2440 results about "Visual positioning" patented technology

The invention relates to the technical field of intelligent detection in the electronic manufacturing industry, in particular to a PCB warping degree intelligent evaluation system integrating visual positioning and multi-modal sensing, which comprises a multi-modal sensing unit, a visual positioning unit, an intelligent evaluation engine and a closed-loop control unit, the multi-modal sensing unit integrates laser displacement, infrared thermal imaging and strain sensors to acquire three-dimensional deformation, temperature and stress data; the visual positioning unit realizes sub-pixel-level positioning by using a high-resolution industrial camera and a feature point matching algorithm, and compensates vibration errors; the intelligent evaluation engine fuses data based on a time-space synchronization protocol, predicts a thermal deformation trend through an improved multi-modal convolutional neural network, and dynamically adjusts a qualified threshold value; and the closed-loop control unit executes sorting and rechecking according to an evaluation result, and optimizes warping and leveling parameters. According to the system, multi-dimensional accurate detection and intelligent control are realized, the PCB warping degree detection accuracy is effectively improved, the process can be dynamically optimized according to the production working condition, and the equipment fault risk is reduced.

The invention relates to the technical field of visual positioning, and provides a visual positioning method and system for precise connector assembly. Performing multi-mode HDR fusion and distortion correction on the original connector image to obtain a multi-layer fusion image matrix; and performing feature region coarse positioning in combination with a composite convolutional neural network to obtain a connector coordinate set, performing geometric feature fine positioning according to the connector coordinate set and the multilayer fusion image matrix to obtain a key point coordinate set, and performing hand-eye calibration and dynamic compensation on the key point coordinate set to generate a pose instruction. And performing vision-force control hybrid assembly processing on the pose instruction according to the sensor feedback data to obtain assembly completion state data. Through multi-modal image fusion, deep learning coarse positioning, precise geometric registration and dynamic cooperation of vision and force control, the speed, precision and stability of precise connector assembly are improved.

The invention relates to the technical field of laser engraving, and discloses a laser engraving method and system capable of automatically correcting coordinates of a galvanometer and a camera, and the method comprises the following steps: calculating a target conversion coefficient based on a calibration size and a pixel size of an image collected by the camera, and controlling the galvanometer to perform laser etching on a cross mark on the surface of a PCB (Printed Circuit Board), recording origin data of a galvanometer coordinate system; adjusting the position of the camera to enable the cross center of the view of the camera to coincide with the cross mark, and calculating the coordinate offset; coordinate space transformation is carried out on all the points to be machined, and galvanometer marking coordinate data are obtained; after laser engraving is carried out on the galvanometer marking coordinate data, an engraving area image is collected, the position residual error and the size proportion deviation are calculated, laser engraving is carried out again, and a laser engraving result is obtained. The technical problem that the visual positioning coordinate and the galvanometer marking coordinate in the laser engraving system are not consistent is effectively solved.

The invention discloses an AI-based neurosurgery auxiliary robot visual positioning system, and relates to the field of visual positioning, which comprises the steps of deploying and initializing structured light scanning equipment and near-infrared imaging equipment, carrying out multi-modal image acquisition on a surgical area, and carrying out standardization processing, space-time alignment and fusion on the acquired image. A multi-modal visual acquisition and preprocessing module, a three-dimensional tissue model construction and registration module, a visual-anatomical feature recognition and extraction module, an AI auxiliary positioning and path optimization module, an intraoperative dynamic perception and feedback control module and a target position confirmation and instruction output module are constructed. According to the method, high-precision identification and dynamic modeling of a brain tissue structure are realized, real-time identification, path planning and position correction can be carried out on a key anatomical structure in an operation process, the precision, the intelligent level and the intra-operation response capability of neurosurgery operation are remarkably improved, the operation risk is effectively reduced, and the positioning reliability and the automatic control efficiency are improved.

