343 results about "Thermodynamic diagrams" patented technology

The invention discloses a large-scale traffic network jam prediction method and device based on a convolutional neural network. The time sequence and spatiality of road network vehicle speed information can be considered at the same time, so that the traffic jam state of the whole road network can be predicted more accurately. The method comprises the steps of S1, collecting GPS data of vehicles, extracting vehicle operating data at each moment on each road section, and generating a matrix M according to the acquired vehicle operating data; S2, generating a space-time thermodynamic diagram of at least one day according to the matrix M, wherein the abscissa of the space-time thermodynamic diagram represents time, and the ordinate of the space-time thermodynamic diagram represents road section ID sequence ranked according to spatial relationship; S3, generating a data set (X,Y) through window sliding on the space-time thermodynamic diagram; S4, establishing a convolutional neural network model, and training the convolutional neural network model by means of the data set (X,Y); S5, inputting a data set to be tested into the trained convolutional neural network model to obtain a prediction result.

The invention provides a vehicle demand thermodynamic diagram generation method and a device thereof and relates to the thermodynamic diagram technology field. The method comprises the following steps of carrying out grid division on an electronic map in an assigned area; acquiring order data corresponding to each divided grid; according to an order total amount in the order data of each grid, and based on a preset merging rule, merging the adjacent grids in the electronic map and acquiring an electronic grid map, wherein the preset merging rule includes rules of the order total amount in the grids and an adjacent position relation among the grids; according to the order data, taking an order generation place of an internal portion edge of each grid in the electronic grid map as a peak, and connecting each peak so as to generate a polygon area; and coloring the polygon area according to a vehicle demand and acquiring a vehicle demand thermodynamic diagram of a current area. In the invention, accuracy and practicality of the vehicle demand thermodynamic diagram are increased.

The invention discloses an end-to-end license plate identification method. The method comprises the steps of firstly segmenting a license plate region to serve as an input according to a license plate locating result; secondly predicting a license plate character thermodynamic diagram by adopting a designed three-layer full convolutional neural network, and obtaining candidate character regions according to the thermodynamic diagram; thirdly performing classification and position correction on the candidate character regions by adopting a designed seven-layer deep convolutional neural network model to obtain character sequences and position sequences; and finally selecting out the character sequences meeting Chinese mainland license plate standards by adopting a template-based optimal path algorithm, wherein the selected character sequences are identification results. The whole network adopts a multi-task combined training mode; the input is a license plate color image, and an output is a license plate number; and the method avoids the defect that a license plate needs to be accurately segmented in a conventional license plate identification method, and can effectively increase the license plate identification rate in a complex scene.

The invention discloses an image target detection method based on weak supervised learning, and belongs to the technical field of machine vision. The method comprises the following steps: firstly, collecting an image data set, and training a constructed deep convolutional neural network model based on a multi-scale feature map by adopting a multi-example learning method; inputting an actual image,and extracting a category thermodynamic diagram of the actual image through the deep convolutional neural network model; and finally, outputting a bounding box of a target in the category thermodynamic diagram by adopting a binarization image connected region analysis method to obtain a target detection result. According to the method, an image target detection task is realized based on a weak supervised learning method. The target detection task can be completed only by using the image-level classification annotation information in annotation in convolutional neural network model training, which is different from the target bounding box annotation information required in the prior art, so that the work of manually annotating the target in the image is greatly reduced, and the completionof the image target detection task has more economic benefits.

The invention provides a method for replaying a user web page operation behavior, the method comprises the steps that according to different event types triggered in the user web page operation, different ways are adopted to collect event information, the collected event information is analyzed, and corresponding operation behavior records of a user are formed; events at all the time points are created according to the operation behavior records, and in the process of replaying, simulated triggering is conducted on the events corresponding to all the time points to display. Accordingly, the invention further provides a system and thermodynamic diagram display device for replaying the user web page operation behavior, and the authenticity and the integrity for replaying the user web page operation behavior are promoted.

