105results about How to "Avoid the impact of accuracy" patented technology

The invention relates to an identification method and device of certificate characters. The method comprises steps of acquiring a to-be-identified certificate picture; using a character region detection model to detect the to-be-identified certificate picture so as to obtain a character region of the to-be-identified certificate picture; using a character identification model to identify the character region of the to-be-identified certificate picture so as to obtain identified characters, wherein the character identification model and the character region detection model are obtained throughthe pre-training based on training samples in a database; when receiving the instruction of character identification errors after the identification confirmation of a user, receiving correct characters input by the user, using the correct characters as output results, adding a to-be-identified certificate picture into the training samples of the database by taking the correct characters as true marks, and updating the database; and based on the updated database, re-training the character region detection model and the character identification model.

The invention relates to a method and device for tracing an online video target. The method comprises the following steps that image characteristics of a start frame in an online video are obtained and an initial background model is established; an image of a next frame is obtained; compared with the initial background model and the image of the next frame, so that a comparison result is obtained, and the initial background model is updated according to the comparison result; a foreground image is obtained and a foreground target is extracted; target characteristics are obtained through the online learning method and the foreground target is located so that the position information of the foreground target can be obtained; the position of the foreground target is marked and an image of a next frame after the marking is conducted is output; all the output image of the next frames are combined so that the moving trajectory of the foreground target can be obtained. According to the method for tracing the online video target, a real-time monitored video is processed, the target is traced at the first time, the phenomenon that the target is traced after all original video images are obtained is avoided, the instantaneity and effectiveness of data are guaranteed, and influence on the accuracy rate after multiple targets intersect and are blocked in an existing tracing mode is avoided.

The invention discloses a scene change detection method. The scene change detection method includes the following steps. Firstly, two images of a monitored scene are captured. Secondly, characteristics of the two images are abstracted with a rapid robust feature calculation method. Thirdly, the characteristics of the two images are matched through a K-dimension data space tree algorithm. Fourthly, estimation of transparent transformational matrixes are performed on matched characteristics of the two images through a modified random sampling consistency algorithm. Fifthly, according to estimated transparent transformational matrixes, space offset amount of the two images is worked out so as to obtain a scene change degree. The rapid robust feature calculation method, the K-dimension data space tree algorithm and the modified random sampling consistency algorithm are applied in the scene change detection method which has the advantages of being rapid, good in stability and high in precision. The invention further discloses a scene change detection system which comprises an image capturing module, an image characteristics abstracting module, an image characteristics matching module, a transparent change matrix estimating module and a scene change degree calculation module. The scene change detection method and the scene change detection system are widely applied in the field of video monitoring.

The invention discloses an experimental use multi-row pressure testing rake device arranged on the surface of a to-be-tested model and comprising a plurality of single-row pressure testing rakes witha rectification effect; the single-row pressure testing rakes are distributed and mounted on the areas of the surface of the to-be-tested model, wherein said areas are same in boundary layer speed distribution; the adjacent single-row pressure testing rakes are distributed on the surface of the to-be-tested model in a spanwise manner; the single-row pressure testing rake comprises a total pressuretube bank formed by breather pipes in a row; isolation blocks are arranged between the total pressure tube bank and the to-be-tested model, and the heights of the isolation blocks are arranged in a graded distribution along with the single-row pressure testing rake distribution. Compared with a conventional pressure testing rake experiment device, the effective measuring point interval is smaller, thus obtaining more test data in a unit length, and effectively improving the spatial resolution in the boundary layer measurement; the experimental use multi-row pressure testing rake device can reduce the interferences caused by the single-row pressure testing rake geometry profile on the flow field, and thus providing more accurate test data.

