232results about How to "Increase the speed of positioning" patented technology

The invention discloses a desktop portable sight line tracking method and apparatus. The method comprises the steps of clicking a preset calibration point on a screen by a user to perform calibration, and capturing a facial image of the user by a system; performing pupil localization and Purkinje hot spot localization on the captured facial image of the user, extracting an eigenvector, and standardizing the eigenvector; obtaining a mapping relationship through calibration point data, and substituting the standardized eigenvector of each frame of image into the mapping relationship to obtain actual fixation point coordinates; and extracting fixation point coordinates in all image frames to realize sight line tracking. According to the method and apparatus, an ellipse fitting-based improvement method is proposed for accurately locating the pupil; two Purkinje hot spots are simultaneously located through a clustering-based method; and the method and apparatus have the advantages of quickness and accuracy for localization.

The invention provides a visual SLAM algorithm based on instance segmentation, and the algorithm comprises the steps: firstly extracting feature points of an input image, and carrying out the instancesegmentation of the image through employing a convolutional neural network; secondly, using instance segmentation information for assisting in positioning, removing feature points prone to causing mismatching, and reducing a feature matching area; and finally, constructing a semantic map by using the semantic information segmented by the instance, thereby realizing reuse and man-machine interaction of the established map by the robot. According to the method, experimental verification is carried out on image instance segmentation, visual positioning and semantic map construction by using a TUM data set. Experimental results show that the robustness of image feature matching can be improved by combining image instance segmentation with visual SLAM, the feature matching speed is increased,and the positioning accuracy of the mobile robot is improved; and the algorithm can generate an accurate semantic map, so that the requirement of the robot for executing advanced tasks is met.

The present invention relates to an actuator for generating a rotational positioning movement, wherein the actuator comprises an electric motor (1) and a reduction gear (2) which consists of two assemblies (4, 5). The first assembly (4) is formed by a gear of the harmonic, cycloid or wobble, type and the second assembly (5) by a gear of the single-step or multistep planetary or Wolfrom type, whereby both assemblies (4, 5) are so tuned to each other, that the reduction gear (2) is not self-locking.

The invention discloses an AGV positioning method based on machine vision. The method comprises the steps: enabling an environment to be digitalized through the design of a positioning digital label firstly; obtaining an image comprising a label through a vehicle-mounted visual system; processing the image through an algorithm, and recognizing the position, content and deflection angle of the label secondly; and finally converting the position, content and deflecting angle into the position and posture of a car in an environment through the calibration of a distance relation table of a camera and the related coordinate conversion, thereby achieving the positioning of the car in the environment. The method is different from other AGV positioning methods based on machine vision, and is small in computation burden in a recognition process. Moreover, the positioning precision and speed can meet the requirements of AGV navigation.

A license plate character partitioning method based on a quick region labeling algorithm and a license plate master space location method belongs to the image processing technical field and relates to a license plate automatic recognition technique. Firstly the license plate region is converted through grey level histogram and grey level stretching conversion to realize reinforcement of the character region on the license plate; secondly a two-valued threshold value is calculated and the license plate grey level image is converted into a two-valued image; thirdly a connectivity analysis of the license plate two-valued image is carried out according to the quick region labeling algorithm and an alternate region of characters is obtain through a region growing method; fourthly a master space location is fixed from the license plate two-valued image; fifthly the final character region is obtained through mending and making up for the character region based on the feature of the license plate master space location; finally the characters are partitioned from the license plate grey level image. The license plate character partitioning method based on the quick region labeling algorithm and the license plate master space location method provided by the invention can effectively improve performances such as systematic versatility and location accuracy.

