100results about How to "Reduce judgment error" patented technology

Disclosed is an automatic detection system for a mobile phone camera. The system comprises a test clamp module for placing a mobile phone; a signal feedback device for leading out a test result; a control unit which is electrically connected with the test clamp module through a test control line so as to control a test to start or end and output whether the test result is qualified and is eclectically connected with the mobile phone through a shooting control line so as to send a shooting instruction and feed a shot test image; and a test cart card module which is opposite to the test clamp module for shooting and comparison of the camera. The mobile phone is placed on the test clamp module, a control interface module sends a signal to be detected to the control unit, the control unit sends a test starting instruction to the mobile phone through the shooting control line so as to drive the mobile phone to start a test, after the test card chart module is shot, the mobile phone transmits a detected image to the control unit through the shooting control line for comparison calibration, and the control unit, according to a comparison analysis result, gives a feedback to a signal feedback device so as to output and display whether the test result is qualified.

The invention discloses a method to on-line monitor the insulated state of power cable metal shield shelves and its equipment, based on the single conductor cable metal circulation model, the method compute relation formula is as follows: Sensors will be packaged on the power cable metal shield shelves and collect circulation significant change signal, the signals are processed by the signal conditioning which is composes of sampling resistor, circuit protection, filtering circuit, after preliminary processing of signals, the signals sent to acquisition card, complete the signal acquisition, and then the data into PC, data is stored and computed by PC, monitoring results of the insulated state of power cable metal shield shelves base on the calculation of circulation relations of the cable metal shield shelves. This invention with less judgment error, no cable damage, through accident prevention of power cables to extend life of power cables, it assure the purpose safe and stable running of power cables.

Embodiments of the invention provide a live streaming video monitoring method and device. The method comprises: in a video sampling layer, extracting a target frame image from a live video stream; in a video analysis layer, inputting the target frame image into a preset classification model to classify the target frame image and outputting image classification information; and in the service processing layer, classifying the live video stream according to the image classification information and obtaining video classification information. The method classifies the live video stream by summarizing the image classification information, reduces the judgment error of a small amount of target frame images, improves the classification accuracy, realizes an online live streaming platform for all people, automatically monitors the image quality of large flow live streaming, and greatly improves the efficiency.

The invention discloses a petrol station oil-unloading area safety control system based on deep learning. The system comprises an oil-unloading area collection camera XC, a vehicle access detector CJ,a personnel real-time tracker RG, an oil-unloading state detector XJ, an oil-unloading operation specification detector GJ, and an oil-unloading state display and alarm ZG. The full-automatic, full-flow, and high-precision vehicle access detection, the personnel real-time tracking, the oil-unloading state detection and the oil-unloading operation specification detection at the petrol station oil-unloading area can be realized based on the standard oil-unloading flow and the safety control specification of the petrol station oil-unloading process, a video stream obtained by overlapping the oil-unloading state detection and the oil-unloading operation specification detection result is used for monitoring in real time / tracing the oil-unloading flow after event; the violation alarm information is output so as to record the worker and the oil-unloading operation specification detection result thereof, thereby solving the action specification detection and safety control problem of the worker at the oil-unloading area.

The invention discloses a method and a device for automatic detection of high-voltage resistance of a spark plug on a production line. The method is characterized in that a clamping mechanism clamps the spark plug to be detected on an assembly line; the ignition thread end is downwards inserted into a gas sealed chamber; high-pressure air is filled into the sealed chamber; a high-voltage pulse voltage string is supplied on a conductive screw rod of the spark plug; a voltage signal is subjected to sampling conditioning and window judgment by a circuit detection board according to a principle of voltage drop after high-voltage breakdown due to an insulation defect; the qualified pulse string is introduced into a controller; and the voltage resistance of the detected spark plug is judged whether the voltage resistance is qualified according to parameter setting.

