94 results about "Shape matching" patented technology

The invention discloses a radioactive measurement data processing method for improving uranium exploration efficiency. The method comprises the following steps: standardized data acquisition: integrating a multi-parameter module, carrying out time-space synchronous acquisition of geological geophysical environment parameters, standardizing protocol alignment data, and supporting three-dimensional modeling; a three-dimensional coupling model is used for processing interference in a sub-module mode, LiDAR-DEM is used for correcting gamma attenuation in the terrain, an optical fiber thermopermeability instrument is used for correcting daughter errors in the hydrology, signal attenuation of a borehole is compensated through a drilling fluid chart, and a space-time continuous interference field is formed; performing dynamic equilibrium coefficient inversion: constructing equilibrium coefficient isoparametric mapping, and inputting rock core, logging and geochemical data into a random forest model; the transfer learning is trained by using historical data, and dynamic inversion and updating of a new area are carried out; intelligent data processing three-dimensional visualization: self-adaptive denoising and spectral shape matching reinforcement abnormity are carried out; constructing a three-dimensional model, and carrying out transfer learning to optimize the precision; the WebGL platform integrates multi-parameter display, virtual drilling functions and high-dimensional data intuitive interpretation.

The invention provides an intelligent boxing method and system. The intelligent boxing method comprises the steps of obtaining boxing task information and generating a boxing scheme. The boxing task information comprises the size, the weight, the stacking limitation, the number and the priority of each cargo, and the size, the bearing capacity and the occupation state of each loading unit. And under the condition that the size, bearing and stacking constraint conditions are met, a plurality of candidate boxing combinations are generated, the loading utilization rate, the residual space distribution and the shape matching degree are evaluated based on a preset optimization target, and an optimal scheme is selected. The system comprises an input terminal, a task information receiving module, a data storage module, a task and scheme storage module, an operation processing module and a final boxing scheme generation module. According to the method, multi-dimensional constraint matching and weight balance are considered, and space waste and transportation risks are avoided; the maximum greedy algorithm and the adaptive large neighborhood search optimization are combined to realize the combination of automation and experience. The method can be expanded to batch task scheduling and grouping sequencing, so that the calculation complexity is reduced, and the overall boxing and transportation efficiency is improved.

The invention discloses an industrial image anomaly detection method and device, equipment and a storage medium, and relates to the field of image recognition. Obtaining a defect-free product graph, making a comparison data set based on the defect-free product graph, and performing pre-storage processing; performing feature point shape matching on the comparison data set based on the received to-be-detected image and the extracted feature point data, and determining a comparison image of the to-be-detected image; taking the to-be-detected image as a target object, correcting a product image area in the image, and aligning the product image area with the product image area of the to-be-detected image according to a pixel shift operation; and comparing the aligned product image areas, and identifying abnormal defects in the to-be-detected image. According to the scheme, the comparison data set of the defect-free product image is constructed, the feature point matching and pixel-level alignment technologies are combined, the problem of false detection caused by product position deviation or form change in a traditional method is effectively solved, the adaptability to illumination change is improved through brightness adjustment and difference matrix analysis, and the detection accuracy is improved. The method has the advantages of high detection precision, high environmental adaptability and high processing efficiency.

The invention discloses a whole-process vascular interventional operation training method and system based on a virtual reality technology, and the method comprises the steps: firstly, developing a vascular interventional mechanics and collision simulation module, and achieving the deformation simulation of a blood vessel and a guide wire through the combination of position dynamics, a region shape matching technology and a Cosseerat elastic rod model, the collision between the blood vessel and the guide wire is treated by adopting a wide-to-narrow three-layer progressive collision detection method; secondly, constructing an immersive virtual operation scene through model construction and rendering optimization; visual angle roaming, instrument operation and interface interaction are fused, and multi-mode interaction control driving intervention training is achieved; finally, based on a'module-event 'dual-drive framework, operation tasks such as anesthesia, puncture, radiography, intervention, stent release and hemostasis are sequentially connected in series through a unified state machine, and whole-process training is achieved. The accuracy and stability of blood vessel and guide wire simulation are improved, the training immersion sense is enhanced, interventional operation whole-process training is achieved, and high teaching and practical value is achieved.

