70 results about "Regional model" patented technology

A system and method for segmenting cardiac images and, in particular, segmenting the left ventricle of the heart using a contour propagation model that integrates visual information and anatomical constraints. The visual information comprises a gradient vector flow-based boundary component and a region component that separates the cardiac contours/regions according to their global intensity properties that reflect the different tissue properties. The anatomical constraints couple the propagation of cardiac contours according to their relative distance. The propagation model comprises a weighted integration of a boundary segmentation model, a region model and coupling function.

The invention discloses a method for extracting video objects, comprising the steps of detecting single image of the video, obtaining feature points of each image in the adjacent two images; matching the feature points in the adjacent images, determining movement equation of the video objects in the adjacent images; selecting an expression model of the video objects; filling the expression model according to a probability filling method to obtain a regional model of the video objects; according to the movement equation, computing an original seed region for the regional model of the video objects; using an image partitioning method to perform partitioning so as to obtain the seed region, finally, obtaining the video objects. According to the scheme provided by the invention, the seed region is selected by the efficient movement estimation algorithm, so that the video objects can be extracted in real time, and the problem that the prior target extraction efficiency is low is solved; in addition, the target extraction accuracy can be effectively improved.

The invention discloses a sponge city test point area low-influence development facility simulation method, which comprises the steps of establishing an overall background model of a test point area in a storm flood management model SWMM, and simulating a rainwater runoff total amount and a runoff peak value before sponge city construction; selecting a target source project in the test point area,utilizing a low-influence development facility simulation module in the SWMM model to input design parameters corresponding to different types of low-influence development facility LIDs in the targetsource project, establishing a project model of the target source project, and simulating a reduction result of the target source project on the total rainwater runoff amount and the peak value, wherein the target source project is generalized into a sub-catchment area in the area integral model; and according to the simulation result of the target source project in the project model, the basic parameters of the sub-catchment area corresponding to the target source project are calculated and adjusted through the parameters, so that the overall sponge city construction effect of the test pointarea is embodied in the area model. According to the method, in large-scale regional simulation, the accuracy of low-influence development facility effect simulation in a regional model is guaranteed.

The invention relates to the technical field of intelligent traffic, in particular to a region-based parking service and management method and system, and the method comprises the steps: obtaining the region information of a region, wherein the region information comprising a region attribute, a region number and a region position; acquiring crowdedness degree information of areas corresponding to different area attributes; adjusting the area attribute of the area, expanding the area range with the high crowding degree to reduce the crowding degree, and reducing the area range with the low crowding degree to improve the space utilization rate. By establishing the area model and the attribute model, the mutual relation between the area and the attribute is cleared, the area attribute of the area can be adjusted, and finally, the parking supply and demand in the area are balanced, so that the problem of time-period and regional imbalance of the parking demand and parking space supply is solved, the area use efficiency is improved, and the parking problem is solved.

Provided are a regional and remote damage-assessment system and method established based on an artificial-intelligence supervised learning linear regression method for different types of vehicle, pertaining to the field of vehicle damage assessment. In order to achieve regional detection after vehicles collide to each other, the system comprises a regional detection subsystem for determining the regions where vehicles collide and for learning regional training data to generate a regional model. The regional model is set up by using the artificial-intelligence supervised learning linear regression method. The remote damage-assessment system and method have following beneficial effects: through the above technical scheme, regional detection of collided vehicle types is achieved; a machine learning method is used in the technical field of remote damage-assessment; and as for the machine learning method, determination accuracy is improved during a damage-assessment process.

The invention provides a logic centralization based global distributed global full-network model maintenance method. The method comprises the steps of building a province-level model service center; checking the mode of the global distributed global full-network model; building an unified model service center; reviewing the information of boundary models. With the adoption of the logic centralization based global distributed global full-network model maintenance method, the regional model information can be centralized and stored; the problems that information congestion and performance reduction caused by large physical centralization can be decreased; in addition, the performance problem caused by centralized model maintenance can be effectively reduced; therefore, the handling efficiency of the system can be increased, and various demand of various applies to a power grid model can be highly supported.