The invention belongs to the data processing of computer systems. As an important part of an emergency coordinating system, a warning situation automatic classification system shares database data with the emergency coordinating system. The core component thereof is a classifier which is a computer processing system, and the software constructed by Bayesian algorithm is used for carrying out the automatic classification of warning situation in the classifier. The process of construction includes two parts: the construction of the classifier and the operation of the classifier. The invention uses the theory of machine learning, takes the data of historical warning situation information as learning data and adopts the modified simple Bayesian algorithm to construct the classifier. New warning situation data is input to generate standard warning situation vector information in classification, and the category of the warning information can be determined by a composite probability formula according to a preset probability threshold value and the occurrence number of the TOKEN serial. The classification system not only is adaptable to various emergency systems, but also can be an independent system. The classification is rapid, accurate, normalized and scientific.

The invention provides an intelligent machine tool working state identification system based on signal fusion. A sensor system is installed on a mechanical system of a machine tool and used for collecting a vibration signal, a temperature signal, a current signal and a speed signal of the machine tool; the vibration signal and the temperature signal are converted into digital signals through a data collecting system and then sent to a monitoring system, and the current signal and the speed signal are sent to the monitoring system through a numerical control system; and the monitoring system isused for controlling the vibration signal, the temperature signal, the current signal and the speed signal to be sent to a state signal database system and / or controlling the vibration signal, the temperature signal, the current signal and the speed signal to be sent to an intelligent data analysis system to achieve real-time data analysis and output an intelligent machine tool working state identification result. According to the intelligent machine tool working state identification system based on signal fusion, erroneous judgment brought by machine tool state identification obtained through signals measured by a power sensor singly is effectively eliminated, information fusion of multiple sensors is combined, a BP neural network method is adopted, therefore, errors in judgment are reduced, the system robustness is improved, and meanwhile basic conditions are provided for dynamic monitoring of the machining technology process.

The embodiment of the invention provides a scene detection method and apparatus of a live video. The method comprises the following steps: extracting a target frame image from a live video stream; identifying image scene information in the target frame image; and identifying live broadcast scene information of the live video stream according to the image scene information. According to the scene detection method and apparatus provided by the embodiment of the invention, the image scene information is summarized to perform scene identification on the live video stream, thereby reducing the judgment errors of a small number of target frame images, improving the scene identification accuracy, realizing a universal network broadcast platform, performing automatic detection on large-flow live broadcast scenes and greatly improving the efficiency.

The invention relates to a pest disaster monitoring method based on machine vision. The method includes the steps of installing a pest inducing device at a pest gathering place, and arranging an imagecollecting device facing the pest inducing device to collect images; identifying pests in the collected images and obtaining the number of the pests; if the number of the pests is greater than or equal to a preset threshold value for the number of the pests, separately extracting all the areas where the pests are located from the images to serve as suspected images of the pests, and judging the identification accuracy of each suspected image of the pests; according to the number of the pests and the identification accuracy of each suspected image of the pests, working out the pest disaster prediction level. By arranging the image collecting device facing the pest inducing device to automatically collect the images of the pests, large consumption of time and labor in manual visual inspection can be avoided, and the pests can be monitored in real time; by combining the number of the pests and the identification accuracy of each suspected image of the pests, the accuracy of calculating the pest disaster prediction level is higher, the obtained result is more significant, and the instructiveness in prevention and control of the pests is improved.

The invention relates to the technical field of air conditioning, in particular to a control method of an air conditioner and the air conditioner. The invention aims to solve the problem of low judgment precision of an existing refrigerant leakage judgment method. In order to achieve the purpose, the control method of the air conditioner comprises the steps that when the air conditioner runs in arefrigerant amount detection mode, the air conditioner is controlled to run in a preset running mode according to set running parameters; when the air conditioner operates stably, the indoor environment temperature, the inlet temperature of an evaporator and the outlet temperature of the evaporator are obtained; and based on the indoor environment temperature, the inlet temperature and the outlettemperature, the remaining state of the refrigerant amount of the air conditioner is judged. Through the control mode, the judgment precision of the remaining state of the refrigerant amount can be improved, and the use experience of a user is improved.