The invention discloses a three-axis permeability testing method, and belongs to the field of coal bed gas exploitation. A rock sample is placed in a model system (1), then pressure pulse is applied on the rock sample; a gas control system (2) supplies test gas to the model system (1); a temperature control system (3) controls the temperature of the model system (1); a three-axis loading system (4) applies confining pressure and axial pressure on the rock core clamper of the model system (1); a vacuumizing system (5) vacuumizes the test device; a gas sample collecting system (6) detects the gas components after the test; and an electrical controlling and monitoring system monitors the test process. The method comprises following steps: filling a sample into a tank, checking the airtightness, carrying out a permeability test, and cleaning the test system. The provided method can simulate a high temperature and high pressure environment of a deep stratum in a lab; pressure pulse is applied on the inlet end of a rock sample, the attenuation data of pressure pulse can be obtained, and then the rock sample permeability is calculated based on the attenuation data. The test efficiency ishigh, the external interference is little, and the test result is more accurate.

The invention provides an even pressure sleeve for measurement of static pressure of a pipeline fracture surface, and belongs to the technical field of engineering fluid mechanics measurement. The even pressure sleeve is composed of a round-ring-shaped top plate, a bottom plate and an outer side plate in an encircling mode on the outer wall of a to-be-measured pipeline, the top plate and the bottom plate are made of transparent materials, the outer side plate is connected with the outer round edge of the top plate and the outer round edge of the bottom plate, the inner round edge of the top plate and the inner round edge of the bottom plate are connected on the outer wall of the to-be-measured pipeline in a tightly abutting mode, eight pressure taking hoes are evenly formed in the same fracture surface, corresponding to the even pressure sleeve, of the wall of the to-be-measured pipeline, and an inner cavity of the even pressure sleeve is communicated with a to-be-measured fracture surface of the to-be-measured pipeline through the pressure taking holes. An exhaust vent with an exhaust valve is formed in the top plate of the even pressure sleeve, and a static pressure exporting hole is formed in the bottom plate of the even pressure sleeve. The even pressure sleeve for measurement of the static pressure of the pipeline fracture surface is reasonable and simple in structure, novel and exquisite in conception, and easy to achieve. Compared with a common pressure measurement device, reliability and accuracy of pipeline static pressure measurement are obviously improved. Due to the fact that the even pressure sleeve is made of the transparent materials, air exhaust conditions in the even pressure sleeve can be conveniently observed.

The invention provides a method for improving automatic driving reliability and electronic equipment. According to the method, firstly, the detection result of an environment characteristic value in apreset sensor work area is acquired; if the detection result is not in the permitted environment characteristic value scope of the preset sensor, and the detection result of the preset sensor is shielded. Namely, for each preset sensor, if accuracy of the output result can not be guaranteed through an external environment characteristic value, the detection result of the preset sensor is not taken as input of a later module, influence of the detection result with the false information on accuracy of obstacle sensing and vehicle positioning is avoided, reliability of the automatic driving process is improved, and hidden safety trouble is eliminated.

The invention supplies a method to gain the link evaluate parameter based on elasticity group circle. It sends data file to end node from sending node, the sending node gaining the sending time, receiving time of responding data file, and process time to the data report and responding data report. The link evaluate parameter would be determined by the sending time, receiving time, and process time. The invention could rapidly and accurately determine the link evaluate parameter and improve the RPR circle network maintainability.

The invention discloses a rigid cage guide defect detection device based on laser light screen scanning. The detection device comprises a driving device, a rotary arm, an encoder, a laser device and a camera, wherein the rotary arm is arranged on the upper end face of a No.2 cage guide by virtue of a support and is close to the inner side face of the No.2 cage guide; the rotary arm is rotatably arranged with the support by virtue of a shaft and is driven by the driving device; the encoder is arranged on the shaft; and the laser device and the camera are arranged in the front of the rotary arm. By locating the laser device at the top end of the cage guide and by scanning the cage guide in a perpendicular direction by virtue of a laser light screen, and by shooting through the camera and splicing images, an image spectrum of the cage guide varying with height can be obtained, so as to rapidly and accurately determine the type and the position of a cage guide defect; the detection device is applicable to the defect detection of a vertical shaft hoisting rigid cage guide, and the detection device is simple in structure, reliable and portable, high in data accuracy, wide in practicability and remarkable in economic benefit.

