123results about How to "Achieve real-time tracking" patented technology

The invention relates to an interaction method and system based on an augmented reality technique, and a terminal. The method comprises following steps that: a server receives a first graph and interaction information uploaded by a first terminal; a second terminal obtains a second image as real scene, extracts second image information as an identified object in real time and sends the second image information to the server; the server searches the interaction information related to the second image information from the preset database as virtual information and returns to the second terminal; the second terminal superposes the virtual information and the identified object in the real scene and displays as augmented reality information; the second terminal receives the operation of a user to the displayed augmented reality information and sends the operated information to the server. The interaction method based on the augmented reality technique of the invention enables the user to learn related information of the photographed object omni-directionally from various angles and enables the interaction information obtained by the user to have imagery; and the user experience is great.

The invention discloses a rapid analysis method for genomes of pathogenic microorganisms and a data analysis platform. The method provides typing and source-tracing functions of bacterial genome sequences based on a bacterial strain database system. A user can only set simple parameters and upload bacterial strain genome sequences. The system can feed back the MLST parting result of bacterial strains, medicine resistance and distribution of virulence genes within the shortest time and make reference to names of bacterial strains and provide phylogenetic trees of all bacterial strains which are uploaded by a user and arranged in a database. The method is suitable for the establishment of a data analysis platform for pathogeny microbiology which integrates study of genetics, genomics and phylogenetics. Compared with the conventional typing technology, the rapid analysis method and system for genomes of pathogenic microorganisms have the following beneficial effects: higher resolution is obtained; bacterial strains of the same clone can be accurately detected; the feedback speed of results is high; and the technology of efficient reference strain setting and genome SNP sequence mapping, data analysis results can be quickly obtained by a user, thereby facilitating real-time tracking and fast source-tracing of clonal spreads in pathogenic microorganisms.

The invention relates to a method for applying a RFID technology and a sensor technology to the supervision and the control of safety equipment, which comprises the following concrete steps of: (a) reading safety equipment adhered with specific electronic tags at real time by utilizing a radio-frequency identification (RFID) monitoring and managing system installed in public places and finishing the real-time monitoring of the working state of the safety equipment; (b) if the monitoring and managing system can not effectively identify the due tag number in a RFID read-write area of the monitoring and managing system, informing an alarm emergency treatment platform of starting working by a background service system; (c) if tag information does not conform to the current state information of the safety equipment in the processes of maintenance, overhaul and check, triggering a corresponding alarm by a RFID read-write unit and informing a manager of carrying out treatment; and (d) if the detection value of a sensor in the public places is abnormal, showing the existence of safety accidents or hidden danger; and when conforming to an alarm condition, issuing an alarm by the background service system and starting an alarm emergency treatment plan. The invention has the advantage of simple and convenient installation, debugging and maintenance.

The invention relates to a GPS-based anti-collision early warning system for switching operation, which comprises a plurality of locomotive terminals, a plurality of station terminals and at least one monitoring management center, wherein the at least one monitoring management center comprises a locomotive real-time dynamic database, a vehicle real-time dynamic database, a track database, a database maintenance unit, a data transmitting unit and a data receiving unit; each locomotive terminal comprises a GPS positioning unit, a locomotive position reporting unit, an analytical unit and a warning unit; and each station terminal comprises a plan execution reporting unit. The invention also provides a corresponding anti-collision early warning method. By the real-time positioning data of the GPS positioning unit on each locomotive terminal and the real-time data of execution conditions of a switching operation plan and a loading and unloading operation plan, which are reported by each station terminal, the invention dynamically updates a plurality of databases, thereby realizing anti-collision early warning for the switching operation.

The invention relates to a method used coherence optical fiber grating set to realize train positioning and real time tracking. It includes the following steps: using wavelength division multiplex n coherence optical fiber grating sets with resin protective layer to cling to the rail inside; the interval of the optical fiber grating is about 200-300m which shortens common track circuit blocking area by ten times; when locomotive drives to the its coherence optical fiber grating set sensitive area, it will produce amplitude vibration with corresponding wave; one end of the optical fiber is connected multiple wavelength laser launcher in the control center by circulator of which another end is connected with fiber amplifier, then connected with the wave separator of which separated wavelength is transformed into electric signal; all electric signal are input into the digital signal processing terminal to exactly figure out train exact position, speed, and acceleration to realize exact positioning and real time tracking.

