90results about How to "Improve discrimination efficiency" patented technology

The invention provides a method and a device for determining the age of a face image. The method comprises: performing rough estimation of the age of the image according to texture and shape characteristics; if the result of the rough estimation falls into an oldest age group or a youngest age group, finishing the operation; otherwise, performing detailed estimation on the effective face image according to face image characteristic library acquired from a specific age group sample, and determining the group to which the age of the face image belongs according to the result of the detailed estimation. The method and the device are used for improving the accuracy of image determination.

The invention relates to the technical field of brain tumor image processing, and specifically discloses a brain tumor segmentation data enhancement method based on a generative adversarial network. According to the method, an image-to-image conversion framework-Tumor GAN network architecture (comprising a generator, a global discriminator and a local discriminator) is constructed; n2-n virtual image pairs can be synthesized from n real data pairs (a multi-modal brain tumor picture set with segmentation tags), and a data enhancement model is obtained through training, so that the quality and authenticity of an image synthesized by the data enhancement model are ensured; on the basis of an attention region provided by a semantic tag, a region perception loss function and a region loss function are added in a generator, training of an image with authenticity to an image translation model is assisted, image details can be reserved, and the generalization performance of the model is improved; by applying the local discriminator to be used in cooperation with the global discriminator, the discrimination efficiency can be improved, and the model is helped to generate a medical image pairwith more real texture details.

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 discloses a holder positive and negative side discrimination device; a holder guiding mechanism, a holder discrimination mechanism, a holder moving mechanism and a holder lifting mechanism are all mounted on a machine frame; the holder guiding mechanism is vertically connected with the holder discrimination mechanism, the holder moving mechanism is located above the holder guiding mechanism, and the holder lifting mechanism is connected with the tail end of the holder guiding mechanism; what a camera shoots is a width of a chamfer angle of a central round hole of a holder, when the central axis of the camera has a 45 degree included angle with a horizontal plane, the face of the chamfer angle of the central round hole of the holder is just vertical to the camera, and at the moment, the chamfer angle width shot by the camera is an actual width. According to comparison of the shot chamfer angle width and a standard chamfer angle width, the positive and negative sides of the holder can be determined; and the determination efficiency is high, and problems of large artificial determination error, low efficiency and the like are reduced.

The invention relates to a method and a device for discriminating image sharpness and belongs to the image processing field. The method comprises steps that an original image of which sharpness is to be discriminated is acquired, and the original image is zoomed to a designated pixel to acquire a target image; a first characteristic of the target image is acquired, and the first characteristic is identified through a gray matrix, an image contrast matrix and a gray gradient maximum accumulation sum matrix; the first characteristic is pre-processed to acquire a second characteristic; the second characteristic is inputted to an artificial neural network model acquired through pre-training, and whether the original image is intelligible is discriminated according to an output result of the artificial neural network model. According to the method, through the gray matrix, the image contrast matrix and the gray gradient maximum accumulation sum matrix, the first characteristic of the image is extracted, image characteristic variety is guaranteed, the image can be reflected from multiple aspects, a discrimination result based on the first characteristic is guaranteed to be relatively accurate, the second characteristic is inputted to the artificial neural network model acquired through pre-training to discriminate whether the original image is intelligible, and a discrimination mode is relatively simple.

The invention provides a detection method and system based on floating car resident trip hot spot regions. The detection method comprises the conversion step and the analysis step executed by a server, wherein longitudes and latitudes are converted into plane distances, and then the analysis step is executed, the longitude and latitude data are summarized for conducting spatial clustering analysis, and therefore the resident trip hot spot regions are obtained. The detection method and system have the advantages that the accuracy of the resident trip hot spot region can be improved, the determination efficiency is improved, and convenience is provided for resident trip selection.

A method for dynamically regulating sequence of service discrimination includes counting up access frequency of various services by counting up service access state in certain time interval when service discrimination is carried out on new service link access by gateway then carrying out dynamic regulation on sequence of service discrimination for raising efficiency of service discrimination.

