137 results about "Coordinate difference" patented technology

The invention relates to the technical field of image state recognition, in particular to an unhooking and rehooking AI accurate recognition grabbing system and method based on pattern recognition, in the system, node construction is conducted through contour changes, boundary difference values and gray level dynamic states of a hook assembly in an image sequence frame, edge displacement accumulation analysis is combined, meanwhile, through a graph neural network, an image sequence frame is obtained, and the image sequence frame is obtained. Cosine values and coordinate difference values between nodes are subjected to combined comparison, and a path hopping sequence is constructed, so that the response sensitivity to state abrupt change is enhanced, a key path of morphological evolution can still be stably extracted under the condition of complex background interference or local shielding, the anti-interference performance and fault tolerance of space path identification are effectively improved, and the space path identification accuracy is improved. Statistical modeling is further carried out on state rate sudden change points through a hidden Markov model, paragraph merging and invalid fragment removing operation are carried out on abnormal point segments by matching a standard state mode, a state label sequence is constructed, and accurate division of high-confidence and multi-segment continuous states is achieved.

This disclosure describes systems, methods, and devices related to object detection in images. A device may input an image, representing an object, to a manual labeling learner system; identify, using the system, first coordinates of an upper left corner of a bounding box representing the object based on a heatmap indicative of a probability of the first coordinates representing the upper left corner; identify, using the system, second coordinates of a bottom right corner of the bounding box based on the first coordinates and a first distance regression map indicative of coordinate differences between the second coordinates and ground truth coordinates input to the machine learning model as training data; generate, using the system, adjustments to the first coordinates and the second coordinates based on a second regression map; and generate, using the system, the adjusted first and second coordinates, the bounding box.

The invention, which belongs to the technical field of artificial intelligence, discloses a multi-modal cross-view-angle air-ground cooperation big model system comprising a heterogeneous agent module, a multi-modal data processing module, a cross-view-angle fusion module, a big language model reasoning module and an action decision module. The heterogeneous intelligent agent module comprises an unmanned aerial vehicle, a ground robot and an unmanned vehicle, the unmanned aerial vehicle is provided with a high-resolution camera, an infrared imaging device, a laser radar and a meteorological sensor, the ground robot is provided with the laser radar, a camera and a force sensor, and the unmanned vehicle is provided with a camera, a radar and an inertial measurement unit. The multi-view features are fused through a cross attention mechanism, precise mapping from the features to real coordinates is realized in combination with a ScaleToRealClip model, and scale and coordinate differences of air and ground data are eliminated; meanwhile, the cross-view fusion module can dynamically update environment coordinates, help an intelligent agent to perceive temporary obstacles in real time, and construct a comprehensive and accurate environment cognition model.

The invention relates to the technical field of resolution improvement, in particular to a video definition improvement method and system.According to the video definition improvement method and system, through window accumulation of adjacent frame gray difference and coordinate difference, minimum displacement is selected, mapping is updated, boundary neighborhood weighted correction is carried out, cross-frame dislocation is reduced, and motion area connection is guaranteed; capturing a three-frame gray scale sequence to calculate a difference value, writing a replacement value according to a monotone and amplitude condition, inputting the replacement value and an alignment frame difference into Fourier transform, decomposing amplitude and phase distribution in a residual field in a frequency domain, and performing amplitude and phase consistency elimination on an abnormal block during frame-by-frame superposition to obtain an abnormal block; row difference and column difference absolute values of neighborhood blocks in the residual iteration texture matrix are accumulated into energy, the energy and a center difference are combined with a threshold value to output a mark, and a texture region and an edge region are distinguished; the brightness of the low-frequency position is updated according to the weight mean value, the maximum difference and the direction difference are combined for the high-frequency position, the edge gradient is extracted through the convolutional neural network, a correction value is generated, edge steps and fuzziness are suppressed, and high-frequency details are continuously kept.