The invention provides a migration scene-oriented adversarial sample generation method and device, a storage medium and a program product, and the method comprises the steps: receiving a to-be-attacked agent model, selecting a plurality of network layers in the agent model, and generating a random mask matched with the parameter shape of the selected network layer; shielding the parameters of the selected network layer based on the random mask to generate a multi-version proxy model; an original input sample is input into the multi-version agent model for forward propagation, gradient information of each agent model for a current adversarial sample is calculated, and a gradient direction is generated; and iteratively updating disturbance by using the gradient direction, and superposing the updated disturbance to the original input sample to generate an adversarial sample. According to the method, high-quality adversarial samples can be generated, and the attack efficiency is remarkably improved.

The invention discloses a photovoltaic module bird droppings defect detection method and system, and the method comprises the steps: carrying out the image screening of at least one first target image according to a preset shape matching strategy, and obtaining at least one second target image; inputting the at least one second target image into a preset neural network, and performing iterative training on the neural network to obtain a defect recognition model; at least one real-time initial image containing the photovoltaic module is obtained, the real-time initial image is input into the defect recognition model, and the defect recognition model outputs and obtains a defect detection result corresponding to the at least one real-time initial image. The remote dependency relationship and the position information of the input image can be coded from the two dimension directions of the horizontal space and the vertical space, so that the network can obtain the characteristics in a larger range, and the precision of the model is improved under the condition of not increasing extra calculation overhead.

The invention relates to the technical field of hot-rolled strips, in particular to a three-point difference dynamic adjusting method in the rolling process of hot-rolled strips, which comprises the following steps: S1, preparation before rolling and parameter initialization: roller shape matching and model parameter preloading; s2, dynamic regulation and control at the threading stage: thickness pre-compensation and temperature gradient suppression; s3, steady-state rolling control is conducted, specifically, three-point difference real-time monitoring and compensation are conducted, and thermal convexity dynamic compensation is conducted; and S4, drifting and exception handling: tail stability control and intelligent transition material insertion. According to the method, a full-process control system covering dynamic roll gap regulation and control, temperature-stress coupling compensation and intelligent decision support is constructed, and the problems that head and tail thickness fluctuation is difficult to restrain and edge thinning exceeds the standard by developing an asymmetric roll bending force distribution model and a rolling rhythm self-adaptive roll type matching strategy through traditional single-point thickness control can be effectively solved.

The invention discloses a power transmission line mountain fire smoke detection method and system, and the method comprises the steps: segmenting a mountain fire smoke candidate region in a to-be-detected power transmission channel picture through a pre-trained semantic segmentation model, and calculating a semantic segmentation confidence score; for each mountain fire smoke candidate area, calculating a shape similarity score through a shape matching method, calculating a position confidence score through a position function, and calculating a color confidence score through color distribution comparison; and carrying out weighted summation on the semantic segmentation confidence score, the shape similarity score, the position confidence score and the color confidence score to obtain a final smoke confidence score, and comparing the final smoke confidence score with a preset threshold to judge whether the corresponding mountain fire smoke candidate region is mountain fire smoke. According to the technical scheme provided by the invention, the forest fire smoke condition in the power transmission line channel picture can be effectively identified, the misidentification conditions of pond smoke, chimney smoke, cloud and mist and the like are obviously reduced, and the method has a wide application scene.

The invention relates to the technical field of digital people, and provides a method and equipment for generating a digital people video, which can be applied to a digital marketing scene. According to the method, the generation of the digital human video is divided into an audio generation process and a video generation process, when the audio is generated, a target text, an emotion parameter and a reference audio are taken as input, and the target audio of the target text is broadcasted by the sound of a reference person in the reference video and a set emotion, so that sound cloning is realized, the degree of violation of the generated digital human video is reduced, and the user experience is improved. And the text features are extracted from two dimensions of characters and pronunciation, so that the pronunciation accuracy during sound cloning is ensured. When the video is generated, the cloned target audio and the cloned reference video are used as input, and the fidelity of the digital human video is improved through expression and mouth shape matching, so that a reference person can quickly and efficiently generate vivid digital human videos with different broadcast contents only by shooting one reference video without shooting one by one, thereby reducing the manufacturing cost of the digital human video, and improving the user experience. And the video production efficiency is improved.