The invention discloses a stability analysis system and method of a time-varying power system. The system comprises a data acqusition module, a wind velocity fitting prediction module, a time-varying system establishment and order reduction module, a stability criterion solving module and a result output module, wherein the data acqusition module is used for acquiring data; the wind velocity fitting prediction module is used for performing fitting prediction on conditional wind velocity trend in short time, sectioning the wind velocity and calculating a conditional characteristic wind velocity probability density matrix of each section according to a fitting model; the time-varying system establishment and order reduction module is used for establishing a continuous markov time-varying power system model considering a wind velocity random characteristic and reducing order of the system; the stability criterion solving module is used for solving whether the system is stable by using a feasibility problem solving method in a linear matrix inequality. Through the stability analysis system and method of the time-varying power system, the problem that the stability of the time-varying power system is difficult to distinguish caused by the wind velocity random characteristic in case of fan connection can be effectively solved.

The invention belongs to the technical field of electric power safety equipment, in particular to a piece of intelligent warning duty equipment. The intelligent warning duty equipment includes a mobile carrier and a duty device shell. The mobile carrier is equipped with an identity authentication module. The duty device shell is equipped with an infrared sensing module, an identity recognition module, a central processing module and an alarm module. The output terminal of the infrared sensing module is electrically connected with a first input terminal of the central processing module. A firstoutput terminal of the central processing module is electrically connected with the input terminal of the identity recognition module. The output terminal of the identity recognition module is electrically connected with a second output terminal of the central processing module. A second output terminal of the central processing module is electrically connected with the input terminal of the alarm module. The identity recognition module and the identity authentication module communicate through a wireless communication protocol. The intelligent warning duty equipment combines an infrared sensing technology, an identity recognition technology and an automatic alarm technology, fits the actual needs of daily work, and has the advantages of high recognition rate and accurate recognition.

The invention discloses equipment and a method for comprehensively and rapidly judging a water bursting source of a mine. The equipment comprises a water level sensor, a light-splitting light path module, a water temperature sensor, an analog-digital converter, a CPU (central processing unit) and a displayer, wherein signal output ends of the water level sensor, the light-splitting light path module and the water temperature sensor are connected with a signal input end of the CPU through the analog-digital converter; and a signal output end of the CPU is connected with the displayer. According to the equipment and the method for comprehensively and rapidly judging the water bursting source of the mine, a mine water bursting source comprehensive judgment model is composited with a conventional hydrochemical analysis method, a conventional water level analysis judgment method and a conventional water temperature analysis judgment method, so that defects in the three conventional methods can be mutually made up, the reliability of a mine water bursting source comprehensive judgment result is improved, the error rate of a mine water bursting source comprehensive judgment is reduced, and a more accurate mine water bursting source comprehensive judgment result is obtained accordingly.

The invention relates to the technical field of computer vision, and discloses a binocular stereoscopic vision based head detecting and counting method which solves a problem that a head detecting and counting scheme in the traditional technology is high in calculation amount, inaccurate in counting and low in accuracy. The method disclosed by the invention comprises the steps of a, calibrating a binocular image acquisition system; b, performing monitoring region image acquisition by using the calibrated binocular image acquisition system, and acquiring a passenger flow scene image; c, performing preprocessing on the acquired passenger flow scene image; d, acquiring a depth image of the image; e, detecting heads with different depths in the depth image by using a contour line searching method; f, tracking the detected heads; g, detecting new heads in a new video frame, and updating the number of heads; and h, outputting the current number of heads. The method disclosed by the invention is applicable to head detecting and counting for passenger flow scenes.

The invention discloses an equivalent modeling method for a direct-drive wind farm of considering an off-load circuit. The equivalent modeling method is characterized by being carried out according to the following steps: (1) building a detailed model of the direct-drive wind farm; (2) generating input wind energy of various units at random; (3) collecting terminal voltage drop values of the units; (4) setting an active current reference value of a grid-side converter; (5) finishing a cluster division according to the conduction condition of the off-load circuit; (6) representing the units with equivalent values; and (7) building an equivalent model of the direct-drive wind farm. According to the equivalent modeling method, accurate representation of the conduction condition of the off-load circuit of each internal unit of the direct-drive wind farm is achieved; and the fitting precision of an equivalent model of the direct-drive wind farm on power external characteristics of the wind farm is improved.