The invention discloses a remote real-time tracking method for a well path, which relates to the technical field of real-time tracking of a wellbore path of a petroleum drilling well. The method of the invention comprises the following steps: remotely transmitting the wellbore path (LWD / MWD) measured in real time on the field, real-time measured data when drilling, well top coordinate data, well deflection data, stratum data, lithological profile data and stratigraphic dip data to a base server, reading the data from a database of the base server, generating a real-time well deflection path tracking chart according to an established mathematical model, and issuing synchronously through Web. The invention excludes the influence of adverse factors of complex geologic structure, obscure stratigraphic dip, thin objective target reservoir and the like, can accurately shoot the target point according to design, and obviously improves the exploratory development effect of a horizontal well.

The invention discloses a GLONASS phase inter-frequency bias real-time tracking and precise estimation method based on particle filtering. The method comprises the following steps that step 1: data preprocessing: the satellite ephemeris and the pseudo range observation value and the phase observation value of the current ephemeris cloud are imported; step 2: a double difference pseudo range and carrier phase observation equation is established, and a single epoch normal equation or a previous epoch accumulation normal equation is obtained after linearization; step 3: the phase bias in the normal equation is corrected according to the particle value to solve the normal equation; and the particle filtering weight is updated, and the value of the phase inter-frequency bias rate and the particle mean square root are calculated; and step 4: the steps 1-3 are repeated to perform real-time tracking or precise estimation of the value of the phase inter-frequency bias rate; and phase inter-frequency bias correction is performed through the value of the phase inter-frequency bias rate after filtering convergence, and the integer ambiguity is fixed so as to realize precise estimation. Real-time tracking of the inter-frequency bias and GLONASS precise positioning without receiver type difference can be efficiently realized.

The invention provides a vehicle real-time tracking method and device based on a monitoring video, and relates to the technical field of intelligent transportation, and the method comprises the steps:constructing a video network according to a plurality of video points; determining a reachable network of each video point according to the video network; determining a travel time interval of the target vehicle according to the reachable network; determining monitoring videos corresponding to the plurality of video points according to the reachable network in the travel time interval to performfeature extraction of the target vehicle; and when the features of the target vehicle are extracted, switching to the corresponding monitoring video for display. According to the invention, a topological relation graph between different video points is formed by constructing a video network, the travel time interval of the target vehicle is determined according to the video network, the subsequentvideo points are started timely to identify the target vehicle, rapid positioning of the vehicle is completed, basic data and a video AI technology are combined, rapid and stable tracking of the real-time trajectory of the vehicle is realized, and the vehicle real-time tracking method and device have a wide application prospect in urban management.

The invention discloses a corner reflection unit and a control method. The corner reflection unit comprises a corner reflection plate, an omnidirectional antenna, a signal receiver, a motor driving device, a motor controller and a motor. The omnidirectional antenna receives a wireless signal to be tracked through the corner reflection plate; the signal receiver feeds back the strength and the direction information of the wireless signal to be tracked to the motor controller according to the received wireless signal to be tracked; and the motor controller drives the motor according to relevant information, so that an output shaft drives the corner reflection plate to rotate around the omnidirectional antenna for a corresponding angle. Because a closed-loop control method is formed between the signal receiving end of the corner reflector and the driving device which controls the corner reflector to rotate, the signal can be fed back in time when the corner reflector is at different positions or in different angles, so that the appropriate adjustment to the optimal signal receiving position is met, the effects that the system cost is lower and the tracking performance is improved are achieved, and the overall performance of the system is improved.

The invention discloses a robot system for spraying anti-pollution flashover coatings on substation insulators and a spraying method. The robot system is characterized by comprising a robot body and aspraying tool, wherein the robot body comprises a lifting mechanism, and the spraying tool is arranged at the front end of the lifting mechanism; the spraying tool comprises a supporting mechanism, asurrounding mechanism and at least one spraying mechanism; the surrounding mechanism is connected with the supporting mechanism and can carry out vertical lifting movement relative to the supportingmechanism; the surrounding mechanism comprises an annular inner chamber and an annular component which can move along the annular inner chamber, and an annular opening is formed in the lower part of the annular inner chamber; and at least one spraying mechanism is arranged under the supporting mechanism, is connected with the annular component and can move along with the annular component along the annular opening. The robot system can realize uniform spraying for the insulators.