The invention discloses a target measurement method and device, electronic equipment and a storage medium, and relates to the technical field of holders. The method comprises the following steps: acquiring a shot image, wherein the shot image comprises a shot target; determining a device parameter and a pixel coordinate difference value, and calculating a correction factor by using the device parameter and the pixel coordinate difference value; and determining a target rotation angle by using the correction factor, and controlling the holder to adjust based on the target rotation angle. Compared with the prior art, the method has the advantages that the correction factor is calculated only through the basic parameters of the image and the equipment under the condition that modes such as laser ranging do not need to be introduced, the measurement precision is improved compared with a mode of trigonometric function conversion, errors caused by lens distortion, projection deformation or environmental factors are reduced, and meanwhile, the hardware cost is also reduced.

The invention provides an intelligent ward nursing information acquisition and processing method and system. The method comprises the following steps: acquiring mattress pressure data to construct a pressure grey-scale map sequence, calculating a pressure barycentric coordinate difference value between continuous frames to generate a patient movement vector set and a body barycentric movement path, extracting a continuous pixel region exceeding a pressure threshold in a single-frame image as a high-pressure region, and counting the duration of the high-pressure region; calculating gradient distribution by using the pressure data, decomposing the gradient distribution into direction and intensity data, and generating a pressure topological graph through dual-channel fusion; integrating the bending characteristics of the moving path, the time distribution characteristics of the duration of the high-pressure area and the spatial gradient characteristics of the pressure topological graph, and constructing a body position dynamic space-time diagram with the central point of the high-pressure area as a node and the moving distance and angle of the node as edges; when it is recognized that the high-risk static body position features exist in the image, a turning-over reminding process is triggered, and a control instruction is generated. According to the invention, accurate recognition and automatic nursing response of high-risk body positions are realized.

The invention provides a positioning method and device and a storage medium, and the method comprises the steps: a first cloud server obtains environment information uploaded by a terminal, carries out the environment reconstruction through the environment information, and obtains a map under a global coordinate system, the global coordinate system is a position coordinate system of map data after environment reconstruction; in response to starting of the terminal, the first cloud server obtains a first coordinate of the terminal in a local coordinate system every preset time interval, and determines a target coordinate difference value corresponding to the first coordinate in a global coordinate system; and the terminal receives the target coordinate difference value sent by the first cloud server, and corrects the first coordinate based on the target coordinate difference value to obtain a second coordinate positioned by the terminal. According to the method and the device, deviation accumulation or positioning loss of the positioning position is avoided, the positioning precision is improved, the situations of mode crossing and drifting in the virtual picture are further avoided, a user does not need to position through a computer backpack, and the user experience is improved.

The invention relates to the technical field of laser ranging, in particular to a novel railway tunnel deformation real-time surveying and mapping system based on a three-dimensional scanning technology, which comprises a laser sampling module, an echo evaluation module, a space repositioning module, a residual evolution module and a deformation surveying and mapping module. According to the method, after continuous distance information of a target position is obtained, measuring point sequence data is constructed based on a time sequence, a signal abnormal area is recognized through echo intensity fluctuation, the signal abnormal area is further converted into a space coordinate structure, track differences are extracted, vector data are generated, deviation characteristics are analyzed in combination with the tunnel space position, and the deformation amplitude is quantified. Compared with a periodic observation mode depending on manual distribution of measuring points, the method has the advantages that fine coordinate difference data can be continuously output, the deformation monitoring frequency and the space coverage range can be improved, manual interference errors can be reduced, and the dynamic mastering capability of the tunnel structure evolution process can be enhanced.