The invention belongs to the field of power generation equipment online monitoring and fault diagnosis, and discloses a power generation equipment parameter warning method based on a DBN network. Thewarning method includes the following steps: (a) acquiring operation parameter data of power generation equipment; (b) building a DBN network model of cross fitting between the operation parameters ofthe equipment; (c) training the DBN network model; (d) using the trained DBN network model to acquire the predicted values of target parameters in real time; and (e) comparing the actual values of the operation parameters of the equipment with the predicted values in real time, and giving a warning for the parameters. Based on the basic idea that the current operation parameters of the equipmentreflect the state of the equipment, a cross fitting model between the parameters is built using the correlation between the parameters of the equipment, and the model is incrementally trained througha small-batch training method. Through the method, the problem that the time series model is inaccurate under variable-load conditions is solved, and meanwhile, the influence of equipment maintenanceand natural deterioration on the accuracy of the model is avoided, and the safe and economic operation ability of the equipment is improved.

An embodiment of the invention discloses an audio processing method and device and a mobile terminal. The method comprises identifying each spokesperson in audio data to be processed; according to audio parameters of the audio data to be processed, splitting the audio data to be processed into a plurality of subaudio data, wherein the subaudio data is corresponding to unit statements; and markingthe spokesperson and speaking time information corresponding to each subaudio data. The audio processing method and device help users to search speeches released by some spokesperson in the audio datato be processed, and prevent influence of speech overlap on speech recognition accuracy and improve speech recognition accuracy.

The invention discloses a monitor of the state of health of a motor. The monitor comprises a data acquisition unit, a data processing unit, a storage unit, a display unit and a transmission unit, thedata acquisition unit comprises a vibration temperature integrated sensor and used for measuring vibration of the motor and a bearing temperature, the data processing unit samples and processes data acquired by the data acquisition unit, the storage unit stores a vibration total value, a vibration frequency spectrum, an enveloping total value, an enveloping frequency spectrum and the bearing temperature obtained by processing of the data processing unit, the display unit displays the vibration total value, the enveloping total value and the bearing temperature obtained through acquisition andprocessing, and the transmission unit transmits the stored data to the outside in a wireless manner by employing automatic timed transmission or manual transmission. According to the motor, the vibration of the motor, the bearing temperature, and the damage state of bearing parts can be simultaneously monitored, and the state of health of the motor is expressed in a comprehensive manner through multiple parameters.

The invention discloses an efficient regression test method based on hash eigen value detection of audio video decoder. The efficient regression test method is realized through the following steps: generating unique and specific hash eigen value from decoded data; and verifying the accuracy of decoding after version upgrading by comparing the hash eigen value of the previous version to that of the current version. On the one hand, it avoids the inaccuracy from omissions caused by subjective factors such as human senses, and on the other hand, all the method needs to do is turn to a comparing tool to compare the files in the previous version and the current version with hash eigen value. This eliminates the procedures and time to play and watch the materials and raises the efficiency greatly. In addition, as the hash eigen value decoded by a standard decoder is unique and accurate, so as long as the contents decoded by the decoder of a new version are consistent with those by the standard decoder, then, that result is deemed to be accurate.

The invention discloses a maximum torque current ratio control method for a built-in permanent magnet synchronous motor. The method comprises the following steps: 1) obtaining three-phase stator currents ia, ib and ic of a motor through sampling, then, obtaining currents idr and iqr under d and q coordinate systems through coordinate transformation, and obtaining a rotor position angle theta and amechanical rotating speed omega r of the motor through a rotary transformer; 2) extracting derivative information dTe / dbeta of the torque to the current angle, taking the derivative information dTe / dbeta as the input of an integrator, generating a given value idref of the d-axis current, and obtaining a given value iqref of the q-axis current under the given torque Teref; adding a feedforward decoupling item to the output of the PI controller to obtain a d-axis stator voltage reference value vdref and a q-axis stator voltage reference value vqref; and 3) adopting a voltage space vector pulsewidth modulation method to obtain six paths of PWM pulse trigger signals, and controlling the two-level voltage source type inverter to work so as to drive the motor to rotate.