The invention discloses a method for detecting an elastic modulus of a concrete material in a reinforced concrete member. The method comprises the steps of wave velocity detection: detecting the transmission velocity of an impact elastic wave in the reinforced concrete member by one of a single-surface reflection method, a single-surface transmission method, a double-surface transmittance method or a surface wave method; elastic modulus calculation: calculating a movable tangent elastic modulus of the reinforced concrete member according to the transmission velocity of the impact elastic wave in the reinforced concrete member; correction: correcting the movable tangent elastic modulus of the reinforced concrete member to obtain a static elastic modulus of the concrete material in the reinforced concrete member. The method disclosed by the invention has the advantages that by the adding of the correction step for eliminating the influence of rebars on the detection of the elastic modulus, the elastic modulus of the concrete material is obtained, so that the safety of a construction / building structure is accurately evaluated.

The invention discloses a broadband multiple correlation flow velocity measurement method. Echo signals are sampled by adopting a time-interleaved sampling technology so that a discrete sampling sequence is obtained; multiple correlation algorithm processing and cross spectrum autocorrelation algorithm processing are respectively performed on the discrete sampling sequence; the number of multiple correlation points is increased in the process of multiple correlation algorithm processing, and real-time performance of flow velocity measurement is guaranteed by adopting odd and even paths of parallel multiple correlation processing; and cross spectrum interpolation time delay estimation is performed on the time-interleaved sampled data by cross spectrum autocorrelation algorithm processing so that a time delay value is obtained, and the time delay value is adopted to roughly measure the flow velocity value to act as the judgment standard of a flow velocity size interval, assist multiple correlation algorithm flow velocity measurement and correct the range of flow velocity measurement. Advantages are that multiple correlation algorithm processing and cross spectrum autocorrelation algorithm processing are supplementary for each other, sampling rate of the signals is enhanced by adopting the time-interleaved sampling technology, and multiple correlation algorithm flow velocity measurement is assisted through combination of the roughly measured flow velocity value of the cross spectrum autocorrelation algorithm so that measurement precision is high and measurement range is wide.

The invention discloses a method for judging flight route deviation and relates to an application technique for judging air route deviation in the route planning field. The method is characterized in that, according a transformation relation between terrestrial coordinates and space rectangular coordinates, the terrestrial coordinates of the current position point of an aircraft and the terrestrial coordinates of the route position point of the aircraft are transformed into the space rectangular coordinates, the coordinates of a position of a route which is closest to the current position point are calculated section by section by adopting a three-dimensional solid geometrical algorithm, the coordinates of the position which is closest to the current position point are inversely transformed into the terrestrial coordinates, and then whether the deviation is within an allowable route deviation range or not is judged by calculating the horizontal distance and the altitude difference between two terrestrial coordinate points. The method for judging the flight route deviation has the characteristics of precise algorithm, rapid judging speed, small error, stable and reliable performance and the like, and is particularly suitable for application fields of air route deviation alarm, flight navigation and monitoring, airspace management and the like.

The invention relates to a nut looseness detection method for a wind power tower drum. The nut looseness detection method includes: drawing a mark on the nut by using a fluorescent material; collecting an initial nut image and a current nut image through a visual sensor; converting the collected original image and the current image into an HSV color space from an RGB color space, converting the image into a binary image through a set threshold value, highlighting the drawn mark, subtracting the binary image of the original image from the binary image of the current image, comparing the position changes of the marks on the two binary images, and judging whether the nut is loosened or not. Compared with the prior art, the method has the advantages that whether the nut in the fan tower is loosened or not can be automatically detected and judged, the interference of background noise generated by factors such as dim light, narrow space and many obstacles inside the fan tower drum on a shotimage is effectively solved, good visual characteristics are provided for a computer to acquire and analyze pictures and judge nut looseness, judgment errors are effectively reduced, and good guarantee is provided for tower drum safety.

The invention relates to a standard colorimetric system used in titrimetric analysis which usually uses the end point color of an indicator to carry out judgment. Standard colors are filled in a standard white back lining, under two common concentration of the standard colors, two self ultimate colors of the indicator form a naturally transferred standard chromatape, namely, a full chromatape; under the working concentration, the error between the standard color of the titrimetric reaction stoichiometric point of typical titrimetric reaction and titration end-point is minus 0.1 percent, namely, the error is plus 0.1 percent between the standard color of the primary half drop of titrant of the stoichiometric point and the titration end-point, that is the standard color of the latter half drop of titrant of the stoichiometric point, a colorimetric standard white back lining with the same texture and chroma is arranged near a bulk colorimetric standard color. The standard colorimetric system can provide evidence for an experimenter to judge the end point color in an intuitionistic manner, and lead an analyst to carry out color matching under the white back lining of coincident color tone so as to reduce errors.