The invention relates to the technical field of limb rehabilitation training, in particular to an artificial intelligence-based cognitive impairment assessment and intervention training system, which comprises a path construction module, an action acquisition module, a matching recognition module, a rhythm regulation and control module and a cognitive intervention module. According to the method, the standard path is constructed in the virtual task and combined with the three-dimensional action data of the user to form the path section response result, so that the spatial integrating degree of the action of the user and the target path can be identified, and quantitative description of the action execution stability and continuity is realized through coordinate difference analysis; a cooperative path record of task execution is obtained in combination with a trajectory coverage rate and a matching relationship, and a trend analysis and parameter regulation mechanism is introduced, so that a task pushing rhythm and a user response rhythm are dynamically matched, and tracking correction of an intervention effect in a training stage and fine evaluation of cognitive limb disorder performance are realized. The analysis granularity of cognitive state and action collaboration is improved, and the mapping precision between the intervention content and the user execution characteristics is enhanced.

The invention discloses a target detection method and device based on visual identification. The method comprises the following steps: acquiring relative coordinates between an image acquisition part and a detection surface of a detection object; when the relative coordinate is not consistent with the reference coordinate, determining a first coordinate difference value; when the relative coordinate is not within the coordinate error range, adjusting the image acquisition component until the relative coordinate is consistent with the reference coordinate; when the relative coordinates are consistent with the reference coordinates, controlling an image acquisition component to perform image acquisition on the detection object to obtain a plurality of images; determining second relative coordinate differences between the plurality of images and the reference image, and selecting a target image; and performing target detection according to the target image. The technical problem that in the prior art, a machine vision system adopts depth-of-field limitation of a fixed camera, so that all targets cannot be ensured to be within the depth-of-field range of the camera when the heights of different models of recognized targets are changed, and recognition cannot be possibly caused is solved.

The invention provides a network RTK base station coordinate precision detection method and system, and the method comprises the steps: carrying out the base station networking based on a Delaunay triangulation algorithm, and determining a baseline list; obtaining satellite precision orbit information, selecting a to-be-solved baseline vector from the baseline list, and resolving the to-be-solved baseline vector by adopting a fast static positioning algorithm to obtain a preset high-precision relative coordinate result; the preset high-precision relative coordinate result is converted into coordinate values of all reference stations, the coordinate values of all the reference stations are compared with preset known coordinate values in a system, coordinate difference values of all the reference stations on all the base lines are obtained through calculation, gross error elimination is conducted based on the coordinate difference values, and processed coordinate difference values are obtained; and integrating the processed coordinate difference values of each reference station on the plurality of base lines, performing consistency analysis and statistical judgment, determining the reference station with abnormal coordinates in combination with a preset threshold value, and outputting a diagnosis alarm result. According to the invention, the reliability and precision of the network RTK service are effectively guaranteed.

The invention discloses an unmanned aerial vehicle positioning method, device and system during signal interruption, and a storage medium. The method comprises the following steps: receiving a broadcast ephemeris broadcasted by a satellite, observation value data and a real-time orbital clock error correction number; judging whether the real-time orbital clock error correction is interrupted or not; if the real-time orbital clock error correction is interrupted, performing short-term forecasting on the correction by establishing a quartic polynomial; combining the correction number with the broadcast ephemeris to generate a real-time precise ephemeris and a real-time track; if the predicted correction number is used, an observation equation is established through epoch difference to settle a coordinate difference; if the received correction number is used, an observation equation is established through a PPP model, and the real-time coordinates of the unmanned aerial vehicle are settled. By adopting the technical scheme of the invention, high-precision positioning of the unmanned aerial vehicle during short-term signal interruption is realized.

The invention discloses an obstacle distance measurement method and system based on panoramic tiled display. Four-way or six-way fisheye cameras are installed on a vehicle, images around a vehicle body are collected, 3D rendering splicing is carried out based on calibration data (internal parameters / external parameters), and a 360-degree panoramic aerial view image is generated. A user clicks two points of a screen through a display terminal, a system converts screen coordinates into 3D normalized vertex coordinates through viewport inverse transformation, relative coordinate differences are calculated and then expanded into homogeneous coordinates, the homogeneous coordinates are converted into physical coordinate differences in combination with a pre-stored rendering transformation matrix inverse matrix (generated by model / projection / view transformation combination), and the 3D normalized vertex coordinates are converted into 3D normalized vertex coordinates. And finally calculating the actual physical distance between the two points. The system supports dynamic refreshing of a ranging result and renders a distance marking line on a panoramic image. According to the method, through multi-camera calibration, coordinate space mapping and a matrix transformation algorithm, interactive obstacle distance measurement in a vehicle-mounted panoramic display scene is realized, and the method is suitable for intelligent driving, vehicle auxiliary driving and other scenes.