The invention discloses an unmanned aerial vehicle visual positioning flight optimization system based on multi-source data fusion, and belongs to the technical field of unmanned aerial vehicle navigation and control. The system comprises a data acquisition module used for acquiring RGB-D image data and multiple physical data; the environment modeling module is used for constructing a local three-dimensional environment map; the path planning module is used for generating an obstacle avoidance flight path; the instruction generation module is used for converting the obstacle avoidance flight path into an attitude control instruction and an accelerator instruction; the state prediction module is used for obtaining an unmanned aerial vehicle prediction state vector at the next moment; the positioning switching module is used for generating compensated fusion positioning data; and the flight control module is used for executing the attitude control instruction and the accelerator instruction. Through the multi-source data fusion and state prediction compensation mechanism, the problems of positioning interruption and trajectory divergence caused by loss of visual signals and unstable flight caused by sudden change of multi-source coordinate switching in the prior art are solved.

The invention discloses a visual positioning method based on an indoor fine three-dimensional model. The method comprises the following steps: acquiring an indoor original image through a camera, synchronously acquiring the attitude angular velocity of the camera, constructing a first feature data set, and correcting to generate a second feature data set; aligning the second feature data set with a pre-constructed indoor live-action three-dimensional model to generate first position estimation data under spatial boundary constraint; according to the first position estimation data, positioning a feature point region from the dynamic image data acquired in real time, and tracking a point movement track in movement track offset to generate a first feature sequence and the like; according to the scheme, through the technical means of multi-source data fusion, dynamic environment adaptability, a local updating mechanism, path planning optimization and the like, the precision, stability and efficiency of indoor visual positioning and navigation are remarkably improved, and an efficient and reliable solution is provided for positioning and navigation in a dynamic environment.

The invention relates to the technical field of unmanned aerial vehicle navigation and control, and discloses an unmanned aerial vehicle indoor and outdoor seamless navigation method fusing Beidou and visual positioning. Comprising the following steps: S1, synchronously acquiring original observation data of a Beidou satellite navigation system of an unmanned aerial vehicle, an image sequence acquired by a visual sensor and inertial data of an inertial measurement unit; s2, processing the original observation data of the Beidou satellite navigation system to obtain the position of the unmanned aerial vehicle, according to the unmanned aerial vehicle indoor and outdoor seamless navigation method and system fusing Beidou and visual positioning, through a self-adaptive fusion filter module, the fusion weight of the unmanned aerial vehicle and visual positioning is dynamically adjusted according to the Beidou signal quality; smooth transition of indoor and outdoor navigation main sources is achieved, pose jump is effectively avoided, a tight coupling depth fusion algorithm is adopted, the precision and robustness of the system are improved, the continuous and stable flight capacity of the unmanned aerial vehicle in complex indoor and outdoor environments is ensured, and the application scene is expanded.

The invention relates to a cutting path generation system based on visual positioning, cutting equipment and a cutting method. The system comprises a scanning camera control module, an image acquisition and recognition module, an image processing module, a cutting path planning module and a cutting control module. According to the system, by combining a Laplace algorithm and an LBP and GLCM texture extraction technology, the boundary recognition problem under the condition that a pattern and a background are in the same color or low contrast is solved; the path smoothness and the cutting quality are improved by adopting contour optimization and curvature compensation algorithms, the cutting sequence is optimized by combining weighting functions of the area, complexity and overlapping degree, and material deformation is reduced. And motor driving is dynamically adjusted through PID control, and stable control over the speed of the tool bit is achieved. The method has the advantages of being high in recognition precision, high in adaptability, intelligent in path planning and stable in cutting process, the automatic cutting efficiency and the finished product quality of complex patterns or same-color materials are remarkably improved, and the overall cutting precision fluctuation is controlled within + / -2%.

The invention discloses an automatic dovetail welding seam grinding control method based on robot visual positioning, and relates to the technical field of visual positioning. The method comprises the steps that after welding is completed, a dovetail welding seam image is collected and converted into a three-dimensional coordinate through a vision algorithm, and a three-dimensional point cloud is generated; smooth interpolation is performed on the three-dimensional point cloud through a B spline curve method to obtain a parameterized curve, and a control point set is optimized through a genetic algorithm to generate a global optimal polishing path; the polishing robot executes a task according to a path, a tail end sensor collects a real-time path and calculates a deviation value with a global path, and when the deviation value exceeds a preset threshold value, inverse kinematics is triggered to solve and correct the path; the hardness of the dovetail weld is measured through laser-induced breakdown spectroscopy, a comprehensive hardness value is obtained in combination with a matrix hardness database, meanwhile, a visual sensor collects the surface state, and polishing process parameters are dynamically adjusted according to the surface state; after the task is completed, the welding seam angle and flatness are detected. According to the method, the optimal path is generated through visual positioning, and automatic grinding of the dovetail welding seam is achieved.