The invention provides a device and method for comprehensively and accurately positioning and monitoring a railway maintenance place. The device comprises a personnel positioning subsystem, a video monitoring subsystem and a central monitoring and analyzing system, the personnel positioning subsystem comprises a reader-writer, a receiving main station and more than one positioning label laid along the direction of railway station tracks and trenches, each positioning label is wirelessly connected with an input end of the reader-writer, an output end of the reader-writer is connected with an input end of the receiving main station, the receiving main station is connected with the central monitoring and analyzing system, and the video monitoring subsystem comprises a video collecting device, a machine room server and a videoing device. Each of operating personnel in an operating area can be positioned accurately in realtime, realtime tracking of operating tracks of the operating personnel is realized, and monitoring video when the operating personnel is at different position points is automatically spliced background; through combination and linkage of positioning and video, the operating personnel can be monitored more comprehensively.

The invention provides a method and system for tracking an underwater robot in real time, and unmanned relay equipment, and relates to the technical field of marine equipment. The unmanned relay equipment comprises the following parts of an underwater acoustic communication device for communicating with the underwater robot; an underwater robot positioning device for positioning the underwater robot equipped with an acoustic beacon; a satellite communication positioning device for satellite communication with a mother ship or a shore-based accusation tower, and for calculating real-time latitude and longitude; an inertial navigation device for providing a three-degree-of-freedom attitude angle and three-degree-of-freedom axial acceleration information; an autonomous control unit for receiving azimuth and distance data of the underwater robot calculated by the underwater robot positioning device, receiving the motion state data provided by the satellite communication positioning deviceand the inertial navigation device, calculating a navigation trajectory according to each received data and performing thrust distribution; and a power and propulsion device for receiving thrust distribution information and propelling the unmanned relay equipment to sail according to the navigation trajectory.

The invention relates to a pedestrian re-identification and tracking method based on spatio-temporal context, which comprises the following steps: training a Mask RCNN network; Using the trained MaskRCNN network to process the original picture set, the training set, the test set and the search set. Training convolution neural network with training set; The test set and the search set are processed using the trained convolution neural network to obtain a first preset number of pictures from the test set for reidentifying the target pedestrian. The invention uses the object detection algorithmand the instance segmentation algorithm to preprocess the picture, removes the background interference information, further improves the model precision, and improves the accuracy of the pedestrian re-identification method. At that same time, the invention solve the problem that the current pedestrian re-identification algorithm lacks the tracking function, and proposes the area prediction algorithm based on the walking speed, and combine the Mask RCNN to reduce the tracking calculation complexity, achieve real-time tracking, and improve the tracking efficiency.

The invention discloses a positioning method for a craniomaxillofacial surgery. The positioning method comprises the following steps of: first, producing an occlusal splint referring tracer according to a plaster model of dentition of a patient; secondly, fixing an occlusal splint in an oral cavity of the patient, performing CT (Computed Tomography) scanning on the patient to obtain CT scanning image data, and segmenting and reconstructing the CT scanning image data through CT image reconstruction software to obtain a triangular mesh model of soft and hard craniomaxillofacial tissues of the patient and the occlusal splint referring tracer; and thirdly, arranging a standard fixing reference frame at the head of the patient, reading a transformation matrix M2 of the occlusal splint referring tracer in a three-dimensional coordinate system of the standard fixing reference frame, solving a transformation matrix M of the standard fixing reference frame in the three-dimensional coordinate system of the head, solving a transformation matrix Mf of a light-reflecting ball array of a tool in the three-dimensional coordinate system of the head, tracing and navigating the light-reflecting ball array of the tool relative to the three-dimensional coordinate system of the head in real time, and positioning the light-reflecting ball array of the tool.

The invention discloses a particle-filtering-based real-time tracking and precise estimation method of a phase inter-system bias of a GNSS. The method comprises: step one, data preprocessing is carried out and a satellite ephemeris value, a current epoch pseudo-range observation value and a current epoch phase observation value are inputted; step two, intra-system and inter-system observation value equations are established and linearization is carried out to obtain a single-epoch normal equation or an epoch-accumulative normal equation; step three, a GNSS observation value normal equation inter-system bias is corrected, the normal equation is calculated and a particle filter weight is updated, and according to a weighted particle, a numerical value of a fractional part of a phase inter-system bias and a particle mean square root are calculated; and step four, the previous three steps are repeated, the value of the fractional part of the phase inter-system bias of the GNSS is tracked in real time to realize real-time tracking; or the previous three steps are repeated, a phase inter-system bias value is estimated precisely and the intra-system and inter-system double-different full-cycle fuzziness values are fixed to realize precise positioning. Therefore, real-time tracking and precise estimation of the phase inter-system bias of the GNSS are realized efficiently; and indistinctive multi-mode GNSS precise positioning of the system is realized.