The application provides a kind of based on remote sensing image's directed lane center line generation method and system, method includes: the remote sensing image data is input into trained directed lane center line segmentation model, obtains the position prediction image and direction prediction image of lane center line;Line segment detection method is used to obtain line segment information from the position prediction image, the actual geographic coordinates corresponding to line segment are obtained using the coordinate information of remote sensing image data;The RGB data in the same place of the direction prediction image with the coordinate position of the line segment is extracted, and the coordinate difference (dx, dy) between the end and the beginning of the line segment is calculated according to the RGB data;According to the coordinate difference (dx, dy), the direction information of the current line segment is calculated;Iterate position prediction image and direction prediction image, according to the coordinate information and direction information of each line segment, adjust the head-tail connection relationship of each line segment, and obtain directed lane center line data.The application reduces the production cost of lane center line data.

The application provides a zero position calibration method and device of a camera, electronic equipment and a storage medium, and relates to the technical field of camera shooting. The camera comprises a main camera and at least one auxiliary camera. The method comprises the following steps: when the main camera is at a first target zero position, controlling the main camera to shoot a first image; determining a first target position in the first image; controlling the auxiliary camera to rotate to a second target zero position based on the first target position and a preset position difference value corresponding to the auxiliary camera; the preset position difference value is a coordinate difference value between the first target position and a second target position; the second target position is a target position in a second image shot by the auxiliary camera and located in the first image. The application realizes the calibration of the second target zero position of other auxiliary cameras based on the first target zero position of the main camera. The calibration of the zero position of each auxiliary camera does not require the participation of an optical interrupter, thereby saving cost and reducing the probability of hardware failure.

The invention discloses a rapid whole-network coordinate resolving method for a large-scale GNSS (Global Navigation Satellite System) reference station network, which comprises the following steps of: dividing a GNSS observation station into a plurality of sub-networks, and applying loose constraint to each sub-network to resolve and generate a clock error and a track file; independently carrying out PPP floating point calculation on each subnet by using the clock correction and the track file; combining the PPP floating point solution results of all the subnets, calculating the PPP solution of the whole net, and obtaining the coordinates and covariance matrixes of all the stations under the original framework; an independent baseline set is constructed, a relative coordinate difference vector is generated, and the independent baseline set converts the station coordinate vector into a baseline difference vector through a coefficient matrix; and introducing a known coordinate of the reference station under the new reference frame, and reversely deducing coordinates of other stations of the whole network based on the relative coordinate difference vector and the known coordinate of the reference station to complete reference frame conversion. According to the method, the independence and robustness of the conversion process are enhanced, and the coordinate precision and reliability of a large-scale network under a unified reference framework can be ensured.

The present application relates to the technical field of image state recognition, in particular to a kind of based on pattern recognition's picking and replacing hook AI accurate identification and capture system and method, in the present application, the outline change of hook assembly in image sequence frame, boundary difference and gray dynamic are carried out node construction, and combined with edge displacement accumulation analysis, while through graph neural network, cosine value between nodes and coordinate difference are compared jointly, construct path jump sequence, to enhance the response sensitivity of state mutation, under the condition of complex background interference or local occlusion, still can stably extract the key path of morphological evolution, effectively improve the anti-interference and fault tolerance of spatial path recognition, further statistical modeling is carried out on state rate mutation point by hidden Markov model, by matching standard state mode, paragraph merging and invalid segment rejection operation are carried out to abnormal point section, construct state label sequence, realize the accurate division of high confidence, multi-section continuous state.