The invention relates to an identification method and system for the motion state of radio frequency identification tags. The method includes the step A that two sides of a check point are respectively provided with a directional antenna, the directions of the two direction antennas both are opposite to that of the check point, and distinction threshold values of forward pass or no-pass and distinction threshold values of backward pass or no-pass are set; the step B that signal strength, read by the two directional antennas, of the same radio frequency identification tags is differentiated to form difference curves; the step C that sine functions are used as generating functions, convolution calculation is carried out on the difference curves, and response curves are obtained; the step D that characteristic parameters on the response curves are selected and compared with the set threshold values at the step A, and then judgment results that whether the radio frequency identification tags forward pass or not pass and backward pass or not pass the check point can be obtained. The system comprises a threshold value setting and storing module, a difference curve generating module, a response curve generating module and a comparing and analyzing module. According to the method and system, identification error is small and both the moving direction and moving speed of the tags can be judged, and the method and the system are simple and convenient to use.

The invention relates to the field of planting, in particular to a peach tree grafting device. The peach tree grafting device comprises a fixed barrel and a cutting mechanism, wherein an air bag is fixed on the inner periphery of the fixed barrel, a fixed cavity communicated with the air bag is formed in the side wall of the fixed barrel, a fixed piston is glidingly connected into the fixed cavity, and fixed springs are further arranged in the fixed cavity; the cutting mechanism comprises a cutting barrel, a scion cutter, a rotating shaft and a stock cutter, positioning holes are formed in oneend of the fixed barrel, and positioning rods are fixed on the cutting barrel; the rotating shaft is in thread fit with the cutting barrel, rolling balls are arranged between the rotating shaft and the cutting barrel, and the stock cutter is fixed on one side, close to the fixed barrel, of the rotating shaft; a scion cutting tank is arranged on the rotating shaft, and the scion cutter is spiral and is fixed on the inner wall of the scion cutting tank. The grafting device in the scheme can cut scions and the stock simultaneously.

The invention relates to the technical field of on-line monitoring of power transformers, and discloses an on-line monitoring method for an oil-immersed transformer. The technical solution thereof comprises the following steps that: (1) a temperature sensor on the top of the oil-immersed transformer detects the temperature of the top oil in the oil tank; (2) the temperature sensor near the oil-immersed transformer detects the ambient temperature; (3) the temperature value of the top oil measured by the temperature sensor per second is compared with the flash point of the transformer oil; the ambient temperature value per second is compared with the preset maximum allowable temperature value; and the top oil temperature value per second is subtracted from the ambient temperature value per second; and (4) the alarm device alarms when receiving an abnormal signal. The method has the effect of improving the accuracy of monitoring the abnormal condition of the oil immersed transformer.

An expressway vehicle intelligent management system based on a vehicle networking technology of the present invention relates to a vehicle networking interaction technology of the vehicle-vehicle-people information applied to the expressways, is based on the accurate speed measurement and positioning principles of an industrial control high-speed coder, and extensively is used for an expressway vehicle comprehensive information real-time and driver interaction management system. The expressway vehicle intelligent management system based on the vehicle networking technology utilizes an international general zigbee networking technology and the working principle of the industrial high-speed coder, carries out the information calculation interaction with a background computer via a 3G or 4G network after acquiring the vehicle accurate speed and the accurate position information, utilizes a conventional mass storage and cloud computing technology to establish an intelligent platform, directly connects a driver intelligent mobile phone APP, and is used to manage, specify and warn the expressway driving vehicles and the driving behaviors. The system is convenient for the expressway administrators and the high speed traffic polices to realize the instant and efficient supervision, thereby comprehensively reducing the high speed peccancy and the accident probability.