The invention belongs to the technical field of overhead line butt joint detection, and particularly discloses and provides a method for detecting overhead conductor steel strands after butt joint, and the method comprises the steps: building and correcting a digital twinborn model of a to-be-detected wire clamp based on a CAD drawing and three-dimensional scanning data of the to-be-detected wire clamp; extracting interface geometric parameters and identifying actual shapes of the interface geometric parameters; generating an initial gridding detection path based on a shape matching grid node distribution strategy; collecting a first ultrasonic data set according to the detection path and identifying an abnormal region; dynamically adjusting the grid density according to the abnormal index of the abnormal region, generating an updated detection path, and collecting a second ultrasonic data set; the first and second ultrasound data sets are combined to form a final detection data set. Through model correction and adaptive path planning, the defect of a fixed point distribution mode is overcome, the transformation from wide screening to accurate focusing is realized, and the defect identification accuracy and the evaluation reliability are remarkably improved.

The invention discloses a small-scale workpiece three-dimensional reconstruction method based on laser line scanning and application. The three-dimensional reconstruction method comprises the following steps: analyzing a key surface, and planning a scanning path and a scanning angle; performing multi-view scanning to obtain multi-view point cloud data; preprocessing to obtain preprocessed point cloud data; carrying out point cloud registration and splicing to obtain registration and splicing data; and generating a reconstructed high-precision triangular mesh surface as a three-dimensional reconstruction model. According to the technical scheme of multi-view scanning path planning, collection of point cloud data is more accurate and reliable, and parameters do not need to be frequently adjusted due to working condition changes; meanwhile, a multi-view point cloud registration and splicing mechanism and a point cloud feature extraction architecture of a U-net architecture are adopted, more details are reserved for surface geometric features of workpiece point cloud, meanwhile, a curved surface is implicitly expressed through a symbol distance function, and the method has remarkable advantages for digital-analog comparison of a workpiece with a complex profile. Three-dimensional shape matching and error evaluation can be efficiently and accurately carried out.

The invention discloses a robot path recognition method and system, and belongs to the technical field of robots, and the system comprises an acquisition module, a recognition module, a coffee cup shape recognition result acquisition module, a coffee cup position recognition result acquisition module, a determination module, a position posture acquisition module and a path recognition module. The method comprises the following steps: acquiring an environment image including a coffee making area, and performing multi-stage identification and processing on the image, including determining candidate block images of suspected coffee cups, obtaining coffee cup structure standard parameters for shape matching, and identifying a coffee making area range for position judgment. And finally, fusing the shape and position identification results to determine a target coffee cup so as to confirm the moving path of the coffee cup. The method can effectively solve the problem that in the prior art, due to the fact that coffee powder is attached to a camera lens, an image is blurred, coffee cup recognition is inaccurate, and path planning deviation is caused.

The invention provides a three-dimensional effect rendering method and device and electronic equipment, and the method comprises the steps: carrying out the zooming processing of a target surface patch according to the model size of a target three-dimensional model and the camera orientation of a virtual camera, obtaining a zoomed target surface patch of which the size is matched with the model size, and keeping the zoomed target surface patch facing the virtual camera; stretching the scaled target surface patch according to the model shape of the target three-dimensional model to obtain a stretched target surface patch of which the shape is matched with the model shape; and moving the stretched target patch to the position of the target three-dimensional model in the virtual scene, deleting the target three-dimensional model in the virtual scene and rendering the stretched target patch to obtain a rendered target patch so as to simulate the three-dimensional effect of the target three-dimensional model through the rendered target patch. According to the method, the target surface patch is stretched in the camera space according to the visual angle, so that the surface patch fitting three-dimensional effect is realized, and a better three-dimensional effect is realized with extremely low rendering performance consumption.

The application belongs to the field of intelligent human-computer interaction, and relates to a kind of intelligent writing and correcting method and system, comprising: utilizing flexible pressure sensor matrix and coordinate positioning system, the multidimensional data of sensing input pen writing, and constructing four-dimensional writing characteristic space;Adopt three-level correcting system, the logical reasoning verification of answer is carried out through depth semantic understanding model;The three-level correcting system successively includes OCR character shape matching, semantic rectification based on large language model and intelligent correction based on process reward model;Cognitive diagnosis model is introduced, and the degree of hesitation of answering is judged by analyzing answering index;Fusion question solving information, the knowledge point of question is identified using large model, and according to the question answering condition, the weak point of knowledge mastery is inferred;With the intelligent writing capture technology based on multi-modal sensor fusion, the writing information is accurately collected, and the intelligent correcting integrated method is constructed combined with cognitive diagnosis model, the efficient identification and accurate correction of writing content are realized.