The invention discloses a live video big data accurate recommendation method and system based on artificial intelligence, and relates to the technical field of data identification and analysis. The method comprises the following steps: selecting and identifying a reference live video; selecting a core reference key frame image; identifying to obtain a core reference high-frequency vocabulary; identifying the live video to be identified; selecting and identifying the to-be-identified core key frame image to obtain a target high-frequency vocabulary; and if at least one core reference high-frequency vocabulary is the same as the target high-frequency vocabulary, carrying out significance detection, carrying out clustering by adopting a spectral clustering method based on positive and negative samples, and determining whether the to-be-identified live video is a target recommended live video or not. According to the method, key frame detection and representative key frame image selection methods are utilized to effectively select the core frame image in the live video, and an OCR-based high-frequency vocabulary matching method and a clustering method are combined, so that effective and accurate judgment is realized, and related video data are accurately recommended for a user.

The invention provides an expressway congestion judgment method, device and equipment and a medium. The method comprises the following steps: acquiring a video frame image acquired by a target expressway monitoring camera; identifying vehicles in the video frame image to generate a vehicle detection frame matched with each vehicle; and according to each vehicle detection frame and a preset highway uplink and downlink region, judging whether the highway uplink and downlink region is congested or not. According to the method, the video frame image collected by the target expressway monitoring camera is acquired, and the vehicle in the video frame image is identified so as to generate the vehicle detection frame matched with each vehicle, so that whether the uplink and downlink areas of the expressway are congested or not can be judged according to the vehicle detection frames and the preset uplink and downlink areas of the expressway. Compared with a mode of monitoring and judging by manually polling real-time monitoring videos, the method of the invention has the advantages that the consumed human resources are relatively small, and the efficiency of judging the traffic jam of the expressway is relatively high.

The invention discloses a driver distraction judgment method based on a transverse driving speed. The driver distraction judgment method based on the transverse driving speed comprises the following steps: 1, collecting transverse traveling speed data at fixed time intervals; 2, making a time domain diagram of the transverse driving speed; 3, removing invalid data segments; 4, performing frequencyspectrum analysis on effective data segments; and 5, judging whether the driver has a distraction phenomenon and a distraction degree in the time interval, wherein the distraction degree comprises three grades of mild distraction, moderate distraction and severe distraction. By the invention, whether the driver has a distraction phenomenon in a driving process can be rapidly and effectively judged, thereby reducing the probability of traffic accidents caused by distraction of the driver.

The invention relates to the technical field of corrugated carton printing machines, and aims to provide an informationized intelligent carton printing system which comprises a control module and a printing module. An acquisition module for acquiring image-text information, image-text position information and color information images of printed cartons is mounted in the printing module, the outputend of the acquisition module is electrically connected with the input end of the control module, a serial port of the control module is connected with a background, and an identification model for judging the carton printing quality is arranged in the background; and before the cartons are conveyed out of the printing module, the identification model obtains the image-text information, the image-text position information and the color information image, collected by the acquisition module, of the printed carton, and judges whether the image-text information, the image-text position information and the color information printed on the cartons meet the printing requirement or not. The problem that an existing carton printing machine relies on manual judgment of the printing quality condition of the cartons and is low in judgment efficiency is solved. The informationized intelligent carton printing system has the effects of automatically judging the printing condition of the cartons andimproving the judgment efficiency.

The invention provides a display method of an address book contact persons and an apparatus thereof. The method comprises the following steps of when a second object which has a same name with a first object is detected to exist in an address book, sending invitation information to a target object, wherein the target object is the first object or the second object; receiving note information which is associated with the target object and is sent by the target object based on the invitation information; and displaying the note information in a business card display menu of the target object. Therefore, according to the note information, the contact persons who have the same name can be accurately and rapidly distinguished so that distinguishing efficiency of a user is increased and simultaneously operation efficiency of the user is improved.

The invention discloses a network interface communication rate matching method and device, equipment and a storage medium. The method comprises the steps of obtaining parameters of a network interface; calculating a maximum throughput rate currently allowed to be sent by a network interface for receiving the message according to the parameters; calculating a current maximum throughput rate of a sending interface used for sending the message; directly sending a message from a sending interface to a receiving interface or sending the message to the receiving interface in a delayed manner by utilizing a comparison result. The device comprises: an obtaining part for obtaining parameters of a network interface; a first calculation component which is used for calculating the maximum throughput rate currently allowed to be sent by the network interface for receiving the message according to the parameters; a second calculation component which is used for calculating the current maximum throughput rate of a sending interface used for sending the message; a control component which is used for sending the message from the sending interface to the receiving interface directly or in a delayedmode. The method, the device, the equipment and the storage medium can carry out communication rate matching forwarding on network interfaces of different systems and different rates to adapt to dynamic changes of a network.