The invention discloses a system and a method for monster distribution in a game scene. The method includes: acquiring a game appointed scene region to obtain scene regional model configurations; according to the regional model configurations, adopting a monster distribution editor for custom configurations of monster division in the region to form a monster configuration data table of the appointed region; in running of a game program, loading the obtained monster configuration data table to generate monsters. The system and the method for monster distribution in the game scene have advantages that a monster designer can use appointed monster distribution rules to allow the program to automatically distribute monsters in the appointed region according to the rules while the designer keeps calling permission, so that monster distribution efficiency is improved, and flexibility in modification is realized; the monsters are distributed according to the rules in running of a random scene, randomness in monster generation is improved, the monsters are different in each-time access of the scene, and accordingly game playability is improved; extending and modifying are facilitated, and the monster distribution rules can be added only by filling and modifying the table.

The invention relates to an establishing method of a regional passenger traffic model based on city attraction. The method is a brand-new regional passenger flow predicting model establishing method where a city attraction model is innovatively introduced based on the accuracy of big data. Application practice is conducted with the Guangdong-Hong Kong-Macao Greater Bay Area as an example, in combination of the advantages of the big data, a regional model is calibrated, and then the development tendency of the Greater Bay Area is analyzed. It is shown through the practical application result that the city attraction model has high practical operability and precision in regional passenger flow interaction research and can be used for guiding the analysis work of the space passenger flow of all national regions developing at high speed at present.

The invention relates to the technical field of order data processing, in particular to an online car-hailing target order prediction method based on CNN-LSTM. The method comprises the following steps: 1, fragmenting a preset region into a plurality of sub-regions; 2, collecting original order data of each subarea in a preset area; 3, obtaining target order data based on the original order data, wherein the target order data comprises the total order quantity, the average order price, the POI characteristics, the weather characteristics and the time characteristics of the same region in the same time period; 4, predicting order quantity data of each region in the next time period based on the CNN-LSTM model: inputting order total quantity data, POI features, weather features and time features in the target order data into the CNN-LSTM model to obtain order quantity prediction data of each region in the next time period; and 5, establishing a regional PVD model to obtain a value thermodynamic diagram of each sub-region. According to the invention, the target order data can be predicted comprehensively and accurately.

The invention provides a real-time simulation analysis system based on the industrial Internet, and the system comprises an industrial topology network construction module which creates an industrial system architecture model, receives a real-time regional simulation model uploaded by a regional model construction terminal, and imports the industrial system architecture model for data integration; a node distribution module, dividing regions for different operation nodes in the industrial system architecture model; a regional model building terminal, collecting static data and dynamic data of the operation nodes in the region for regional simulation, and obtaining a regional simulation model; and a node calibration module, calibrating the interaction relationship between the nodes involved in the dynamic data according to the static data of the current operation node, and outputting the calibrated interaction relationship between the operation nodes to the regional model construction terminal for construction of a regional simulation model. According to the method, the industrial topology network is constructed through regional node modeling, simulation analysis of dynamic data is carried out, iterative optimization of production plan simulation is realized, and the research and development and production efficiency is effectively improved.

The invention relates to a reverse construction method and system for automobile typical parts. The reverse construction method comprises the following steps: carrying out three-dimensional measurement on a sample automobile typical part to obtain automobile body point cloud data, and carrying out regional modeling on the automobile body skeleton part point cloud data and non-sheet metal part point cloud data to obtain a plurality of skeleton part basic feature curved surfaces and a typical structural part regional model; splicing the basic characteristic curved surfaces of the framework pieces to obtain basic characteristic curved surface transition surfaces of the framework pieces, and processing the basic characteristic curved surface transition surfaces of the framework pieces to obtain a vehicle body structural surface; respectively processing each typical structural part region model by adopting a section plane method to obtain a plurality of sketch contours and processing each sketch contour into a plurality of basic curved surfaces; and stitching the plurality of basic curved surfaces to obtain a stitched curved surface, and operating the stitched curved surface to obtain the non-sheet metal part model. According to the method and system, the construction accuracy of the automobile typical parts can be improved, and the constructed automobile body better conforms to theexpected effect.