The invention discloses a method and a circuit for detecting default phase of a three-phase alternating-current supply. The method comprises the following steps: sampling each phase of alternating-current signal of the three-phase alternating-current supply in a resistance step-down manner; rectifying the sampled three-phase alternating-current signal into a direct current signal; and detecting avoltage waveform of the direct current signal by utilizing an optical coupler, and determining default phase conditions of the three-phase alternating-current supply according to the voltage waveformof the direct current signal. With the adoption of the detection method, the obtained result is more accurate, the false alarm default phase condition caused by capacitance deviation or capacitance drop and the like by adopting a capacitor can be avoided in the resistance step-down manner, and the reliability of the default phase report is improved. In addition, the voltage waveform is detected byadopting the optical coupler, weak current output by a post-stage circuit is influenced by strong current output by the three-phase alternating-current supply, and accurate results are obtained.

A cell state supervision apparatus supervises the state of at least one cell in a cell set. The cell state supervision apparatus comprises a measure circuit measuring the state of the cess state and a caculation circuit. The measure circuit comprises a temperature measure part for measuring the temperature of the cell unit, a current measure part for measuring the current of the cell unit and a voltage measure part for measuring the voltage of the cell unit. Based on the cell temperature measured by the temperature measure part and the current measured by the current measure unit, the predetermined voltage corresponding to the predetermined charging state of the cell unit is caculated by the caculation circuit through the caculation expression predertermined by temperature secondary exponential function and temperature exponential function, the voltage of the cell unit measured by the voltage measure part is compared to the predetermined voltage calculated by the expression by the calculation circuit to determined the predetermined charging stage of the cell unit.

The invention relates to a fent color decision method for dacron / spandex textile. The testing method comprises the following steps: unfolding a dyed fent fabric, setting the transverse enumerator at N columns, measuring the transverse total length of the N columns, setting the longitudinal enumerator at N rows, and measuring the longitudinal total length of the N rows; taking a bulk final fabric made of the same material as the fent fabric, similarly setting the transverse enumerator at N columns, measuring the transverse total length of the N columns, setting the longitudinal enumerator at N rows, and measuring the longitudinal total length of the N rows; stretching the dyed fent fabric through a needle plate, and finally forming, wherein the pattern dimensions are as follows: the transverse length h=A*H, and the longitudinal length z=A*Z; according to the final forming process for bulk production, simulating on a small forming machine, feeding the fent fabric, and setting the temperature and the holding time; and using the formed small fabric samples for color contrast. The fent color decision method provided by the invention is simple to operate, and can greatly reduce the difference between bulk samples and small samples during color decision.

The invention provides a side board ribbing distribution optimization method, a side board and a washing machine provided with the side board, and relates to the field of washing machines. The optimization method is used for obtaining optimal ribbing layout, the frequency fixing characteristic of the side board is improved, the stiffness of the side board is high, and the washing machine is small in vibration amplitude and low in noise during work. According to the side board ribbing distribution optimization method, modal analysis is conducted on a side board original model, then morphologic optimization of reinforcing rib layout is conducted on the side board, a side board optimized model with optimal ribbing distribution is obtained, and then whether design requirements are met or not is verified. By means of the method, the obtained side board is applied to the washing machine, the problems of insufficient static stiffness and dynamic stiffness of a washing machine side board and resonance are solved, and accordingly the side board deformation problem of the washing machine in product transport and product falling processes are well improved.

The invention relates to a simple depth touch gauge. The simple depth touch gauge comprises a touch gauge through end, a measuring body, a first tensioning sleeve, a three-color handle, a second tensioning sleeve and a touch gauge stop end, wherein the measuring body is connected with the touch gauge through end in a sliding manner, and capable of sliding along the axial direction and rotating along the radial direction of the touch gauge through end, respectively; the touch gauge through end is fixed at one end of the three-color handle by the first tensioning sleeve; and the touch gauge stop end is fixed at the other end of the three-color handle by the second tensioning sleeve. Threads are arranged on each of the touch gauge through end and the touch gauge stop end; and the each of the touch gauge through end and the touch gauge stop end is smooth. The simple depth touch gauge has the advantages of being simple, compact and reasonable in structure, convenient to operate, low in manufacturing cost, high in detecting efficiency and the like. The measuring body can rotate on the through end, so that the step surface of the measuring body faces operating personnel for convenience in vision or touch judgment of the operating personnel; and moreover, the operating personnel use fingers to touch to improve detecting precision and reduce error in judgment.