This invention provides a temp sensing device for sensing if the region with a mark object temp surpassing a preset value is a major region. The device includes an image pick up unit, an image database and a comparing unit, among which the image pick up unit is for extracting a surface temperature image of the mark object, the image process unit is for labeling the region with the temperature of the surface temperature image surpassing the preset valure with a preset method. The image database stores multiple surface major region images for defining the major regions of the surface temp images, the comparing unit carries out an image discriminating technology.

The present application relates to a data processing method, apparatus, computer readable storage medium and computer device. The method includes: acquiring virtual object interactive animation data,wherein the virtual object interactive animation data corresponds to multiple interactive animation frames; Identifying the target interactive scene to which the interactive animation frame belongs according to the preset interactive scene rules; segmenting The virtual object interactive animation data according to the target interactive scene to which each interactive animation frame belongs, andobtaining the virtual object interactive animation scene data corresponding to the target interactive scene. Establishing The relationship between the target interactive scene and the corresponding virtual object interactive animation scene data. The scheme provided by the present application can improve the accuracy of distinguishing the virtual object interaction scene and improve the distinguishing efficiency of the virtual object interaction scene.

The present invention discloses a method for automatically diagnosing cancerous cells, comprising: S1: obtaining a first tissue image and a second tissue image by focusing under different electron microscopes through different focal lengths; S2: performing edge detection on the first tissue image to obtain the nucleus boundaries of all the cells in the first tissue image; S3: aiding in the calibration of the center points of all the cells in the first tissue image according to the second tissue image; and S4: statistically counting the cell membranes among adjacent cells of the first tissue image to determine whether canceration occurs in the tissue in the first tissue image or not based on the statistical result. The method of the invention has the following advantages: utilizing the differences existing between cancer cells and healthy cells; and using the computer vision method for automatic characteristics extraction to distinguish the cells at high distinguishing efficiency and accuracy.

The embodiment of the invention provides an indoor and outdoor user terminal judgment method and a base station. The method comprises the steps that MDT data sent by an undetermined user terminal in apreset area is acquired, and if it is judged that the MDT data do not comprise GPS information, the MDT data sent by a plurality of auxiliary judgment user terminals in the preset area are acquired;Wherein the MDT data sent by the auxiliary discrimination user terminal comprises GPS information; According to the MDT data of the undetermined user terminal and the MDT data of each auxiliary discrimination user terminal, determining a target auxiliary discrimination user terminal in the plurality of auxiliary discrimination user terminals; And performing indoor user terminal or outdoor user terminal judgment on the undetermined user terminal according to the RSRP values respectively corresponding to the undetermined user terminal and the target auxiliary judgment user terminal. According tothe method and the base station provided by the invention, the indoor and outdoor user terminal discrimination efficiency is improved.

The invention provides an intelligent monitoring method for a potential fracture area of a reinforced soil retaining wall. The intelligent monitoring method comprises the steps that information geogrids are laid in a reinforced soil retaining wall in a to-be-monitored area in a layered mode; a digital three-dimensional model of the reinforced soil retaining wall in the to-be-monitored area is established; the strain position of an intelligent reinforced belt and corresponding strain mechanical data are acquired in real time, and the information is imported into the digital three-dimensional model; and the corresponding strain position information meeting the strain threshold value is extracted, all the strain positions meeting the strain threshold value are sequentially connected into a curve, and the area where the curve is located is the potential fracture area. The invention provides the brand new intelligent method for monitoring the potential fracture area of the reinforced soil retaining wall, the potential fracture area can be quickly and accurately monitored, early judgment and early warning are realized before a slope body is damaged, greater loss is avoided, the judgment difficulty of the potential fracture area is greatly reduced, and the discrimination efficiency of the potential fracture area is improved.

The invention relates to the technical field of computers, in particular to a legality judgment method and device for vehicle scheduling behaviors and computer equipment. The method comprises the following steps: receiving a scheduling request sent by a user terminal, wherein the scheduling request carries a scheduling vehicle identifier; searching scheduling behavior data of a scheduling vehiclecorresponding to the scheduling vehicle identifier; judging the legality of the scheduling behavior data by using a preset judgment algorithm; and when each judgment algorithm judges that the scheduling behavior data is legal data, judging that the scheduling behavior is a legal behavior. By adopting the method, the efficiency and accuracy of judging the legality of the vehicle scheduling data canbe improved.