The application discloses a power transmission line mountain fire smoke detection method and system, comprising the following steps: for the power transmission channel picture to be detected, a pre-trained semantic segmentation model is used to segment out the mountain fire smoke candidate area in the picture and calculate the semantic segmentation confidence score; for each mountain fire smoke candidate area, the shape similarity score is calculated through a shape matching method, the position confidence score is calculated through a position function, and the color confidence score is calculated through color distribution comparison; the semantic segmentation confidence score, the shape similarity score, the position confidence score and the color confidence score are weighted and summed to obtain the final smoke confidence score, which is compared with a preset threshold to determine whether the corresponding mountain fire smoke candidate area is mountain fire smoke. The technical scheme provided by the application can effectively identify the mountain fire smoke condition in the power transmission line channel picture, and significantly reduce the misidentification conditions such as pond smoke, chimney smoke and cloud and mist, and has a wide application scenario.

The invention relates to a precise infinitesimal displacement photoelectric measurement system and a precise infinitesimal displacement photoelectric measurement method, and belongs to the technical field of photoelectric measurement, and the precise infinitesimal displacement photoelectric measurement system comprises a curvature flow contour extraction module, a geodesic constraint positioning module, a Riemann optimization module and a displacement calculation module. Then parameterized representation and optimization are carried out on the light spot contour by using a curvature flow theory to generate high-quality contour description data; constructing a geodesic distance field based on a differential geometry principle, identifying a local maximum point as a candidate circle center, and evaluating the shape matching degree of each candidate circle center to determine an optimal candidate circle center; converting a sub-pixel positioning problem into an optimization problem on a Riemannian manifold, and determining a target coordinate of sub-pixel precision through gradient descent of a parameter space; finally, target displacement change between adjacent image frames is calculated, a micro displacement measurement result is output, and the displacement measurement precision is improved to 0.01 pixel and is improved by 5-10 times compared with a traditional method.

The application discloses a polygon decomposition method, an electronic device and a computer readable storage medium. The polygon decomposition method comprises the following steps: obtaining a to-be-decomposed polygon of an integrated circuit layout; obtaining a plurality of candidate polygon templates from a template library; performing shape matching processing and polygon update iteration processing on the to-be-decomposed polygon and the candidate polygon templates to obtain a shape matching result; and performing polygon decomposition processing on the to-be-decomposed polygon according to the shape matching result to obtain a trapezoidal set for manufacturing a mask after decomposition. According to the scheme of the embodiment of the application, the polygon in the integrated circuit layout can be efficiently decomposed, thereby reducing the decomposition time consumption and improving the universal applicability.

The invention discloses an electrical secondary drawing small target detection method, system and device based on an improved Faster R-CNN, and a medium, and belongs to the technical field of drawing small target detection. On the basis of optimization technologies such as a lightweight conditional generative adversarial network, multi-modal feature extraction and fusion, an adaptive Anchor mechanism and shape matching loss function, a dynamic receptive field and the like, the electrical secondary drawing small target detection task is completed with high precision and high efficiency.

The application discloses a kind of photovoltaic module bird dropping defect detection method and system, method includes: according to the shape matching strategy of preestablished to at least one first target image is carried out image screening, obtain at least one second target image;At least one second target image is input into the neural network of preestablished, and the neural network is iteratively trained, and defect identification model is obtained;At least one real-time initial image containing photovoltaic module is acquired, and the real-time initial image is input into the defect identification model, and the defect identification model outputs and obtains the defect detection result corresponding to at least one real-time initial image.It can be encoded to remote dependence relationship and position information from horizontal and vertical space two dimensional directions to input image, so that network can obtain greater range of features, improve the accuracy of model without increasing additional computing overhead.