The invention discloses a drill rod monitoring method and a drill rod monitoring device. The drill rod monitoring method comprises the following steps: (S1) the tension break variable and the drill rod length value in the break of a main windlass tension value are obtained; (S2) drill rods being lifted at present are determined according to the drill rod length value; and (S3) when the tension break variable is greater than or equal to the weight of outer-layer rods of the drill rods being lifted, the drill rods being lifted at present generate the rod carrying, and warning signals are output. The drill rod monitoring method, provided by the invention, can judge out a rod carrying area so as to bring convenience for operators to adopt remedial actions, and is low in judgment error.

The invention discloses a magnesium-aluminum alloy automobile covering piece die craftsmanship digital conversion manufacturing method, and solves the technical problem that an existing magnesium-aluminum alloy automobile covering piece mold is difficult to manufacture. The method comprises the steps of geometric profile center area determination, category / radian rebound correction column construction, unequal gap male die construction, trimming area reservation upheaval, core forming arc sharpening, programming and numerical control machining, craftsman reserved trimming area adjustment and shaping lapping A large amount of craftsman experience, workmanship and experimental results form a systematic data processing system, so that the problem of large-scale craftsman debugging in the mold manufacturing process is solved through early digital processing and machining, the manual labor amount of craftsmen is reduced, the judgment error of craftsman is reduced, and the workmanship requirement is reduced; from an experiment, the workload of craftsmen can be reduced by 65%-95%.

The invention relates to a spinal cord injury condition monitoring method based on a somatosensory evoked potential signal, which is mainly and technically characterized by comprising the following steps: step 1, obtaining the somatosensory evoked potential signal, decomposing the somatosensory evoked potential signal on a time domain and a frequency domain by adopting a short-time Fourier transformation method to obtain a time domain signal and a frequency domain signal; step 2, carrying out abnormal early warning judgment according to the change rate of the somatosensory evoked potential signal on the time domain signal and the frequency domain signal. The spinal cord injury condition monitoring method has the benefits that the somatosensory evoked potential (SEP) signal is decomposed into the time domain signal and the frequency domain signal, and the change rate of the somatosensory evoked potential signal is monitored, so that the function of monitoring the spinal cord injury condition is automatically realized, the problem of error caused by incomplete criterion in the prior art is solved, the monitoring accuracy is improved, the judgment error is reduced, and the judgment time is effectively shortened; meanwhile, an early warning can be technically sent out if the abnormal situation occurs, so that the occurrence of spinal cord injury accidents is prevented.

The invention discloses a sensor arrangement method for a test of an air conditioning system, and relates to the technical field of air conditioning systems, in particular to the sensor arrangement method for the test of the air conditioning system. The arrangement directions of an air conditioning system respectively comprise an air conditioning assembly air outlet, a foot blowing air outlet, anarrangement principle of a thermocouple, an evaporator temperature sensor, a warm air core body, a condenser temperature sensor, an air conditioning tube temperature and pressure sensor, system flow,system oil content, a head temperature sensor, a foot temperature sensor, an overall vehicle air outlet and ambient temperature outside a vehicle. According to the sensor arrangement method for the test of the air conditioning system, the situation of icing on the surface of an evaporator can be effectively avoided; the temperature distribution on the surface of the evaporator under different working conditions cannot be understood accurately if the temperature sensors on the surface of the evaporator are arranged unreasonably, so that the arrangement of the temperature sensors of the evaporator is affected; and by the arrangement method, icing risks in an operating process of the system can be reduced or completely eradicated as far as possible.

The invention relates to a method for improving the linearity of an interpolated pixel, applicable to a touch device provided with multiple sensors. The method comprises the steps as follows: sensing multiple voltage differences by sensors which are arranged in a sudoku form by taking a touch point as the center; respectively calculating a horizontal axis vector, a longitudinal axis vector, a first inclined vector and a second inclined vector according the voltage differences; allocating the first inclined vector and the second inclined vector to a horizontal axis and a longitudinal axis; and calculating a horizontal axis value and a longitudinal axis value according to the sum of vectors distributed on the horizontal axis and the vertical axis, so as to determine the position of an interpolated pixel. The method in the invention can effectively reduce judgement error of an interpolated pixel point and improves the linearity of pixel display.