The invention discloses an attitude recognition method and a system based on a thermodynamic diagram and an offset vector, and a storage medium. The method comprises the following steps: obtaining a to-be-recognized target image; performing feature extraction on the to-be-recognized target image; predicting key point positions according to the extracted features; correcting the predicted key points, and determining final positions of the key points; and determining attitude information of a to-be-recognized target according to the key points. According to the method, the features of the imageare extracted, then the key point positions are predicted, the prediction result can be corrected, and finally, the attitude information is obtained through recognition. More accurate attitude information can be obtained, and the method can be widely applied to the technical field of deep learning.

The invention provides a method for generating a thermodynamic diagram, comprising the following steps of: firstly, recording current coordinates of a cursor in a webpage in a coordinate system with the current resolution; secondly, converting the current coordinates into standard coordinates of the cursor under the standard resolution; and finally, recording the standard coordinates and generating the thermodynamic diagram of the webpage according to the standard coordinates. Correspondingly, the invention also provides a terminal for recording data required for generating the thermodynamic diagram and a server for generating the thermodynamic diagram. By implementing the method and the device provided by the invention, the thermodynamic diagram with higher accuracy and correction can be generated and a heat point webpage area on the webpage can be accurately displayed.

The embodiment of the invention provides a computing method based on a neural network. The method comprises the steps of receiving an original image; extracting a feature tensor of the original image; and processing the feature tensor based on a trained fixed-point neural network to generate an image thermodynamic diagram. The embodiment of the invention provides the computing method in which the neural network is achieved by a fixed-point method, as fixed-point computation is adopted, a computation amount is small, fewer resources are occupied, and thus requirements on hardware is low.

The invention provides a thermodynamic diagram showing method and device. By adopting the embodiment, a showing ratio of a target image is obtained and thermodynamic parameters are further obtained according to the showing ratio so that a thermodynamic diagram corresponding to the target image is shown according to statistical index data of the target image and the thermodynamic parameters; the thermodynamic diagram is shown by adopting the thermodynamic parameters corresponding to the showing ratio of the target image, so that the problems in the prior art that set thermodynamic parameters are adopted so that a distribution condition of the statistical index data of the target image can not be clearly shown by the thermodynamic diagram can be avoided, and thus, the showing reliability of the thermodynamic diagram is improved.

The invention discloses a dense crowd counting and personnel distribution estimating method based on deep learning, which includes the following steps: preprocessing a target crowd image to be detected; selecting an appropriate data set to train a designed deep learning network model; inputting target data to the trained network for analysis and identification; and finally, getting a thermodynamic diagram of people in the coverage area, and getting the number and distribution of people in the crowd. Through the method, the number and distribution of people in a high-density crowd can be estimated accurately, and the problem that people in a crowd are blocked due to high density is well solved. The method is applicable to video images under any lighting condition, in any environment background and of any resolution.

The invention provides a thermodynamic diagram drawing method and apparatus and belongs to the image processing field. The method includes the following steps that: a map to be processed and thermodynamic data corresponding to the map are obtained; the map is divided into a plurality of cells; thermodynamic density values corresponding to each cell are calculated according to the thermodynamic data; color descent gradient is calculated according to a preset color shallowest value, a preset color deepest value, a maximum thermodynamic density value and a minimum thermodynamic density value in the thermodynamic density values corresponding to all the cells; and a thermodynamic diagram can be drawn on the map according to the color descent gradient, the thermodynamic density values corresponding to each cell, and the preset color shallowest value.