The invention discloses an electromagnetic induction device and a touch positioning method thereof, belonging to the field of electromagnetic handwritten input. The touch positioning method provided by the invention comprises the following steps of: simultaneously gating at least two groups of transmitter coils and receiving coils for scanning, detecting and recording an electromagnetic pen signal; primarily positioning an electromagnetic pen according to the maximum value of the detected electromagnetic pen signal and a coil combination corresponding to the maximum value after all boundary transmitter coils and boundary receiving coils in an antenna plate zone are scanned; and accurately positioning the electromagnetic pen according to the coil combination corresponding to the maximum value of the electromagnetic pen signal if the maximum value of the electromagnetic pen signal is smaller than a preset initial positioning threshold value, or else processing the combination of the receiving coils and the transmitter coils corresponding to the maximum value of the electromagnetic pen signal to confirm the position of the electromagnetic pen. The method provided by the invention can be adopted to rapidly realize the accurate positioning of the electromagnetic pen and can effectively avoid the problems of contact loss, delayed positioning, and the like when the electromagnetic pen rapidly moves.

The invention relates to an active maglev electromagnetic catapult for aircraft carriers. The active maglev electromagnetic catapult comprises a maglev guide rail fixed on a catapult base, a catapult platform sleeved on the guide rail, and a linear motor. The shipboard aircraft catapult platform is supported by active maglev supporting technology and is driven by the high-power linear catapult motor, and catapult-assisted take-off of shipboard aircrafts is achieved. The active maglev supporting technology is introduced, a catapult platform position control system and a speed control system are provided, and accordingly positioning precision and catapulting speed of the catapult platform are increased greatly, electrification is achieved and the service life is prolonged. The active maglev electromagnetic catapult has the advantages of adjustable supporting rigidity and damping, no friction, no need of lubrication, no environment pollution and the like.

The invention discloses a jar cover surface defect detection method based on machine vision. The method comprises the following steps that: obtaining a jar cover image, and calculating the center of gravity coordinate (Xg, Yg) of the jar cover image; obtaining jar cover edge points, and adopting a three-point random circle detection method to carry out fitting on the jar cover edge points to obtain a fit circle, wherein the fit circle is a jar cover area; dividing the jar cover area into four areas, and independently radially unfolding four areas; obtaining the dimension features of a center area, carrying out classified identification on the basis of a preset center area classifier to obtain the defect detection result of a center area; and independently obtaining the dimension features of an embedded ring area, a glue injection area and an edge curling area, and on the basis of a preset embedded ring area classifier, a glue injection area classifier and an edge curling area classifier, carrying out classified identification to obtain the defect detection results of the embedded ring area, the glue injection area and the edge curling area. The method is quick in implementation andaccurate in positioning, an error problem caused by the irregular texture information of the center area is overcome, and the robustness of the method is better.

The invention relates to an airborne SAR image automatic target positioning method, which convert the coordinates of a target in the SAR image into an actual geographic coordinates of the target on the earth by utilizing the flight information (location, speed, etc.) of an aerial carrier and the SAR image information (resolution, center distance, central Doppler frequency, etc.). Based on a range-Doppler equation, the invention combines an earth model equation to carry out an target location, which not only can realize an absolute localization without a control point, but also considers the influence of the earth model on positioning accuracy; on the other hand, Taylor series expansion is used for carrying out the linearization of nonlinear equations; an iterative method is adopted to solve the equation, thus improving the positioning speed; meanwhile, the initial value point of an iterative computation is appropriately selected according to the operation mode of radar, thereby controlling a converging direction in the iteration and avoiding a target confusion. As a positioning method without the control point, the automatic target positioning method of the invention has the advantages of high accuracy and high speed, thereby being capable of carrying out a real-time processing and suitable to various airborne SAR systems.

The invention discloses a network fault positioning method and a device. The method comprises the following steps: establishing a propagation model; processing time slice information; filtering noise; confirming an assumption set; calculating the reliability; and updating prior probability. The invention has low arithmetic complexity, fast positioning speed, strong anti-noise capability and high positioning accuracy degree.