The invention relates to the technical field of image processing, and discloses a visual positioning system and method for a complex box five-axis numerical control comprehensive processing machine. The system comprises a calibration module for collecting multi-angle images and utilizing mark point calibration coordinates to obtain a parameter mapping table, an extraction module for carrying out edge detection and contour extraction splicing features to obtain a three-dimensional data set, a marking module for calculating the actual position of a box and marking inner cavity machining areas, a recording module for analyzing and determining cutting parameters and deformation of all the areas, and a display module for displaying the cutting parameters and the deformation of all the areas. The processing module generates an obstacle avoidance cutter path and a thin-wall area machining strategy, and the comparison module monitors and compares actual effects to dynamically adjust parameters to obtain compensation data. According to the five-axis numerical control comprehensive processing machine, high-precision visual positioning and intelligent processing parameter optimization of the complex box on the five-axis numerical control comprehensive processing machine are realized, and the processing precision, efficiency and reliability of the complex box are improved.

The invention provides a pathological image visual positioning method and system, equipment and a storage medium, and belongs to the technical field of image recognition, and the method comprises the steps: extracting visual features based on a target pathological image, and determining a semantic feature vector and a knowledge feature vector based on first text description; the target pathological image is a pathological image to be subjected to target area positioning, and the knowledge feature vector is used for representing knowledge information associated with the content of the target pathological image; fusing the semantic feature vector and the knowledge feature vector to obtain a fused text feature; performing cross-modal fusion on the fused text features and the visual features to obtain fused multi-modal features, and obtaining fusion representation based on the fused multi-modal features; and based on the fusion representation, positioning a target area in the target pathological image through a multi-layer perceptron to obtain position information of a bounding box of the target area. The method can improve the capability of accurately and flexibly positioning the pathological image region level.

The invention belongs to the technical field of unmanned aerial vehicle autonomous navigation, provides an unmanned aerial vehicle trajectory planning method and an unmanned aerial vehicle design applying the method, and aims to realize real-time and efficient environmental perception and unmanned aerial vehicle autonomous obstacle avoidance trajectory generation. A group of primitive sets is predefined in a three-dimensional state space to explore the whole search space so as to realize complete coverage of a feasible region, multi-mode perception input of'depth image, current state and target direction 'is adopted, and the depth image is acquired by a depth camera; the current state is obtained by the airborne vision positioning module; multi-modal sensing input is processed by a deep learning network, future expected position, speed and acceleration information is calculated according to output of the deep learning network and serves as input of a bottom layer controller of the unmanned aerial vehicle for trajectory tracking, and finally obstacle avoidance flight in a complex environment is achieved. The method is mainly applied to unmanned aerial vehicle design and manufacturing occasions.

The invention discloses an automatic stacking control system based on visual positioning. The automatic stacking control system comprises a multi-mode visual positioning module, a model building module, a dynamic path planning module, a mechanical arm execution module, an error correction module and a storage module. The multi-mode visual positioning module comprises an RGB camera, a sensor and a laser radar and is used for obtaining three-dimensional space information of goods and a stacking area in real time. The multi-modal visual positioning module is mounted on one side of the stacking area; the model building module is used for building a stacking area, the cargo position and the position of the mechanical arm in a three-dimensional space, when the system is used, information about the position of the stacking area and information about whether new cargoes can be stacked in the stacking area or not can be obtained, then movement at any position in the space can be achieved through the three-axis cooperation mechanical arm, and the three-axis cooperation mechanical arm can move at any position in the space. The application range of the system is widened, and cargoes can be accurately conveyed to the position of a stacking area through the dynamic path planning module and the error correction module.