The application provides a flight control method and system for unmanned aerial vehicle (UAV) to perform waypoint tasks based on differential coordinates, which comprises: north calibration: determining the angle between the self-built coordinate system and the NED coordinate system by capturing a mark line pointing to the north direction from an arbitrary origin, and obtaining the parameters required for coordinate conversion; establishing a conversion relationship between the self-built coordinate system and the NED coordinate system based on the parameters required for coordinate conversion, so that the waypoint coordinates in the UAV waypoint flight task, i.e. the coordinates in the self-built coordinate system, are output as the coordinates in the NED coordinate system, and can completely reflect the relative positions of the coordinate points in the self-built coordinate system, so that the path is completely consistent with the flight path required by the waypoint task; based on the new coordinates in the NED coordinate system after conversion, the UAV can obtain the relative positions of the waypoints, so as to obtain the flight control instructions based on the coordinate differential operation, and perform the fixed-point flight.

The invention provides an image bright field correction method and device for a laser plasma accelerator, electronic equipment and a storage medium, and relates to the technical field of X-ray imaging. The method comprises the following steps: acquiring dark fields and initial X-ray beam spot images of an X-ray bright field calibration detector and an X-ray imaging detector to obtain coordinates of an initial X-ray beam spot center; under the condition that the imaging sample is moved in, a real-time unprocessed X-ray beam spot image obtained by the X-ray bright field calibration detector and a real-time unprocessed projection image obtained by the X-ray imaging detector are obtained, so that a real-time beam spot image and a real-time imaging image are obtained, and coordinates of a real-time X-ray beam spot center are obtained; according to the coordinate difference between the real-time X-ray beam spot center and the initial X-ray beam spot center, the X-ray beam spot center is obtained, and a final real-time projection image is obtained. According to the invention, the accuracy of bright field correction is improved by measuring the X-ray beam spot in real time and combining geometric space mapping and coordinate transformation.

The present application relates to the field of construction engineering and geological monitoring technology, and discloses a vertical deformation monitoring method and system based on three-dimensional scanning and differential pressure level gauge. The method comprises the following steps: collecting point cloud data of a building structure in real time by using a three-dimensional scanner, and collecting vertical displacement data of the building structure in real time by using a differential pressure level gauge; obtaining the vertical coordinates of the point cloud by using a three-point registration algorithm, and calculating the vertical coordinate difference value of the point cloud to obtain a first difference value; calculating the average of the first vertical displacement data of the current monitoring period and the average of the second vertical displacement data of the previous monitoring period, and calculating the difference value between the average of the first vertical displacement data and the average of the second vertical displacement data to obtain a second difference value; and weighting the first difference value and the second difference value according to the standard deviation of the point cloud data and the standard deviation of the vertical displacement data, and calculating the final vertical deformation of the building structure by using a weighted average method. By implementing the present application, the cost of manual measurement is reduced, and the defect of single data source is changed.

The embodiment of the invention provides a mining area vehicle attitude calculation method and system based on satellite positioning, and the method comprises the steps: defining the map direction of rail transit and the positive direction of a satellite antenna, setting the coordinate system output parameters and attitude calculation basic parameters of the satellite antenna, and calculating the attitude of a vehicle when a train travels in the positive direction of the map; recording relative position parameters of the satellite antenna, and storing the calibrated reference course angle; acquiring real-time coordinates of the main antenna and the auxiliary antenna, calculating a real-time course angle based on a coordinate difference, and smoothing the real-time course angle by adopting a filtering algorithm to obtain a stable attitude angle; carrying out difference calculation on the attitude angle and the reference course angle to obtain an angle deviation, and judging whether the train turns around or not according to the angle deviation and the duration time; and outputting the vehicle attitude data according to a preset period, and carrying out validity verification on the output vehicle attitude data. According to the invention, the limitation of traditional single threshold judgment is solved, and the turn-round identification accuracy is improved.