The invention discloses a workpiece contour detection method and a related device. The method comprises the following steps: acquiring a pre-dispensing image and a post-dispensing image of a target workpiece; performing shape matching on the image before dispensing according to the contour feature point distribution template to obtain matching data before dispensing; acquiring a grid matching range according to the matching data before dispensing; performing segmentation processing on the dispensed image to obtain a current contour area and a current glue path area; and according to the grid feature point distribution template, performing shape matching on the current contour area and the current glue path area in a grid matching range, and obtaining current contour data of the target workpiece in the image after glue dispensing. According to the method, the image of the target workpiece before glue dispensing and the image of the target workpiece after glue dispensing are respectively collected, the target workpiece is primarily positioned according to the image before glue dispensing, the image after glue dispensing is secondarily positioned according to the result of the primary positioning, and the positioning error caused by tiny offset in the glue dispensing process is made up by secondarily positioning the target workpiece.

An embodiment of the present invention provides a method and system for adjusting a silicon-based resonant device based on image recognition, which reduces the influence of image distortion on the accuracy of image processing through a camera calibration method; calculates the center position of the laser spot by using an adaptive threshold segmentation method and a gray center-of-gravity method; quickly identifies the microstructure in the silicon-based resonant device through a shape matching algorithm; accurately measures the relevant dimensions of the microstructure by using a Hough detection algorithm; the image processing system serves as the feedback for motion control and laser processing, and the frequency detection circuit serves as the feedback for adjusting the performance of the silicon-based resonant device, and cooperatively controls the motion parameters of the four-axis motion platform and the energy parameters of the ultrafast pulsed laser, thereby realizing the automatic and precise trimming of the structural dimensions of the silicon-based resonant device. The present invention has the advantages of controllable trimming method, high adjustment accuracy, high adjustment efficiency, etc., and is applicable to the batch manufacturing of high-performance silicon-based resonant devices.

The invention provides an automatic drawing auditing method based on deep learning semantic segmentation and shape matching, and the method comprises the steps: converting a to-be-recognized drawing from a vector format to a bitmap format, obtaining an image corresponding to the to-be-recognized drawing, inputting a pre-constructed target recognition model for recognition, and obtaining a first target object of the to-be-recognized drawing, performing morphological optimization to obtain the contour shape and geometric attribute information of the first target object; selecting a target contour shape from a pre-constructed standard shape library, and calculating an intersection-to-union ratio of the contour shape to the target contour shape; if the intersection-union ratio is larger than the intersection-union ratio threshold value of the first target object, compliance auditing is conducted on the first target object, a compliance auditing result is obtained, and a structured auditing report is generated, so that the target objects in the drawings of various formats are rapidly and accurately recognized, the target objects are verified based on a shape matching method, and the verification efficiency is improved. And the efficiency, flexibility and accuracy of engineering drawing auditing are improved.

The application belongs to the technical field of overhead line butt joint detection, and specifically discloses an overhead conductor steel strand butt joint detection method, which comprises the following steps: constructing and correcting a digital twin model of a to-be-detected clamp based on CAD drawings and three-dimensional scanning data of the to-be-detected clamp; extracting interface geometric parameters and identifying the actual shape thereof; generating an initial grid detection path based on a shape matching grid node distribution strategy; collecting a first ultrasonic data set according to the detection path and identifying an abnormal area; dynamically adjusting the grid density according to the abnormal index of the abnormal area, generating an updated detection path, and collecting a second ultrasonic data set; and merging the first and second ultrasonic data sets to form a final detection data set. Through model correction and adaptive path planning, the application overcomes the shortcomings of a fixed point distribution mode, realizes a transition from extensive screening to accurate focusing, and significantly improves the defect identification accuracy and evaluation reliability.

The application provides a trackless navigation texture information redundancy analysis method, comprising: processing original image data of a current wear area by using an edge detection algorithm to obtain wear boundary contour coordinates and gray distribution data; extracting boundary coordinate point sets from a reliable anchor point area and determining an optimized update frequency value according to the position and shape matching degree of the boundary coordinate point sets and historical path map data; incrementally updating the path map by using the optimized update frequency value, integrating the gray distribution data of the reliable anchor point area and the texture detail loss degree value, and generating an enhanced positioning reference data set; updating the enhanced positioning reference data set according to supplementary data and performing periodic matching degree evaluation according to the optimized update frequency value to maintain long-term positioning consistency.