The invention discloses a semiconductor mechanical arm autonomous navigation and positioning method based on AI vision, and relates to the technical field of mechanical arms, and the method comprises the following specific steps: visual perception and multi-modal information fusion: collecting multi-modal data through configured industrial, depth, light field and polarized light cameras, pre-processing the multi-modal data, and carrying out multi-modal information fusion; realizing information fusion by using a deep learning network containing a feature extraction layer and a fusion layer and an attention mechanism, and outputting a unified visual description; according to the method, a deep learning target detection model is utilized, an attention mechanism is introduced to enhance the attention and extraction capability of small target features, the recognition result is optimized in combination with surface feature information recognized by polarized light imaging, the category and position of a target object can be accurately determined, meanwhile, a mechanical arm motion error model is established, and the recognition accuracy is improved. And the fused visual positioning information is fused with sensor data such as a mechanical arm joint encoder and a gyroscope by utilizing a sensor fusion technology, and a positioning result is optimized and compensated.

The invention relates to a tin ring automatic focusing laser soldering tin fixing and 3D welding spot detecting system, and belongs to the technical field of industrial automatic laser processing. The system is characterized in that a tin ring preparation module identifies pins in real time and dynamically optimizes winding parameters through visual positioning and a self-adaptive winding head, and performs defect identification and pre-repair analysis through multispectral imaging; the soldering tin positioning module is used for switching stations through a rotary table, and precise positioning and clamping of a product are realized by combining laser contour sensing and a focusing compensation algorithm; the laser tin soldering module can visually identify tin materials and automatically call welding parameters, laser power and action time are regulated and controlled in real time by establishing a regional heat conduction model, and gradient cooling is implemented after welding; and the 3D detection module adopts dual-mode scanning, constructs a welding spot three-dimensional model based on fusion data, recognizes microcracks through super-resolution processing, and discriminates the welding spot quality. And automation and intellectualization of the whole process from tin ring preparation to welding spot quality detection are achieved.

The invention belongs to the technical field of augmented reality (XR) and computer vision crossing, relates to fusion application in immersive digital performance, and provides a virtual actor reconstruction and immersive presentation system based on 4D Gaussian splash modeling and XR space positioning. The system comprises a set of spherical multi-camera-position high-synchronization camera shooting matrix used for capturing dynamic images of actors; performing dynamic modeling on the multi-angle image through a 4D Gaussian splashing technology, and outputting a virtual actor point cloud model which can be used by XR equipment; vPS visual positioning and an SLAM tracking module are combined, and precise mapping positioning of a performance space is achieved in AR / MR equipment. The method supports the immersive watching of the actor image at the audience end at a 360-degree free visual angle, and realizes the natural presentation of the virtual actor without dead angles and wearing in cooperation with shielding judgment and a real-time rendering engine. The system is widely applicable to on-site entertainment scenes such as immersive theaters, text travel performances, brand activities, concerts and television programs.

The invention provides a spatial heterogeneous vision positioning system and positioning method of a pineapple picking robot, relates to the technical field of picking robot positioning, and realizes high-precision spatial positioning and attitude estimation of the robot in a complex terrain environment by comprehensively utilizing multi-source information such as vision, inertial measurement, multispectrum and structured light. The positioning stability and the real-time performance are obviously improved; by embedding a multi-modal deep learning model, efficient non-destructive detection and intelligent classification of internal defects of pineapple fruits are realized, a mechanical arm is supported to dynamically adjust a picking strategy, fruits with better quality are preferentially selected for picking, and the picking quality and efficiency are improved; and meanwhile, the preset three-dimensional terrain model and the heuristic search algorithm are combined, the motion path of the robot is effectively planned, it is guaranteed that the robot safely and stably moves in the complex environment, and the mechanical damage risk is reduced.