A defect in an edge region of a wafer is detected with high sensitivity. A defect detection device (50) comprises: a data input interface (51) configured to acquire parameters used for defect detection; an image input interface (52) configured to acquire a captured image of a wafer; and a control unit (53) configured to analyze the acquired image based on the parameters and detect a defect in the wafer. The parameters include a threshold for detecting a protruding defect and a defect determination level.The control unit (53) extracts a rectangular area from a captured image, containing a ring-shaped image that shows an edge of the wafer, as an image of a region of interest (ROI); generates a binarized image by binarizing the ROI image such that a higher luminance value is set as a first value and a lower luminance value is set as a second value, based on the threshold for detecting a protruding defect; acquires image coordinates corresponding to an inner circumference of an area represented by the first value in the binarized image as inner circumference coordinates; calculates degrees of variation of irregularities in a circumferential direction of the inner circumference coordinates as coordinate difference values; and detects the defect by comparing the coordinate difference values ​​with the defect determination degree.

The invention discloses a real-time wire diameter detection method and system for multiple metal welding strips, and relates to the technical field of photoelectric measurement. The method comprises the following steps: enabling the multi-metal welding strip in linear motion to pass through a constant backlight illumination area to generate a high-contrast projection; the linear array CCD is controlled to perform continuous line scanning in a scanning mode in the vertical movement direction to obtain image frames; determining left and right edge coordinates of sub-pixel precision based on a gray gradient edge extraction algorithm; and calculating a line diameter measurement value according to the edge coordinate difference value and the calibration coefficient, and performing real-time filtering. The system correspondingly comprises a backlight illumination module, an image acquisition module, an edge processing module and a wire diameter calculation module. Reflection interference is eliminated through backlight illumination, the linear array CCD is used for high-speed scanning to adapt to the motion state, the precision is improved in combination with a sub-pixel algorithm, and the problem of high-precision real-time detection of the wire diameter of the metal welding strip under the high-speed condition is effectively solved.

The invention relates to the technical field of GIS (Geographic Information System) modeling, in particular to a digital terrain resource GIS model construction method, which comprises the following steps of: determining space complexity based on the confusion degree of all data points in the point cloud of each tree; based on the distance from all data points in any cylindrical area of the point cloud of each tree to the mass center of the cylindrical area and the coordinate difference of all data points, the point cloud wave disorder degree is determined in combination with the space complexity; on the basis of the distance between each tree in each region and each tree in the neighborhood of each tree and in combination with the point cloud wave disorder degree, determining the tree distribution density so as to adjust the grid window size of each region; and constructing a GIS model of the target forest region based on the rasterized three-dimensional point cloud. According to the method, the size of the grid is adjusted in a self-adaptive manner, so that the construction precision and efficiency of the digital terrain resource GIS model are improved.

A computer vision vehicle locating fusion method includes: receiving an instant driving image from a camera, extracting multiple image features from the instant driving image and the multiple image features being extracted with the pre-stored feature sets in the storage device to fuse an inertial measurement parameter and the pre-stored satellite locating coordinate corresponding to one of the pre-stored feature sets that best matches the instant driving image to generate a first candidate coordinate; using the one satellite measurement coordinate received from a satellite locating device as a second candidate coordinate; calculating a first difference between the first candidate coordinate and an estimated reference coordinate, and calculating a second difference between the second candidate coordinate and the estimated reference coordinate; and determining and outputting the first candidate coordinate or the second candidate coordinate that has the smaller difference with the estimated reference coordinate.