The embodiment of the invention provides a product quality detection method and device and electronic equipment. The method comprises the following steps: acquiring a shot original image of a to-be-detected product and a binarization threshold corresponding to the to-be-detected product by detection equipment, and performing binarization processing on the original image according to the binarization threshold. The detection equipment can perform de-noising processing on first pixel points according to binarization values of multiple second pixel points adjacent to the first pixel points, namely binarization values corresponding noisy points subjected to binarization processing are adjusted, so that the influences of the noisy points in the image on the accuracy of subsequent product scratch detection are avoided. Finally, the detection equipment determines whether the image subjected to de-noising processing contains scratches. The product quality detection method is totally executed by the detection equipment, the detection equipment detects the quality of the to-be-detected products by adopting the unified detection standard, human intervention does not exist in the detection process, and the product quality detection accuracy is improved.

The invention discloses a textile fabric cutting device, including a base, wherein an operation table is fixedly connected to the top of the base, a control console is fixedly connected to the right side of the front face of the operation table, a protection frame is fixedly connected to the top of the operation table, moving grooves are formed in the outer walls of the two sides of the protection frame, driving frames are arranged in the middles of the outer walls of the two sides of the operation table, the middles of the driving frames penetrate through the protection frame, and a cutting device is arranged at the positions, located on the left side of the operation table, of the tops of the driving frames. The invention relates to the technical field of textile equipment. According to the textile fabric cutting device, the cutting device, a cutting mechanism, the operation table and a fixing block are arranged, the cutting mechanism moves at the bottom of the cutting device, a certain sliding space appears, the fixing block scrapes residues on the surface of the cutting mechanism, a cutter is prevented from being easily damaged when the cutting thickness is large, the situation that wrinkles are generated in a cutting process, and cloth is easy to damage is avoided, and cutting deviation is not easy to cause.

The invention discloses an uncertain data frequent item set publishing method based on differential privacy, which comprises the following steps: (1) giving a data set D, carrying out frequent item set mining on the D to obtain the support degree [lambda] of a frequent 1-item set ILIST and a kth frequent item set, and adding noise to obtain the support degree [lambda] of the frequent 1-item set ILIST and kth frequent item set; (3) extracting a length estimation value of a maximum frequent item set containing each item x from the D', and taking the length estimation value as hierarchical information; and (4) setting an initial node as a null set in the D', and screening by using hierarchical information constraints and subset frequent constraints to obtain a candidate item list NodeList ofthe current node; and (5) screening a top-k frequent item set and a support degree thereof in the FISet, and publishing the topk frequent item set and the support degree thereof. The whole process ofthe method meets differential privacy, and the privacy security of data can be improved.

An embodiment of the invention provides a continuous blood pressure measuring method and device based on a volume pulse wave signal, equipment and a storage medium. The method includes: acquiring thevolume pulse wave signal of a to-be-measured object, wherein the volume pulse wave signal at least comprises a pulse wave period; acquiring a blood pressure feature value according to the periodic value of each pulse vale and the amplitude change value of the pulse wave from the starting point of each pulse wave period to preset time; inputting the blood pressure feature value into a pre-trained linear blood pressure model of the to-be-measured object to obtain the blood pressure value of the to-be-measured object. The method has the advantages that a cuff type blood measuring method which needs to wait for cuff inflation and deflation during blood pressure measuring is avoided, practical application is facilitated, and continuous blood pressure measuring is achieved; the blood pressure feature value uses the amplitude information and speed information of the volume pulse wave signal at the same time, blood pressure measuring accuracy is increased, the linear blood pressure model is simple in calculation, and blood pressure measuring efficiency is increased.