The invention discloses a material weighing and conveying control method and system, and relates to the technical field of self-adaptive control systems. The method comprises the following steps: setting a visual positioning identifier in an unloading area, and configuring an identification code for a material; establishing an information management database, associating the identification codes of the materials with material attributes, and generating a material association attribute table; collecting an identification code of the material, and calling a material association attribute table based on the identification code; and based on the information management database, obtaining a production line state table of each production line and a warehouse state table of the current warehouse, generating a candidate transportation list, determining a comprehensive priority score of each candidate transportation target point, and selecting the candidate transportation target point with the highest score as a final transportation target point. Compared with the prior art in which fixed rules are used for scheduling, the material transportation targets can be dynamically distributed according to real-time production requirements, manual intervention is avoided, the logistics efficiency is improved, high-requirement materials are preferentially distributed, and production interruption caused by material shortage is prevented.

The invention relates to the technical field of red tide anomaly detection, in particular to a red tide anomaly detection method and system based on an improved multi-mode Transform. The method comprises the following steps: acquiring a remote sensing image and text data; respectively carrying out data preprocessing according to the obtained remote sensing image and text data; performing visual positioning and text selection based on the preprocessed data; performing cross-modal feature learning on the basis of a hierarchical Transform of a multi-modal capsule mechanism; guiding an attention mechanism based on a semantic path to carry out image-semantic feature alignment optimization; and carrying out multi-modal knowledge distillation on the optimized features. According to the method, an image and text preprocessing module, a visual positioning module, a keyword extraction module and other modules are combined, multi-angle accurate perception of a complex red tide scene is achieved, and the bottleneck that a red tide area is difficult to accurately recognize under the condition that data are single and information dimensions are limited in a traditional method is broken through.

The invention discloses a multi-modal multi-scale retrieval enhancement generation method, system and device applied to external knowledge questions and answers and a medium. The method comprises question perception, multi-modal multi-scale query fusion coding, dense recall and answer generation. Analyzing a key query phrase from the question through a fine-tuned instruction language model, and accurately positioning a region of interest corresponding to the phrase in an image by using an open set visual positioning model; multi-source information is compressed and distilled into an optimal query vector through a deep fusion network integrating multi-head self-attention and an information bottleneck theory; executing a maximum inner product search to recall related knowledge; guiding the large language model to synthesize all information to generate a final answer; the system, the equipment and the medium directly perform feature fusion in the vector space based on the method, so that challenges such as information loss and cascading errors caused by a traditional normal form can be effectively dealt with, high correlation and high accuracy of retrieval knowledge are ensured, and accurate and reliable image-text questions and answers are realized.

The invention relates to an intelligent photovoltaic cleaning method for air-ground cooperation of an unmanned aerial vehicle and a cleaning robot, and belongs to the technical field of photovoltaic cleaning. The method comprises the steps that the real-time pose of the unmanned aerial vehicle is determined in a visual positioning and laser positioning fusion mode; establishing a wind speed estimation model by taking the wind speed as a predictable state quantity; solving an optimal control sequence of the unmanned aerial vehicle in a prediction time domain based on a wind speed estimation model so as to compensate wind disturbance in advance; the track of the unmanned aerial vehicle is corrected to achieve accurate butt joint of the unmanned aerial vehicle and the cleaning robot; and an intelligent decision-making method is used for realizing the timely transfer of the cleaning robot by the unmanned aerial vehicle and the cleaning work of the photovoltaic panel of the cleaning robot. Through centralized control of the central platform, a single unmanned aerial vehicle assists multiple cleaning robots, dynamic task allocation and path planning are achieved, and the problems that desert photovoltaic cleaning efficiency is low and the coverage range is limited are solved.

A subway tunnel unmanned aerial vehicle inspection system disclosed by the present invention comprises a flight platform, a depth camera, a binocular camera, a 3D laser radar, a TOF laser radar, a 4K holder, a crack analysis module and a track following flight module, the depth camera is used for infrared track identification, the binocular camera is used for providing a visual positioning function and collaborative path planning, and the 3D laser radar is used for 3D laser radar. The 3D laser radar is used for generating three-dimensional point cloud data of a tunnel environment and performing real-time mapping, the crack analysis module marks cracks based on a crack analysis algorithm when receiving environment real-time acquisition videos or pictures, and the track following flight module controls the course of the unmanned aerial vehicle based on rail features acquired by the depth camera. Through a multi-sensor fusion technology, a high-precision flight control algorithm and an intelligent path planning method, autonomous navigation, obstacle avoidance and inspection tasks of the unmanned aerial vehicle in a tunnel environment are realized, and the method is suitable for scenes of subway tunnel structure health monitoring, equipment state detection, potential safety hazard investigation and the like.