The invention discloses a construction method of a network-intensive type mini-passenger car travel demand thermodynamic diagram. The method comprises: the system is open to all registered users; division is carried out by using a traffic zone as an area dividing means; color marking is carried out on areas of the traffic zone and color filling is completed; a pie chart of the statistics is used as a marking way for a network-intensive type mini-passenger car travel demand thermodynamic diagram of the traffic zone; and for a network-intensive type mini-passenger car travel demand thermodynamic diagram facing the arrival demand, only a correspondence relation needs to be adjusted, marking is carried out based on the arrival demand thermodynamic diagram and the pie chart of the correspondence relation is screened to express the arrival demand thermodynamic diagram of the traffic zone, thereby forming a network-intensive type mini-passenger car travel demand thermodynamic diagram facing the arrival demand. With the construction method, according to actual demands, different demand objects can be selected to use the network-intensive type mini-passenger car travel demand thermodynamic diagram to analyze the passenger demands so as to transport capacity allocation reasonably, thereby realizing transport capacity balancing and guaranteeing satisfaction of the travel demands of citizens.

The invention discloses a method and a device for displaying thermodynamic diagrams. The method includes acquiring a plurality of web page frames in dynamically displayed web pages; acquiring thermodynamic diagrams relevant to each web page frame and generating thermodynamic diagram sets; arranging slide keys on the various thermodynamic diagrams in the thermodynamic diagram sets; dynamically displaying the thermodynamic diagrams in the thermodynamic diagram sets after users trigger the slide keys. The dynamically displayed page frames are page frame sets displayed in a rolling manner. The method and the device have the advantage that the thermodynamic diagrams of the dynamically displayed page frames can be visually displayed by the method and the device.

The invention relates to a deep license plate detection method based on a thermodynamic diagram and key point regression. The method includes an off-line learning phase and a detection phase. The off-line learning phase includes the following four steps of: (1) designing network depth; (2) preparing a training sample set; (3) performing sample labeling; and (4) performing training. According to the step (1), a deep learning network structure is designed, inputted images are adjusted and unified, first-layer convolution and activation function operation are performed on the images, second-layerconvolution and activation function and pooling operation are performed, and third-layer convolution and activation function and pooling operation are performed, and a plurality of task branches areformed, one branch learns the coordinates of plate number plates through one convolutional layer, another branch learns the thermodynamic diagram of the license number plates through another convolutional layer. According to step (2), a batch of vehicle front or rear images is obtained and adopted as a sample set for offline learning, and the larger the total number of the classes of samples is, the better a training effect is, and the sizes of the samples are normalized. According to the method, the offline trained deep network is adopted to characterize a target, and therefore, license platedetection can be performed on the target quickly and steadily.

The invention relates to a method and system for accurate online advertisement putting based on big data drive. The method comprises the steps that a user management system is established; user behavior data classification is conducted on interested users stored in the user management system, and a user behavior database is established; screening and searching are conducted according to user behavior data database classification items; thermodynamic diagrams, shown on corresponding maps, of the corresponding user behavior databases are established. The historical changes of user interests are visually shown in the modes of the thermodynamic diagrams, the thermodynamic diagrams with more interested users are deep in color, the thermodynamic diagrams with fewer interested users are light in color, the changes of the user interests are known according to the thermodynamic color changes of the users on the maps, data information is prospectively obtained according to the changes of thermodynamic diagrams, the user behavior data can be effectively, continuously and accurately collected, an advertisement putting effect can be improved, and a coaching effect can be played on marketing decision making.

The invention discloses a thermodynamic diagram generation method. The thermodynamic diagram generation method comprises the following steps that: evenly dividing an area to be processed into a plurality of subareas in advance; after a mobile collection record reported by each mobile collection base station in a statistical period is obtained, determining the subarea to which a GPS (Global Positioning System) of each mobile collection record belongs; according to each mobile collection record, determining the collection record number of each subarea; and finally, according to the collection record number of each subarea and the center point of each subarea, generating the thermodynamic diagram of the area to be processed. Therefore, the thermodynamic diagram of a whole area can be generated through a data collection record reported by the collection base station which continuously moves, and the problem of inaccurate thermodynamic diagrams since the base station and a geographic position can not be fixed due to the movement of the base station in the prior art is overcome.