The invention relates to the field of a picking end effector, in particular to a large-tolerance strawberry picking end effector capable of picking up multiple strawberries once based on a hand-eye system, and a working method thereof. The invention aims to provide a strawberry picking end effector which can pick up clustered fruits, has low demanding degree on picking positioning of a mechanical arm, and a working method thereof. According to the picking end effector and the working method, the strawberry fruit region (single grain or adjacent graininess mature strawberry fruits) is taken as the picking objects; the separation of adhering mature fruits is not necessary; the algorithm complexity and calculation time consumption of identifying and positioning the target fruits can be reduced; the image degree coordinate of the target fruits can be exactly converted into a world coordinate according to an internal reference and an external reference of a camera, and the positioning precision and speed can be improved; many strawberries can be picked up once; under the condition that the mature fruits are close or adhered, the picking efficiency can be greatly improved; the picking end effector has large-tolerance; under the condition that the positioning precision on the mechanical arm is not demanding, the picking success rate can be improved.

The embodiment of the invention discloses an information processing method and device, equipment and a storage medium, and the method comprises the steps: determining an image feature set of a currentphotographing interface and the current position of electronic equipment; obtaining a target anchor point map corresponding to the current position, wherein the target anchor point map at least comprises a historical image feature set and corresponding historical anchor point information; matching the image feature set of the current shooting interface with a historical image feature set in the target anchor map to obtain a target image feature set; and displaying historical anchor point information corresponding to the target image feature set in the current shooting interface.

The invention discloses a hydraulic control system and a device used for the location of a ship body by sections. Three hydraulic cylinders that are mutually orthogonal are adopted; wherein, the piston rod of a hydraulic cylinder with a displacement sensor moves up and down; the other two hydraulic cylinders push the hydraulic cylinder to move front and back as well as left and right on a horizontal plane. The moving displacement on the horizontal plane of the hydraulic cylinder the piston rod of which moves up and down is monitored and recorded by a displacement sensor; the recorded data is input into a PCL control circuit, thereby further controlling the expansion amount of the other two hydraulic cylinders to be capable of ensuring the hydraulic cylinder the piston rod of which moves up and down to make translational motion on the horizontal plane. The system and the device of the invention can adjust and control the precise positions of up and down, front and back as well as left and right of the substances arranged at the hydraulic cylinder the piston rod of which moves up and down. Location with high precision can be carried out on the ship body by sections according to the free combinations of three or multiple devices.

The invention discloses an indoor positioning and mobile track monitoring system and a method thereof. No extra infrastructure is needed; and with a WiFi signal in wide coverage in a building, an indoor positioning system is designed and realized in an Android smart phone. The system is mainly applied to specific occasions like a large shopping mall or a supermarket or in a large indoor environment like a library or an airport. Therefore, a user can find his/her location conveniently and find out an exit or a bathroom conveniently. A good foundation is laid for building a platform to realize a new positioning system, updating algorithm researches, and developing the position service. On the basis of WiFi-position-fingerprint-based indoor positioning, relevant factors that include an AP position, the number, a collection time delay, and processing ways after collection and influence the positioning precision are studied. A reasonable method for improving the positioning precision is provided; the correlated AP strength processing algorithm is improved; reasonability of the original processing signal is enhanced; and the positioning speed and the accuracy are improved.

The invention discloses a terminal location method and device. The method comprises: acquiring a measurement report (MR) uploaded by a terminal; forming a cell vector by using a cell identity (ID) of each measurement cell, and forming a level vector by using a coverage level value of each measurement cell; acquiring a cell ID of a serving cell from the MR, and searching multiple training samples in a base station, to which the cell ID of the serving cell belongs, from an MR sample library; determining training samples corresponding to training cell vectors, which have the same cell IDs, in the multiple training samples as to-be-compared training samples; determining probability values of each level vector and a training level vector of each to-be-compared training sample through a neural network model; and determining training position information corresponding to a to-be-compared training sample, the probability value of which is larger than a probability threshold, to be the position of the terminal. According to the cell ID of the serving cell, the training samples of the base station, to which the serving cell belongs, are found out, and speed and efficiency of terminal location are improved.