The embodiment of the invention provides a warehouse body intelligent unblocking method and system based on visual identification and air cannon linkage, and the method comprises the steps: collecting images in a warehouse in real time through a plurality of high-definition anti-explosion cameras disposed in the warehouse, carrying out the fusion processing of the collected images in the warehouse, and constructing a warehouse three-dimensional dynamic model; based on the warehouse three-dimensional dynamic model, key features are extracted by adopting a neural network; taking the extracted key features as input, analyzing associated parameters through a time sequence of a long-short-term memory network, establishing a congestion cause knowledge base, and labeling congestion types in a classified manner; constructing a congestion risk grading model based on the congestion cause library and a machine learning algorithm; and activating a corresponding air cannon based on the level corresponding to the three-level early warning and a visual positioning matching result, dynamically adjusting injection parameters, and executing unblocking operation according to an optimized time sequence. According to the embodiment of the invention, closed-loop intelligence from visual perception, reason modeling, artificial intelligence decision making to accurate execution is constructed.

The invention belongs to the technical field of unmanned aerial vehicles, and provides an omni-directional vision positioning system and method and an unmanned aerial vehicle, the omni-directional vision positioning system is installed on a fuselage of the unmanned aerial vehicle, and the omni-directional vision positioning system comprises an image acquisition device used for acquiring a real-time scene image in an omni-directional range in the flight process of the unmanned aerial vehicle; the inertial measurement sensor is connected with the image acquisition equipment and is used for synchronously acquiring motion state data of the unmanned aerial vehicle; the navigation system is connected with the image acquisition equipment and the inertial measurement sensor and is configured to execute at least one of distortion correction processing and feature extraction processing on the real-time scene image to obtain a processed image; and according to the motion state data and the processed image, determining current pose information of the unmanned aerial vehicle in a space coordinate system constructed by taking a take-off point of the unmanned aerial vehicle as an original point. The positioning precision and robustness of the unmanned aerial vehicle in a complex environment are improved, and the requirement of high-precision autonomous flight is met.

The invention provides a robot production line article grabbing method based on visual localization, which comprises the following steps: preprocessing a multi-modal image to obtain an original image; performing grid mapping on the original image to obtain a multi-level feature descriptor; based on the multi-level feature descriptors, a dynamic mapping relation among the three coordinate systems is established through a non-rigid coordinate system alignment algorithm, and coordinate offset of movement of the conveyor belt is compensated in real time; performing hierarchical feature matching on the multi-level feature descriptors and an article template library, identifying article categories and extracting contour geometric features; based on the contour geometric features and the surface curvature distribution, the three-dimensional pose and candidate grabbing points of the object are obtained through a geometric constraint optimization algorithm; and generating a robot obstacle avoidance track according to the candidate grabbing points and the robot motion model. According to the method, accurate coordinate compensation is realized through multi-modal data fusion and dynamic adaptive grid mapping, and the article positioning accuracy is improved in combination with hierarchical feature matching.

The invention relates to the technical field of precise motion control, in particular to an XY motion platform positioning error compensation system and method, which extracts spatial-temporal characteristics through multi-modal data fusion and normalization processing in combination with a neural network hybrid model, and dynamically manages time sequence errors by using a forgetting gate, an input gate and an output gate. And a compensation parameter is updated by adopting an online adaptive training mechanism of error source classification. The problems that in the prior art, due to mechanical abrasion and thermal deformation of an encoder, precision is attenuated, nonlinear errors are difficult to process through a PID algorithm, pure vision positioning is prone to interference and complex in calibration, and an online learning mechanism is lacked can be effectively solved, the positioning comprehensive error is reduced to the micron order, the anti-interference robustness and the real-time compensation capacity of a system are improved, and the system reliability is improved. And the positioning precision and the production efficiency are obviously improved.