The invention discloses a ship multi-target tracking method based on a YOLO V5 algorithm, and the method comprises the steps: carrying out the screening and marking of collected ship image data, building a data set, and dividing the data set into a training set, a verification set, and a test set; training the YOLO V5 network by using a training set and a verification set in the self-established data set to obtain a ship detection model based on the YOLO V5 network and a weight file; detecting the test set by using the trained YOLO V5 detection model, outputting a detection result, and evaluating the detection model; based on the trained YOLO V5 detection model, generating a tracking model through DeepSort algorithm processing; and carrying out real-time verification on the generated DeepSort tracking model. According to the invention, marine ship detection and multi-target tracking can be realized, the detection precision is high, the real-time performance is good, and the speed is fast.

The embodiment of the invention provides a mobile terminal as well as a tracking method and a tracking system thereof and solves the problem of tracking the position of a user. The tracking method comprises the steps of: sending and storing request information containing the position of the mobile terminal into a server after entering an emergency mode; receiving authentication information fed back by the server, wherein the authentication information is generated after the server receives the request information; and sending the authentication information to a pre-set receiving party, so as to enable the pre-set receiving party to log in the server to look up the position of the mobile terminal by the authentication information. The mobile terminal as well as the tracking method and the tracking system of the mobile terminal can be used for realizing the real-time tracking on the position of the user.

The invention discloses a real-time control method based on an intelligent device camera and an intelligent device, which relate to the intelligent control field. When students study independently, the problems encountered by students are difficult to be solved in time, which has a great impact on students' learning efficiency. The invention relates to a real-time control method based on a cameraof an intelligent device, which comprises the following steps: the camera is turned on in real-time, and an object image is obtained in an image capturing area; When the voice signal of the user is collected, speech recognition is performed on the voice signal to recognize the problem raised by the user; Performing image recognition on the target image to identify the content of the user-designated area on the book; Combining the questions posed by the user and the contents of the designated area on the book, searching for answers to the questions posed for the contents of the designated areaon the book; Displays answers to questions about the contents of a specified area on a book. When the students study independently, the problems encountered by the students can be solved in time, andthe learning efficiency of the students is increased.

The invention discloses a face projection method for a facial makeup based on face features, and the method comprises the steps: building a facial makeup dermatoglyphic pattern database; obtaining a video frame comprising a human face, and carrying out gray processing; obtaining all Haar characteristics of all positions on a gray scale image, and solving characteristic values; carrying out the training of the characteristic values through employing an Adaboost algorithm, and obtaining the optimal characteristic point of each position; adding the characteristic points to an active appearance model, calculating the coordinates of a nose tip of the human face, a plane deflection angle of the human face and the depth deflection angle of a side surface of the human face, and building a human face model; selecting a needed facial makeup dermatoglyphic pattern, selecting a facial makeup dermatoglyphic pattern control point, carrying out the fitting of the facial makeup dermatoglyphic pattern, and obtaining a fitted curve; carrying out the deformation of the fitted curve, and obtaining a facial makeup model through combination; adjusting the projection angle according to the coordinates of the nose tip of the human face, the plane deflection angle of the human face and the depth deflection angle of the side surface of the human face; and projecting the facial makeup model onto the human face model. The method is flexible and variable, is simple and quick, and enables the facial makeup to be projected onto the human face accurately in real time.