The invention provides an automatic slope rill width extracting and grading method and system based on point cloud data, and relates to the technical field of soil erosion. The method comprises the following steps: firstly, acquiring three-dimensional point cloud data covering a rill region to be detected, and establishing a complete slope point cloud and a corresponding space coordinate system through control point registration, denoising and cutting; constructing a digital elevation model as a rill boundary verification reference according to the slope point cloud; in the slope surface point cloud, calculating the coordinate difference of adjacent points in the cross slope direction and the vertical direction to obtain a first error threshold value for determining point pairs in the same cross section; for the point pairs meeting the first error threshold value, calculating the coordinate difference in the down-slope direction, comparing the coordinate difference with a preset rill width threshold value, and extracting rill boundary points; and based on the extracted rill boundary points, verifying the spatial consistency with the digital elevation model, calculating the average width of each rill, and outputting a width grading result. According to the invention, errors generated during manual measurement can be greatly reduced; and the measurement precision of the width of the rill is improved.

The invention provides a video coding method and device, equipment and a storage medium, and relates to the technical field of computers, in particular to the fields of artificial intelligence, computer vision, video coding and the like. The specific implementation scheme is as follows: reading a to-be-coded current frame image and a to-be-coded reference frame image; respectively extracting feature points and descriptors from the brightness components of the current frame image and the reference frame image to obtain a first type of feature points and descriptors corresponding to the current frame image and a second type of feature points and descriptors corresponding to the reference frame image; performing feature point matching according to the first type of feature points and descriptors and the second type of feature points and descriptors to obtain coordinate differences of a plurality of feature matching points on the current frame image and the reference frame image, and taking the coordinate differences as motion vectors; and performing video coding according to a coding mode determined by the direction consistency of the motion vectors.

This invention provides a method, apparatus, and computer program product for establishing table topology relationships, relating to the field of electronic digital data processing technology. By identifying the vertex coordinates of cells, selecting matching points, calculating the coordinate difference, and comparing it with length and width error thresholds, the invention identifies the relationships between cells in the row and column directions. It then sorts the cells within rows or columns using the coordinate values ​​of specified marker points, directly dividing rows and columns and reconstructing spatial relationships based on geometric rules. This reduces performance degradation caused by excessively large image sizes, does not rely on the integrity or clarity of table borders, and can adapt to complex table scenarios involving borderless tables and tables with numerous empty rows and columns, while significantly improving processing speed.

The invention relates to the technical field of optical distortion correction, in particular to an OCT optical distortion correction method based on deep learning, a medium and equipment. The method comprises the steps of inputting a distorted calibration plate image into a target convolutional neural network to obtain a corrected image, generating a correction pixel difference according to a pixel coordinate difference value of calibration points of the distorted calibration plate image and the target convolutional neural network, obtaining a scanning position correction difference through coordinate conversion, and further converting the scanning position correction difference into X-direction and Y-direction driving compensation voltages so as to correct an optical galvanometer driving voltage in an OCT imaging unit. According to the technical scheme, the multi-layer convolutional neural network is introduced, the adaptive convolutional layer, the pooling layer and the full connection layer are included, parameters are optimized through supervised learning, distortion features are automatically learned, a nonlinear mapping relation is established to generate the correction pixel difference, and the correction precision and efficiency are improved. The method realizes automatic accurate correction, solves the problems of insufficient accuracy and the like in the prior art, avoids manual intervention, simplifies the process and reduces the calculation complexity.

This application relates to a method, apparatus, and storage medium for determining the state of an active pen. The method includes: when an active pen has established a communication connection with a terminal and movement of the active pen is detected, determining the current coordinate difference information of the active pen within the current time period; the current coordinate difference information includes the coordinate difference between the position coordinates of two adjacent moments within the current time period; determining the current velocity of the active pen within the current time period based on the current coordinate difference information; if the current velocity is less than or equal to a preset velocity threshold, determining the total displacement of the active pen within the preset time period based on the current coordinate difference information and the preset time period; wherein the preset time period is located within the current time period and includes the current moment; and determining the target pen state of the active pen based on the total displacement and the preset displacement threshold. This method can improve the anti-interference capability and recognition accuracy of active pen state recognition.