The invention discloses an electric power operation supervision method and system based on AI behavior recognition and semantic analysis, and belongs to the technical field of electric power system safety, and the method comprises the steps: arranging a control architecture and collection equipment of an electric power system, and continuously collecting voiceprint data and video stream data; verifying the identity of an operator through voiceprint recognition, carrying out semantic analysis on a voice instruction based on an electric power standard term library, and judging the consistency between the instruction content and a task issued by a system; identifying the position of an operator through visual positioning, performing spatial matching with the interval number of the target equipment, and detecting the normalization of an operation behavior through video analysis; and generating a voice interaction prompt according to instruction fuzziness, triggering hierarchical alarm through multi-modal data fusion, and synchronously recording and storing violation operation data. The power system supervision efficiency is improved, authorization operation is ensured, human errors are reduced, safety is guaranteed through visual monitoring, alarm is triggered through multi-mode fusion, violation is responded in time, evidence is recorded, and traceable safety management is supported.

The invention discloses a multi-modal remote sensing visual positioning method and device based on scene knowledge enhancement and a medium, and relates to the technical field of remote sensing visual positioning. The method comprises the following steps: firstly, preprocessing a plurality of remote sensing images, generating scene knowledge enhanced text description, forming a visual positioning data set, and dividing the visual positioning data set into a training set, a verification set and a test set; constructing a visual positioning model, and obtaining an optimal model through training, verification and testing; inputting a to-be-queried text to obtain remote sensing image coordinates. According to the method, cross-modal fusion of knowledge enhancement is realized, scene knowledge is embedded into a visual positioning framework, and the problem of inference of implicit semantics in a remote sensing scene is solved; multi-scale image features and scene knowledge are fused layer by layer through multi-round cross-modal attention iteration, and semantic understanding from coarse granularity to fine granularity is achieved; a similarity threshold is introduced to screen a high-correlation image region, and background interference is reduced in combination with loss constraints. The LLaMA2 is subjected to efficient fine tuning in combination with the LoRA technology, end-to-end coordinate generation is supported, and both performance and calculation efficiency are considered.

The invention relates to the technical field of image processing, in particular to an image local enhancement super-resolution method based on text prompt. According to the technical scheme, the method comprises the steps of obtaining a low-resolution image and a text prompt provided by a user; and inputting the low-resolution image and the text prompt into a visual positioning model to generate a region-of-interest mask which is used for identifying the position of a target region corresponding to the text prompt in the image. According to the invention, the visual positioning model is driven to automatically identify the region of interest of the image through text prompt, fine reconstruction branches are configured for the region of interest, lightweight reconstruction branches are configured for the background, and global visual consistency is guaranteed in combination with the fusion module, so that the readability and the identifiability of details of the region of interest are remarkably improved; the method is advantaged in that accurate requirements of intelligent monitoring, document OCR, medical image and other scenes are satisfied, calculation resource distribution is substantially optimized, calculation power waste of a background area is avoided, and image super-resolution efficiency and practicality are improved.

The invention discloses a port internal and external fleet positioning system and method, and relates to the technical field of intelligent positioning. The method is used for solving the problems of frequent occurrence of positioning blind areas, sensitive dynamic interference, uncontrollable error accumulation and low anchor point resource scheduling efficiency in a complex scene of a port. Anchor point coverage is dynamically calculated through metal environment signal attenuation characteristics, blind area information is generated in combination with vehicle trajectory clustering, and signal resource allocation is optimized. Ground texture gray scale entropy segmentation and cross-frame Bayesian correlation are carried out in a blind area, interference is eliminated, synchronous pose data are generated, and visual positioning robustness is improved. An error propagation chain model is constructed based on communication topology among vehicles, accumulated errors are corrected by fusing confidence coefficients of multiple vehicles, and the motorcade positioning consistency is enhanced. And finally, analyzing a port scheduling plan and a vehicle motion state, constructing a space-time prediction model, dynamically generating an anchor point instruction, and realizing on-demand scheduling of anchor point resources. The port vehicle positioning precision and the resource utilization efficiency are remarkably improved, and reliable support is provided for port automatic operation.