This invention discloses a traffic sign detection method based on spatial completion and multi-scale focusing, comprising: performing category balancing enhancement on the traffic sign dataset to construct a balanced dataset; embedding a feature complementarity mapping module in the Q4 layer of the YOLOv12 backbone network, fusing color and spatial information through channel segmentation and bidirectional attention mechanism to complete the features of small targets; embedding multi-scale convolutional attention modules in the P3, P4 and P5 layers of the detection head respectively, using differentiated strip convolution kernels and color attention mechanism to achieve feature focusing of traffic signs at multiple scales; using a composite loss function combining SIoU loss and shape matching loss for network training optimization; performing color enhancement, multi-scale feature extraction and non-maximum suppression algorithm post-processing in the inference stage to output the traffic sign category, bounding box and confidence score; this invention significantly improves the accuracy of small target detection and multi-scale adaptability, meeting the real-time detection needs in complex traffic scenarios.

A workpiece detection system capable of detecting a workpiece accurately and rapidly comprises: a three-dimensional sensor that measures a three-dimensional shape of an object in a field of view; and an information processing device that uses a visually recognizable shape in each point of view of a three-dimensional model as a search target shape, and, when a matching partial shape that matches the search target shape is found from the three-dimensional shape measured by the three-dimensional sensor, outputs the matching partial shape as a search result. The information processing device excludes a part of the search target shape or a part of the search result on the basis of a preset condition.

A vehicle control system incorporates a variety of neural networks used to compute multiple wheel slip estimates for multiple wheels based on wheel speed measurements for each wheel and outputs from other vehicle sensors, such as an IMU or GNSS receiver. Separate neural networks estimate the wheel slip for each wheel. These wheel slip estimates can be used to select wheel speed measurements to update a Kalman filter. The Kalman filter further updates its state based on GNSS measurements that were not rejected according to a rejection algorithm that includes shape-matching with respect to a trajectory estimated by the Kalman filter.

The present application provides a kind of personnel identification and counting method and system based on single vision, comprising: using latitude correction algorithm to carry out equidistant cylindrical development to wide-angle image of single vision, eliminate edge human distortion;Personnel detection model is trained through data set, and the feature response of occluded target is adaptively enhanced;The virtual trajectory of the occluded target is maintained by establishing a "parent-child association" state machine;Trajectory matching algorithm is used to perform time sequence shape matching on the trajectories inside and outside the door, accurately retrieve the lost exit target due to occlusion and perform counting compensation.The present application realizes high-precision counting of complex occlusion behaviors such as side by side and close contact under the premise of using only single vision and without depth sensor, with a counting error rate of less than 2%, significantly reducing the deployment cost of the system.

The invention discloses a bracket-free invisible orthodontic appliance based on direct 3D printing and a digital design and manufacturing method of the bracket-free invisible orthodontic appliance. According to the method, a digital closed-loop process is constructed; firstly, a correction target is set through intraoral scanning and digital tooth arrangement; constructing a smooth envelope space of the appliance by utilizing a contraction wrapping algorithm; accurately segmenting the dental crown by adopting an AI technology based on PointNet + + so as to define a force boundary; generating an appliance digital model with an accurate form and stable mechanics through a thickness control and edge optimization algorithm; finally, based on the DLP technology, the orthodontic appliance is directly printed with biocompatible resin, and the performance and safety of the orthodontic appliance are ensured through strict post-treatment such as secondary curing. According to the method, a traditional film pressing process is omitted, accumulative errors are eliminated fundamentally, personalized gradient design of mechanical properties of the appliance can be realized through process control, and an invisible correction technology is marked to step from shape matching to function matching.

The application provides a trackless navigation texture information redundancy analysis method, comprising: processing original image data of a current wear area by using an edge detection algorithm to obtain wear boundary contour coordinates and gray distribution data; extracting boundary coordinate point sets from a reliable anchor point area and determining an optimized update frequency value according to the position and shape matching degree of the boundary coordinate point sets and historical path map data; incrementally updating the path map by using the optimized update frequency value, integrating the gray distribution data of the reliable anchor point area and the texture detail loss degree value, and generating an enhanced positioning reference data set; updating the enhanced positioning reference data set according to supplementary data and performing periodic matching degree evaluation according to the optimized update frequency value to maintain long-term positioning consistency.