the abnormal data in the operation process of the linear motor through the construction of the first graph data structure, does not depend on manual discrimination, solves a problem of low fault discrimination efficiency caused by manual discrimination, and improves the fault discrimination efficiency.

The invention discloses a speed reducer capable of automatically judging faults, and the speed reducer is characterized in that a plurality of first sensors are installed on at least one of a speed reduction transmission mechanism, a transmission shaft, a transmission bearing, a box body, a ventilation device, a lubricating device and an oil seal, and are used for detecting and uploading sensing signals of corresponding parts of the speed reducer. The speed reducer is additionally provided with a fault automatic judgment function, in the process from sensing signal collection to fault information determination, fluctuation of the end of a signal is avoided through fitting and decomposition, more accurate input is provided for signal decomposition, virtual noise is constructed through irrelevant signal components obtained after signal decomposition, and the speed reducer has the advantages of being simple in structure, convenient to use and high in practicability. Therefore, the extracted effective signals are more real, the noise reduction effect is better, and automatic fault judgment of the speed reducer and the elevator is achieved.

The invention provides a pipeline corrosion detection method based on wavelet transformation, which comprises the following steps: S1, knocking a pipeline by using a force hammer, and collecting signals of an acceleration sensor; S2, selecting a cmor wavelet as a mother wavelet, and carrying out continuous wavelet transformation and mapping on an input signal x (t) to obtain a time-frequency domain expression of x (t) wavelet transformation; S3, performing screening and filtering on the time-frequency domain expression to obtain a WTfilter (t, f); S4, calculating a double integral of the WTfilter (t, f) about frequency and time in the time range; and S5, setting pipeline corrosion degree judgment indexes y1 and y2, and judging the quality of the pipeline according to the integral value. According to the method, after threshold screening is creatively extracted, the integral of the characteristic range in the time-frequency plane is used as the pipeline quality discrimination characteristic, and a new thought is provided for a characteristic extraction algorithm based on the time-frequency domain; and the testing method is simple and convenient, the limitation of searching damage characteristics by manual comparison is avoided, and the automatic judgment of the pipeline corrosion quality is realized.

A basket capable of judging scoring and counting belongs to the field of sports goods and is composed of an installing plate, an iron ring, a phototube installing strip, a digital tube, a photoelectric cell unit, a storage battery pack, a switch, a control panel, an outer cover plate, shading tubes, an emitting tube, a receiving tube, a circuit board and an installing hole, wherein the shading tubes, the emitting tube, the receiving tube, the circuit board and the installing hole are arranged on the phototube installing strip. The iron ring is fixedly arranged on the installing plate, the phototube installing strip is arranged in the iron ring, the circuit board is arranged on the inner side of the phototube installing strip, the emitting tube and the receiving tube are arranged on the circuit board symmetrically, the shading tubes with certain length are installed on the outer sides of the emitting tube and the receiving tube, the control panel is arranged on the installing plate, and the outer cover plate covers the control panel to prevent the control panel from being damaged by a basketball. The photoelectric cell unit is connected with the storage battery pack, the switch is arranged between the storage battery pack and the control panel, and the digital tube is connected with the control panel. A single-chip microcomputer on the control panel of the device can judge scoring and counting and display scores through detection of signals of the receiving tube automatically.

The invention provides a vehicle information marking method, device and equipment and a storage medium, and relates to the technical field of image processing, in particular to the fields of computer vision, artificial intelligence, intelligent traffic, automatic driving and the like. According to the specific implementation scheme, a target area is determined in a road image; the boundary of a diversion area of the target area is determined, the boundary of a vehicle signal in the target area is determined, and the vehicle signal comprises a road marker line in the road image; marking the vehicle information by using the position relationship between the boundary of the vehicle information and the boundary of the diversion area of the target area; the positional relationship is determined based on a traveling trajectory of the vehicle. In this way, the error condition of manual labeling can be overcome. The position of the vehicle information in the target area is automatically marked, and data support can be provided for continuously expanding a map navigation database subsequently.

The invention discloses a transmitter test system and method based on a real-time digital-analog simulation system. The system comprises a tested transmitter, an information feedback device and the real-time digital-analog simulation system, wherein a transmitter test simulation model is established by the real-time digital-analog simulation system according to test parameters required by the tested transmitter, and a current and / or voltage signal is outputted in a small power signal mode to the tested transmitter; in combination with an electrical signal outputted by the tested transmitter and set values of the corresponding tested parameters, whether the electrical signal at an output end of the tested transmitter meets the excessive logic test condition is detected by the information feedback device according to excessive logic test pressure plates, whether the electrical signal at the output end of the tested transmitter meets the under-logic test condition is detected by the information feedback device according to under-logic test pressure plates, and the detection result is fed back to the real-time digital-analog simulation system in the form of a contact signal; whether the test result is correct is determined by the real-time digital-analog simulation system according to the received contact signal. The system is advantaged in that test efficiency of the transmitter can be greatly improved.

The invention relates to the technical field of power transmission line overhaul, in particular to a power grid damper defect automatic discrimination method. The power grid damper defect automatic discrimination method comprises the steps of that a source image is extracted, stored and analyzed; a wire image is extracted; a background interference image is filtered; self-adaptive matching of gray enhancement threshold values of the images is conducted, and the images are subjected to enhancement processing; an image library is subjected to intelligent learning, and automatic discrimination of an abnormal defect of a damper is achieved. By means of the power grid damper defect automatic discrimination method, patrol data are stored, processed and analyzed; the data are subjected to segmentation, the images are subjected to extraction, background noise reduction and enhancement, and a damper installation image library is established; an intelligent identification engine is adopted for learning, and image data subjected to extraction, noise reduction and enhancement are input; the abnormal position of the damper can be automatically identified, the workload is effectively lowered, the efficiency of patrol data interpretation is improved, the accuracy rate and timeliness rate of the abnormal position of the damper are increased, and a prerequisite is provided for safe operation of a power grid.