The invention provides a positioning and tracking method for nodes in a wireless sensor network. The method is applied to the wireless sensor network. The wireless sensor network comprises an upper computer, a base station, the anchor nodes, the beacon nodes and the blind nodes, wherein the number of the anchor nodes is larger than or equal to 3, and each anchor node comprises a positioning module. The method includes the following steps that the anchor nodes are positioned by themselves through the positioning modules and send positioning information to the surrounding environment, wherein the positioning information includes coordinates of the anchor nodes; the beacon nodes are positioned through the positioning information sent by the anchor nodes and send positioning information to the surrounding environment, wherein the positioning information includes coordinates of the beacon nodes; a state equation model and an observation equation model are established for the blind nodes according to the positioning information sent by the beacon nodes or the anchor nodes; the positions and the speeds of the blind nodes are calculated according to the state equation model and the observation equation model. The method is simple in network deployment, accurate in positioning and tracking and wide in application range.

The invention relates to the radio frequency identification (RFID) indoor positioning technology, specifically a RFID indoor positioning algorithm based on deep deterministic policy gradient (DDPG). The indoor positioning algorithm comprises the following steps: establishing an action network and an assessment network, wherein the action network comprises an action estimation network and an actiontarget network; the assessment network comprises an assessment estimation network and an assessment target network; performing single-step updating on the policy gradient by using an action-assessment method, wherein the policy gradient can be applied to a continuous action to perform screening; adding deterministic in the screening process, and outputting an action value on the continuous action, thereby determining a location of the target label. Since the RFID indoor positioning action is continuous, the DDPG and the RFID indoor positioning are combined, thereby well solving the positioning continuity problem. Compared with the traditional indoor positioning algorithm based on a neural network , the algorithm is more continuous on the positioning action, the positioning precision is further improved, and the algorithm is especially suitable for the condition that the lable information is huge.

The invention discloses an RFIC indoor-positioning method based on particle swarm optimization (PSO). The RFIC indoor-positioning method is used for three-dimensional positioning of an article with an RFID electronic tag. According to the method, M reader-writers and N reference tags are deployed in an indoor environment in advance, and the reader-writers and the reference tags are distributed on the three-dimensional space in a grid mode; the distance between the electronic tag and some reader-writer is measured according to tag signal receiving strength and a wireless signal propagation model, and distance-measuring errors are reduced through the reference tags; after the distances between the electronic tag and the reader-writers are acquired, an error accumulation function can be obtained, the PSO is used for solving the reference value when the error accumulation function obtains the minimum value, and in other words, the spatial position coordinates of the electronic tag are obtained. Compared with a traditional method, the selecting process of the reference tags is optimized, measurement of parameters of the wireless signal propagation model is optimized, the PSO is introduced to solve the position coordinates of the electronic tag, and under the conditions that no hardware devices are added, the positioning speed is increased, and the positioning accuracy is improved.

The invention disclose a method for positioning vertebras and intervertebral discs, comprising the following steps of: determining the rough positioned zones of the vertebras and intervertebral discs in a positioned image; generating a vertebra edge image according to the rough positioned zone; determining the horizontal position of each intervertebral disc on the vertebra edge image; and determining the horizontal position of each intervertebral disc according to the vertebra edge image and determining the upper edge positioning line and the lower edge positioning line of each vertebra and the positioning line of each intervertebral disc. Meanwhile, the invention also discloses a device for positioning the vertebras and the intervertebral discs. By adopting the method and the device in the invention, the positioning speed and precision of the vertebras and the intervertebral discs can be improved.