The invention provides a method for determining an AC value of an orthokeratology lens. The method comprises the following steps of: S1, acquiring parameters of a cornea of a patient, wherein the parameters comprise a cornea flatness K value, a cornea e value and a cornea diameter; S2, calculating a cornea rise according to the parameters acquired in the step S1; S3, matching a lens rise closest to the cornea rise in a lens database; and S4, taking the AC value of the orthokeratology lens corresponding to the lens rise as the AC value of the orthokeratology lens needing to be produced. The invention further provides a system for determining the AC value of the orthokeratology lens, which is used for realizing the method. According to the method and the system, the influence of different experiences of different optometrists on the accuracy of measurement of the AC value of the orthokeratology lens is avoided, the difficulty of lens matching is reduced, and the standardization degree of lens matching is improved.

The embodiment of the invention provides a device for improving automatic driving reliability and an automatic driving system. In the device for improving the automatic driving reliability, a characteristic acquisition module firstly acquires detection results of environment characteristic values in preset sensor working regions; and in a processing module, if the detection results are not in an environment characteristic value permissible range of preset sensors, the detection results of the preset sensors are shielded. In other words, for each preset sensor, if the external environment characteristic value cannot guarantee the accuracy of an output result of the corresponding preset sensor, the detection result of the corresponding preset sensor is not adopted as input of following modules, the situation that the detection results with wrong information are adopted and affect obstacle sensing and vehicle locating accuracy is avoided, the reliability of the automatic driving process is improved, and potential safety hazards are eliminated.

The invention discloses a cutting device for automobile air conditioner condenser pipe processing. The cutting device comprises a supporting seat, a pushing device, a clamping and moving device, a driving device, a cutting device body, a cooling device, a fixing device, a dust collecting device and a cleaning device; the supporting seat is provided with a mounting groove, and the clamping and moving device is arranged in the mounting groove; the pushing device is arranged on the side wall of the clamping and moving device, the driving device is arranged at the top of the supporting seat, the cutting device body is arranged on the driving device, and the cutting device body is fixedly connected with the driving device; and the cooling device is arranged on the side wall of the cutting device body, the fixing device is vertically arranged at the bottom of the driving device, the fixing device is fixedly connected with the driving device, and the cleaning device is arranged on the side wall of the supporting seat. According to the cutting device for automobile air conditioner condenser pipe processing, the condenser pipe is fixed through the clamping and moving device, the condenser pipe is pushed through the pushing device, the condenser pipe is cut through the cutting device body, and the condenser pipe is cleaned through the cleaning device.

The invention discloses a riverway water quality detection device which comprises a bracket, a detection device and a sampling device; the bracket comprises a base and a supporting rod vertically arranged on the base; the sampling device comprises a sampling outer pipe and a sampling inner pipe; the sampling outer pipe comprises a sampling transverse pipe and a sampling vertical pipe; the detection device comprises a shell; the shell is internally provided with a detection unit and a sample placement groove; the detection unit comprises a testing electrode, a rotation mechanism used for driving the testing electrode to rotate and a lifting mechanism used for driving the testing electrode to move up and down; and a first lifting mechanism used for driving the detection device to move up anddown is also provided. For the riverway water quality detection device provided by the invention, by combining sampling and detection, detection can be performed directly in the sampling process, thedetection step is saved, and the detection accuracy is high; and in addition, the riverway water quality detection device performs sampling by enabling the sampling outer pipe to directly extend intoa riverway, avoids that the detection accuracy is influenced by movement due to water flow impact and is high in detection precision.

The invention provides an IMU-enhanced event camera optical flow estimation method and a system, and the method comprises the steps: firstly reconstructing a fuzzy brightness image at any moment intoa clear brightness image through an EDI model, building a basic optical flow model through combining with a brightness constant hypothesis, and enabling the EDI to represent event-based double integration; then adding the IMU as a constraint into a basic optical flow model to realize continuous optical flow estimation in any motion consistent scene; separating a background and a foreground for processing when a hybrid motion scene appears, introducing IMU constraints during background optical flow estimation, introducing sparse constraints during foreground optical flow estimation, jointly estimating the background optical flow and the foreground optical flow of the hybrid motion scene through an alternate iterative updating mode, and finally, the overall continuous optical flow of the scene is obtained through combination. The method is suitable for scenes comprise a motion consistency scene in which a single background moves or a foreground object moves consistently, and a mixed motion scene in which the background and the foreground move in different directions and sizes.