The invention relates to a tracing transmission mechanism suitable for a solar concentrating system. A centring round rail is arranged at the bottom of a horizontal rotating disc, a driving friction wheel and driven friction wheels are movably connected with the centring round rail, and a pitching mechanism is arranged above the horizontal rotating disc. The tracing transmission mechanism further comprises a friction wheel height adjusting device connected with the driving friction wheel. Friction between the driving friction wheel and the centring round rail is used for driving the horizontal rotating disc to rotate, and therefore the motor is prevented from burning out; the pitching mechanism is used for adjusting the orientation of a solar concentrating mirror; the friction wheel height adjusting device prevents the driving friction wheel from falling off the centring round rail by adjusting the height of the driving friction wheel; the horizontal rotating disc is used for adjusting the horizontal direction of the concentrating mirror to trace the azimuth angle of the sun, the pitching mechanism is used for adjusting the vertical angle of the concentrating mirror to trace the elevation angle of the sun, and by means of the cooperation of the horizontal rotating disc and the pitching mechanism, sunlight is radiated perpendicularly to achieve photovoltaic conversion at the maximum efficiency.

An internet of things perception method for furniture flexible production is provided; the furniture enterprise mechanization and industrialization level in our nation needs to develop; the internet of things perception method for furniture flexible production comprises the following steps: raw material storage information acquisition and process; material use and material cutting information acquisition and process; board material deep processing information acquisition and process; sorting and packaging information acquisition and process; transfer warehouse information acquisition and process; quality examination and reworking repair information acquisition and process. In a parts processing process, the method uses a digital mode to store information, thus reducing manual errors, and improving production efficiency; in a storage management process, the cargo warehouse in / out information can be updated in real time, thus automatically monitoring the stock, saving massive tedious examination and inventory works, and saving cargo inventory time.

The invention discloses a numerical control machine tool workpiece machining allowance determination method based on image recognition, and the method comprises the steps: obtaining image data of a to-be-machined workpiece, carrying out the feature recognition of the image data, determining the workpiece shape features in the image data, and determining a reference 3D model corresponding to the workpiece shape features based on the workpiece shape features, the reference 3D model is used for reflecting the workpiece shape of the to-be-machined workpiece after machining is completed; acquiring laser point cloud data of the to-be-processed workpiece, and determining a real-time 3D model of the to-be-processed workpiece based on the laser point cloud data; and according to the real-time 3D model and the reference 3D model, the machining allowance of the workpiece to be machined is determined. According to the invention, the machining allowance can be tracked in real time, and the workpiece to be machined can be machined conveniently.

The invention discloses a method for implementing one-to-one real-time logistics. The method comprises the steps of setting an order-releasing account as a consignment account and an order-receiving account as a shipper account, receiving a logistics order released by the consignment account, and pushing logistics order information to a matched shipper account; if the shipper account releases a relay carrying order, pushing relay carrying order information to other matched shipper accounts; after a delivered command is received, pushing order-delivered information to the consignment account; and after delivered confirmed information is received, ending an order process. The invention also discloses a system for implementing one-to-one real-time logistics. The system comprises an order receiving-sending module, a relay carrying order judgment module, a relay carrying order receiving-sending module, an order delivery module and an order delivered confirmation module. Extremely-rapid, on-time and reliable services are provided for a user by using a compact car as the shipper through crowd-sourcing sharing and relay transportation modes, the resource utilization rate is improved and the maximization of logistics transportation efficiency is realized.

The invention belongs to the field of revolution cabinet inventory verification and positioning systems. A revolution cabinet real-time inventory verification and positioning system is characterized in that an access window is formed on a front panel of a revolution window, an RFID antenna packaging box is clamped between transverse columns on two sides of the revolution cabinet, the RFID antenna packaging box is internally provided with an RFID antenna, the RFID antenna is connected with feedback lines, and the feedback lines are connected with a reader on the RFID antenna packaging box; the revolution cabinet is internally provided with a transmission device, and the transmission device drives bucket plates in the revolution cabinet to move around the RFID antenna along an arc track; articles in the bucket plates are provided with electronic labels; the RFID antenna is corresponding to the access window; and when the articles with the electronic labels right face the RFID antenna through windows on back sides of the bucket plates, the reader finishes inventory verification and positioning. The revolution cabinet inventory verification and positioning system is simple in structure, the radio frequency antenna is parallel to the bucket plates, and barrier-free communication between the antenna and the article labels is realized.