The invention discloses a tobacco shred specification customization system based on long tobacco shred reprocessing. The tobacco shred specification customization system comprises a rack, and further comprises a long tobacco shred sorting device, a long tobacco shred pull-apart device and a cluster tobacco shred slitting device which are arranged on the rack, the long tobacco shred sorting device is adjacent to the long tobacco shred pulling-apart device, the long tobacco shred sorting device comprises a plurality of first needle rollers which are sequentially arranged in the advancing direction of the tobacco shreds, and the first needle rollers are driven by a first transmission mechanism and a first driving mechanism to rotate clockwise in the same direction; the long tobacco shred pulling-apart device is adjacent to the long tobacco shred sorting device, the long tobacco shred pulling-apart device comprises more than one second needle roller which is sequentially arranged in the advancing direction of the tobacco shreds, and the second needle rollers are driven by a second transmission mechanism and a second driving mechanism to rotate; and the cluster tobacco shred slitting device is adjacent to the long tobacco shred pull-apart device and comprises more than one third needle rollers which are sequentially arranged in the advancing direction of thetobacco shred, and the third needle rollers are driven by a third transmission mechanism and a third driving mechanism to rotate. According to the system, the uniformity of the tobacco shred structure of the thin cigarette is improved, so that the rolling requirement of the thin cigarette is met.

The invention discloses a screening device with the crushing function for beneficiation. The screening device comprises a mounting base, a mounting frame is arranged at the upper end of the mounting base, a crushing tank is arranged on the left side of the upper end of the mounting frame, a motor is arranged in the middle of the upper end of the crushing tank, a rotating shaft penetrating through the bottom of the crushing tank is arranged at the output end of the motor, a mounting plate is arranged on the surface of the rotating shaft, the mounting plate is connected with an ore breaking edge through a pin shaft, a material leaking opening is formed in the bottom of the inner cavity of the crushing tank, the bottom of the crushing tank is provided with a discharging guide groove penetrating through the top of the mounting frame, a screening tank is arranged in the middle of the upper end of the mounting base, a vibrating motor is arranged in the middle of the bottom of the screening tank, a filtering plate is arranged at the upper part of the inner cavity of the screening tank, and a discharging plate is arranged at the lower part of the inner cavity of the screening tank. The screening device is simple in structure and reasonable in design; the ore is crushed and then screened through the screening device, and the distinguishing efficiency is high; and the screening device is an ideal device for distinguishing minerals in size and specification.

The invention relates to a fusion method of a passive millimeter wave image and a visible light image in a human body security inspection. The method comprises the following steps of: determining a low brightness temperature suspicious target such as a metal in an original millimeter wave image and displaying the low brightness temperature suspicious target as a dark color area; Suspicious objectswith high brightness temperature, such as liquid, are displayed as bright color areas, and the suspicious objects in the final fusion image are represented as bright areas; For dark suspicious targets, After wavelet decomposition of millimeter wave image and visible image, In the low-frequency coefficient matrix of the fusion image, the target region comes from the low-frequency coefficient matrix of the millimeter-wave image, the non-target region comes from the low-frequency coefficient matrix of the visible image, the high-frequency coefficient matrix of the fusion image is the high-frequency coefficient matrix of the original visible image, and finally the fusion image is obtained by wavelet transform. For the bright suspicious target, the fixed threshold segmentation is used to obtain the millimeter-wave image suspicious target, and the final fusion image is obtained by region selection fusion: the target region is from the millimeter-wave image, and the non-target region is fromthe visible image.

The invention provides an intelligent toll evasion behavior classification method, a storage medium and a terminal, and belongs to the field of highway billing and charging. The method comprises the steps of collecting historical toll evasion cases and normal passing cases, establishing a toll evasion occurrence and type judgment rule model, and verifying and applying the model. The final fee evasion occurrence and type judgment rule model is associated with a highway station charging system, real-time passing vehicles are monitored, whether fee evasion behaviors occur or not is judged, the fee evasion type is judged, and finally a monitoring result is visually displayed to workers. In the method, a machine learning technology is introduced, machine learning model training is carried out by utilizing extracted features through random forest, LASSO, a decision tree, logistic regression, GBDT and other algorithms in a model, and a corresponding machine learning model for detecting whether toll evasion occurs or not and judging the type is generated; besides, new abnormal features are screened in model application, a new type is manually judged and added, and the learning ability and the application range of the model are improved.