The invention discloses a park worker positioning method and device, and the method comprises the steps: achieving the positioning of a worker through a worker positioning system, and determining an indoor region and an outdoor region of a park corresponding to a region model which is established for the park in advance; determining an area where the positioning terminal is located in the area model through the identifier of the positioning terminal and the initial placement position corresponding to the identifier; selecting a corresponding positioning mode for positioning in a plurality of pre-stored positioning modes according to the located area, and if the located area is changed, replacing the corresponding positioning mode with other positioning modes. By dividing the park into the indoor area and the outdoor area, different positioning modes can be selected according to different area types in the park, the positioning stability and accuracy are ensured, when the area where a worker is located changes, the positioning modes can be synchronously switched, and the situation that positioning signals are interrupted is avoided.

The invention discloses a building equipment management system based on a construction information model technology, relates to the technical field of construction equipment management, and solves thetechnical problem that in the prior art, the engineering progress speed is reduced due to the fact that regions cannot be reasonably distributed.The method includes: receiving a region model througha region division unit; reasonably planning the installation position of the large equipment for the regional model, dividing the regional model into a plurality of rectangular regions, then obtainingregional information of a region i, analyzing the regional information of the region i, obtaining an equipment planning coefficient Xi of the region through a formula, and obtaining the occupied areaof the large equipment; sorting the occupied areas from large to small, then grading the equipment, and sending the graded equipment to an equipment management platform; and reasonably distributing regions. The working efficiency of equipment is improved, and the situation that the engineering progress speed is reduced due to improper arrangement of equipment positions is reduced.

The invention relates to a PDF character extraction method and device, a computer device and a storage medium. The method comprises the steps of obtaining a PDF character extraction request; determining the range of a data area in the PDF document according to the PDF character extraction request; performing column cutting and row cutting on the data area; dividing the data area into a plurality of cells according to the results of column cutting and row cutting; and respectively extracting characters in the plurality of cells. According to the invention, aiming at the editable PDF file, the data area is dotted, segmented and mapped through the visual front-end tool in the frameless table area, the data area is subjected to unit gridding, and the character extraction is carried out throughthe regional model text of the unit gridding.

The invention discloses a system and a method for establishing the regional remote damage assessment of different vehicle models on the basis of an artificial intelligence unsupervised learning FuzzyC-Means clustering algorithm, and belongs to the field of vehicle damage assessment. In order to solve the problem of regional detection after vehicle crashes, the system has the technical key points that the system comprises a regional detection subsystem, wherein the regional detection subsystem judges a crashed region when the vehicle crashes, and learns region training data so as to generate a region model, and the region model establishes and uses the intelligence unsupervised learning FuzzyC-Means clustering algorithm. The system has the following beneficial effects: by use of the technical scheme, the regional detection of a vehicle crash can be realized, a machine learning method is used in the technical field of remote damage assessment, and therefore, discrimination accuracy is improved by the targeted machine learning method in a damage assessment process.

The invention discloses a cross-region communication flow prediction method based on heterogeneous model reuse. The method comprises a data acquisition step, a model reuse step, a new region model training step and a new region model test step. The method comprises the following steps: firstly, collecting communication data of different regions, including feature data and meta-information description data of features; training a model on the previous regional data, and mapping the heterogeneous model into the feature space of the current region by adopting a model reuse and optimal transportation technology; and finally, establishing a new model on the new area by using a small amount of communication data, soso to train a usable model with good performance by using a small amount of marked data. According to the method, the difficulties that feature spaces of new and old regions are different and only data in the new region is difficult to establish a new model can be solved, and themethod occupies few resources and is high in applicability in the implementation process.

The invention discloses an analysis method and system based on an energy technology supply and demand side response model, and the method comprises the steps: simulating the technology of an energy supply end and an energy processing conversion technology according to a preset TIMES model, and constructing an analysis model of an energy system index of a target region; through a scientific prediction technology, predicting to obtain the medium-and-long-term energy demand of the target region, and constructing an energy efficiency database and an emission factor database of the target region; simulating a set energy supply and demand side change scene according to the analysis model, and calculating to obtain an energy supply and demand relationship under different energy supply and demandside change scenes; and according to the energy supply and demand relationship, analyzing to obtain the development direction of the energy system of the target region. According to the method, the energy efficiency database and the emission factor database of the target region are constructed, and the comprehensive supply side scene and demand side scene are innovatively designed, so that the problem that an actual emission reduction path cannot be analyzed due to the fact that an existing region model mostly adopts a total constraint target can be solved.