The invention relates to an authenticity analysis method and system. The method comprises the steps that authenticity query request information is received, and the item identification code of the authenticity query, the authenticity query place and the authenticity query user are determined according to the authenticity query request information; the last authenticity query place of the item identification code, each first historical authenticity query user and each related person of each first historical authenticity query user are determined according to the historical authenticity query record information of the item identification code; if the authenticity query user is different from each first historical authenticity query user and each related person of each first historical authenticity query user, or the authenticity query place is different from the last authenticity query place, the number of independent users of the authenticity query place is accumulated; and the authenticity analysis is carried out on a to-be-verified item corresponding to the item identification code according to the number of independent users of the authenticity query place. According to the scheme of the invention, the accuracy of the authenticity analysis can be improved.

The invention provides a control method for torque output during parking and starting, a motor controller and an automobile. The control method comprises the steps of in a parking state, when a signal that an accelerator pedal is treaded is received, controlling a motor to output torque according to an initial torque change rate; determining a target gradient according to the obtained current output torque, the longitudinal acceleration signal and the current parking state information; when the current output torque reaches the target starting torque value, initiating a slow gradient output stage with the slow gradient torque change rate smaller than the initial torque change rate, and in the slow gradient output stage, switching the EPB from the locking state to the unlocking state; and the target starting torque value and the slow gradient torque change rate being determined according to the target gradient. The rising speed of the output torque of the motor is slow when braking is released, brake millstone sound caused by counterbalance of the motor and the EPB is avoided, and the safety and comfort of vehicle starting are guaranteed.

The invention provides a statistical method and device for the utilization rate of HSDPA (High Speed Downlink Packet Access) service code resources. The method provided by the invention comprises the following steps: obtaining the emitting carrier power, the emitting carrier power which is not used by HS-PDSCH (High Speed-Physical Downlink Shared Channel) transmission or HS-SCCH (High Speed-Signaling Control Channel) transmission in all code channels, the maximum downlink timeslot emitting power, the single BRU (Basic Resource Unit) power of the HS-PDSCH and the number of the available HS-PDSCH BRU in the carrier frequency; according to the emitting carrier power and the emitting carrier power which is not used by the HS-PDSCH transmission or the HS-SCCH transmission in all the code channels, determining the actual emitting power average value of the HS-PDSCH in the carrier frequency; according to the single BRU power of the HS-PDSCH and the number of the available HS-PDSCH BRU in the carrier frequency, determining the maximum permissible emitting power of the HS-PDSCH in the carrier frequency; and according to the actual emitting power average value of the HS-PDSCH in the carrier frequency and the maximum permissible emitting power of the HS-PDSCH in the carrier frequency, determining the utilization rate statistical value of the HSDPA service code resources. The invention can improve the accuracy of the utilization rate statistics for the HSDPA service code resources and reduces the technical realization difficulty.

The embodiment of the invention discloses a behavior vector generation method and device, a terminal and a storage medium, and belongs to the information security field. The method comprises the stepsthat an operation record is acquired, wherein the operation record comprises behavior identifiers of at least one operation behavior generated in a target application; according to multiple preset groups, the at least one behavior identifier in the operation record is replaced with other behavior identifiers belonging to the same group, wherein the operation behaviors corresponding to the behavior identifiers in the same group have the association relation; and behavior vectors are generated according to the behavior identifiers obtained after the at least one behavior identifier is replaced.The generated behavior vectors can describe the operation behaviors generated in the target application to obtain an accurate application detection model through training, the operation behaviors ofa terminal user cannot be restored according to the behavior identifiers in the behavior vectors, therefore, the condition that the user privacy is leaked is avoided, and the security is improved.