The invention relates to a low-complexity 5G NR channel estimation method based on compressed sensing, and belongs to the field of wireless communication. The method comprises the following steps: solving a channel time domain information pair by using a compressed sensing theory, and performing Fourier transform to obtain corresponding frequency domain information to calculate a channel frequencydomain autocorrelation matrix RHH; quickly estimating a signal-to-noise ratio SNR by using the extracted pilot signal; according to the characteristics of a cyclic matrix, solving an LMMSE estimationmatrix, introducing a Toeplitz matrix, and calculating a pilot frequency LMMSE channel estimation value based on compressed sensing. The method starts from traditional LMMSE channel estimation, the characteristics of the compressed sensing theory and the cyclic matrix are combined, the frequency band utilization rate is increased, and meanwhile the complexity is reduced.

The invention discloses a method and system for identifying a target object in a region range. The method comprises the following steps: determining a pre-divided recognition region where the target object is located; activating a plurality of image recognition sensors within the recognition area; receiving image data acquired by the plurality of image recognition sensors for the same target object at the same time; constructing three-dimensional image data according to the plurality of acquired images; analyzing the three-dimensional image data to obtain a first type of identifier for identifying an identity; and identifying the target object in real time in the images acquired by the plurality of image identification sensors according to the first type of identifier. According to the method and the system, the action track of the target object in the identification range can be obtained, the real-time tracking of the target object can be realized through splicing of different identification ranges, and the difficulty of dynamic target object image identification is solved.

The invention provides an equipment state tracking method and a device thereof. The method is applied to a first proxy server. The first proxy server is in communication connection with the acquisition equipment; wherein the acquisition device is in communication connection with at least one acquired device, the acquired device comprises at least one data point, and the method comprises the following steps: acquiring a data packet communicating with the acquisition device, the data packet carrying real-time state data of the at least one data point of the acquired device acquired by the acquisition device; under the condition that the equipment state is normal, analyzing the data packet; updating the historical state data of the data points by using the real-time state data; and under thecondition that the equipment state is abnormal, discarding the data packet, and updating the state into an abnormal state. Under the condition that the equipment state of the acquired equipment is abnormal, the data packet can be directly discarded, and the data packet can be analyzed under the condition that the equipment state is normal, so that the network communication pressure is reduced, thewaste of computing resources is avoided, and the data processing efficiency is improved.

The embodiment of the invention provides a message processing method, apparatus and device. The method comprises the following steps: receiving a service instruction, and generating a corresponding service track log according to the service instruction, wherein the service track log is used for recording the life cycle of the service instruction; parsing the track log, and determining a service track identifier in the track log; judging whether the service track identifier exists in a preset subscription service identifier database; and if so, converting the track log corresponding to the service track identifier into a to-be-pushed message, so as to push the to-be-pushed message to a client subscribing the service track identifier based on the CometD service. According to the message processing method, apparatus and device, the message is directly delivered to the subscription client without falling to the ground, and the time interval from data generation to data presentation is shortened; and accurate positioning is achieved according to the service track identifier, a tracking service processing track is displayed according to the track log, the processing speed is improved, manual waiting is shortened, and the effects of real-time tracking and simultaneous tracking multiple client interfaces are achieved to improve the user experience.

The invention discloses a method for identifying a polystyrene product. The method comprises the following steps: preparing a radio-frequency identifying label of the product; implanting the radio-frequency identifying label into the polystyrene product during injection molding of the polystyrene product; and arranging a radio-frequency identifying system at production and marketing, logistics or application places of the polystyrene product, and reading the radio-frequency identifying label with a reader of the radio-frequency identifying system to identify the polystyrene product. The invention also discloses the polystyrene product, the identifying system and a bicycle race information system with the identifying system. The method for identifying the polystyrene product improves the identifying efficiency, accuracy and instantaneity, and improves the judgment efficiency and fairness of a bicycle race.

The embodiment of the invention provides a human eye sight line tracking method. The human eye sight line tracking method comprises the steps: obtaining human eye feature samples of a target human eye at a plurality of preset sight line positions at a current position; according to the human eye feature samples of the preset sight line positions at the current position and human eye features, obtained in advance, of multiple preset sight line positions, located at the reference position, of the target human eye on the screen, obtaining a function relation between the human eye feature samples of the preset sight line positions at the current position and the human eye features at the reference position; and obtaining target human eye features of the target human eye at an unknown sight line position on the screen at the current position, obtaining human eye features at a reference position corresponding to the target human eye features according to the function relationship, and obtaining the unknown sight line position according to the human eye features at the reference position corresponding to the target human eye features. According to the embodiment of the invention, real-time tracking is realized, and the tracking precision is high.