The present disclosure relates to a foreign language teaching evaluation information generating method, a foreign language teaching evaluation information generating device, an electronic apparatus, and a storage medium. The method includes the following steps that: foreign language test information is outputted, video signals collected by a video capture device are detected, video picture information and audio information in the video signals are extracted; the lips of a user in the video picture information are positioned, and the mouth shape features of the lips of the user are captured; pronunciation information corresponding to the foreign language test information is matched with information in a pre-established foreign language test information and pronunciation information model, wherein the pronunciation information contains standard pronunciations and standard mouth shape features; and the first feature deviation value of the deviation of the mouth shape features of the lipsof the user from the standard mouth shape features which are obtained by means matching is calculated, and teaching evaluation information is generated according to the first feature deviation value.According to the method of the invention, the evaluation of foreign language teaching is realized on the basis of the positioning and comparison of the mouth shape features in the foreign language test information.

The invention discloses positioning equipment, a positioning base station, and a space positioning system and method. The positioning equipment comprises a housing and a photosensitive sensor which is arranged on the housing. The technical scheme of the embodiments of the invention is featured in that the position of the positioning equipment relative to the positioning base station is determined according to three laser scanning signals; and as the precision of laser measurement is millimeter-scale and the positioning speed is millisecond-scale, the positioning precision and the positioning speed are greatly improved compared with the prior art, and the technical effect of precisely and rapidly achieving space positioning is realized.

The invention discloses a rapid visible satellite selection method based on a multimode GNSS receiver. The method comprises the steps that (1) elevation angles, received by the multimode GNSS receiver currently, of all visible satellites are calculated, the satellites with the satellite elevation angles smaller than 10 degrees are eliminated, and a usable visible satellite set S is obtained; (2) according to satellite position information of the satellites and receiver preliminary position information, a coefficient matrix and a weight coefficient matrix are calculated; (3) a GDOP value corresponding to the visible satellite set S is calculated; (4) a contribution value delta Gi of each visible satellite to GDOP is calculated and the contribution values delta Gi are ranked from large to small, so that a set SG is formed; (5) when the multimode GNSS receiver needs to select N visible satellites for navigation and location, the front N satellites in the set SG are selected, and a corresponding visible satellite subset is a selected usable satellite constellation. The method has the advantages of being convenient and quick to operate, simple in implementation, high in accuracy, wide in application range and the like.

The invention relates to a precision mechanism light driving technology, which comprises a composite outrigger which is formed by a set of piezoelectric ceramic sheet with two electrodes and a base chip, a hinge, a reflecting lens and frame, and a supporting frame. The lens frame is connected with the base chip by the hinge; the base chip is fixed by the supporting frame. When the positive end of the power is paralleled with the two sets of the electrode area, the negative end of the power connects with the base chip, so that the composite outrigger generates a bend deformation; when the positive end and the negative end of the power are separately connected with the two sets of the electrode area, the composite outrigger generates a torsion deformation. The composite outrigger can also be formed by four cruciform symmetrical distributing outriggers with the reflecting lens in the center of the cruciform symmetric or by two vertical outriggers with the reflecting lens on the vertical point; the lens frame is connected with the supporting frame.

The invention discloses an upper pull rod fault detection method based on deep learning, and belongs to the technical field of upper pull rod fault detection. According to the method, the problems ofhigh possibility of errors and missing detection and low detection efficiency when the existing method is used for detecting the faults of the upper pull rod are solved. The method comprises the following steps: firstly, acquiring a train image to be detected, and acquiring a region-of-interest image from the acquired image; carrying out enhancement processing on the region-of-interest image to obtain an enhanced region-of-interest image; after the image is segmented, outputting the brake cylinder part positions in the segmented sub-images through an improved SSD model; extracting a to-be-identified image containing the upper pull rod according to the position of the brake cylinder; and finally, carrying out fault identification on the pull-up rod in the image to be identified by adoptinga FastRCNN model and a Unet model. The method can be applied to upper pull rod fault detection.