The invention discloses a multi-sensor arrangement and evaluation method for intelligent automobile environment perception, and the method comprises the steps: building detection region models of various sensors, setting an arrangement scheme evaluation index, building an evaluation hierarchical structure through employing an analytic hierarchy process, and enabling the evaluation hierarchical structure to comprise a scheme layer, a criterion layer and a target layer, wherein the scheme layer comprises the multi-sensor arrangement scheme, the criterion layer comprises the evaluation indexes, the target layer comprises an evaluation result, and a judgment matrix between the evaluation indexes is constructed by using a consistent matrix method; determining the weight of each evaluation index in the criterion layer by using a judgment matrix; and calculating according to the value and the weight of each evaluation index to obtain an evaluation result of the multi-sensor arrangement scheme. According to the method, the corresponding quantitative indexes are formulated, so that the evaluation method is more scientific, and the multi-sensor arrangement method based on the evaluation method is high in redundancy, few in blind area and low in cost.

The invention discloses a multi-model coupling waterlogging early warning method and system based on Docker cluster management. The method comprises the steps that basic hydrological data and flood prevention data are collected, and a pipe network runoff model and a river channel model are established; performing model segmentation on the pipe network runoff model and the river channel model to obtain regional models, and integrating the regional models into a plurality of complete regional hydrological models according to a runoff process; in the regional hydrological model, a simulation result of the pipe network runoff model is used as a boundary condition of a river channel model for model coupling, a regional waterlogging coupling model is obtained, and a data processing analysis function module is established; respectively packaging the regional waterlogging coupling model and the data processing analysis function module into a plurality of Docker containers, respectively establishing a model cluster and a data processing cluster in a Docker cluster management mode, and deploying the model cluster and the data processing cluster into a cloud server; and performing waterlogging early warning analysis and visual display based on the model cluster and the data processing cluster. The waterlogging early warning model can give consideration to timeliness and accuracy at the same time.

The invention discloses a self-adaptive cross-regional base station energy consumption model migration and compensation method. The method mainly comprises two parts of generating an auxiliary log file by utilizing a Wassertein generative adversarial network and constructing a model for predicting the state of a target regional base station by utilizing a migration source regional model. Accordingto the method adopted by the invention, the method comprises the following steps: firstly, acquiring a log file of each regional base station, and generating a large number of auxiliary base stationlog files by utilizing a Wassertein generative adversarial network; and generating a classification model by using the log files of a large number of source area base stations and a small number of target area base station files for estimating the performance of the target area base station.

The techniques herein include using an input context to determine a suggested action. One or more explanations may also be determined and returned along with the suggested action. The one or more explanations may include (i) one or more most similar cases to the suggested case (e.g., the case associated with the suggested action) and, optionally, a conviction score for each nearby cases; (ii) action probabilities, (iii) excluding cases and distances, (iv) archetype and/or counterfactual cases for the suggested action; (v) feature residuals; (vi) regional model complexity; (vii) fractional dimensionality; (viii) prediction conviction; (ix) feature prediction contribution; and/or other measures such as the ones discussed herein, including certainty. In some embodiments, the explanation data may be used to determine whether to perform a suggested action.

The invention discloses a method for displaying component space information of an IFC model on the Web, which comprises the following steps: 1) in IFC modeling, carrying out space naming on a component of the model by adopting a space identifier; 2) analyzing the IFC file in a BIM engine to obtain graphic data and spatial structure information of the model; 3) generating a 3D model from the IFC file, attaching the spatial name extracted from the IFC file to the 3D model, assembling a tree structure through a regular expression according to the identifier of the spatial name, and storing the tree structure into a storage library; 4) calling a tree structure of the storage library data, taking a spatial name contained in a tree node as an index, traversing all components in the 3D model, andfinding a component matched with the index; and displaying the spatial information by hovering a component in the 3D model. The method overcomes the difficulties that the IFC Web display cannot obtain the space structure tree and the regional model component cannot be selected, and has the advantages that the range is gradually narrowed from large to small for viewing, the model component is quickly positioned, the affiliation information of the model component is displayed, and the model operation is carried out.