The invention relates to an advertisement putting method and system based on groups interested in directional house property. The method includes establishing a building information database, and displaying house attributed on a map according to a building geographical location; based on group Internet-surfing behavior analysis, establishing a group interest database, and performing screening search on user interests; associating the building information database with the group interest database, and showing on the map; building a thermodynamic diagram of a user building interest database displayed on the map, and searching or screening attention groups corresponding to buildings with designated features; and performing accurate putting of building advertisements according to the interested groups by an advertisement putting system. Access tracks of groups are analyzed and classified, user interest preference is obtained, and general user portraits of users are judged, groups interested in house property are obtained and are displayed on the map in the form of the thermodynamic diagram, thereby realizing accurate positioning of users interested in different buildings, and completing accurate pushing of building advertisements according to classifications to users with different interest.

The embodiment provides a target integrity detection method and a device. The method comprises steps that an original image of a target is inputted into a neural network to acquire a thermodynamic diagram of the original image, and pixel values of the thermodynamic diagram represent pixels of the corresponding positions of the original image belonging to probability of the target; edges of the target in the original image are determined according to the thermodynamic diagram; and whether integrity of the target in the original image is realized is determined according to the edges. Through the method and the device, when integrity of the target in the original image is detected, robustness of interference signals such as busy background of the original image can be realized, moreover, on the condition that the thermodynamic diagram is acquired in the neural network, subsequent operation can be accomplished only through several relatively simple image processing steps, and usability is improved.

The invention discloses a web page thermodynamic diagram loading method and device. The web page thermodynamic diagram loading method includes: adding a screenshot of a to-be-loaded web page to a web page thermodynamic diagram configuration interface which is used for configuring interfaces of web pages needing to load thermodynamic diagrams; acquiring click rate of the to-be-loaded web page; according to the click rate of the to-be-loaded web page, loading the thermodynamic diagrams of the to-be-loaded web page on the screenshot of the to-be-loaded web page. Through the method and device, the problem of long time for web page thermodynamic diagram loading in the prior art is solved.

The invention discloses an urban people flow prediction method based on a Seq2Seq generative adversarial network, and the method comprises the steps: abstracting the urban people flow data at different times into image frames, and representing the people flow through a thermodynamic diagram; Dividing the observation data into training data and labels according to time, and converting the problem into an image problem; The idea of WGAN generative adversarial network is generally adopted, a generator generates pedestrian flow in a certain period of time in the future on the basis of historical data by using a Seq2Seq method, and external factors such as weather are added at the same time; The discriminator uses a Waserstein distance to discriminate true and false data; In the training process, the generator and the discriminator are continuously optimized by combining the generative adversarial loss and back propagation. And finally, when the discriminator cannot discriminate the authenticity, the optimized generator is used for predicting the future urban pedestrian flow. According to the method provided by the invention, the generative adversarial network is used for carrying out urban people flow prediction for the first time, and the problems of fuzzy prediction and slow algorithm convergence are solved in combination with external environment factors.

The invention discloses a webpage thermodynamic diagram generating method and device. The webpage thermodynamic diagram generating method includes the steps that webpage click rates of a plurality of webpage areas on a webpage are detected; a preset reference webpage click rate is obtained; the webpage click rates of the webpage areas are respectively compared with the reference webpage click rate; datum areas are obtained, wherein the datum areas are the areas with the webpage click rates larger than the reference webpage click rate; coloring chromatic aberration of the click rate per unit in the datum areas is determined; based on the determined coloring chromatic aberration of the click rate per unit, the datum areas are colored to obtain a webpage thermodynamic diagram. Through the webpage thermodynamic diagram generating method and device, the color contrast ratio between the webpage areas with the different click rates in the webpage thermodynamic diagram is increased, and the effect of enhancing the vision experience degree of a user is achieved.

The invention discloses a customized bus planned route travel demand thermodynamic diagram construction method, which comprises the steps that all regions in a city are subjected to regional division according to square shape of big bus stops, three layers, including a departure demand layer, an arrival demand layer and a travel demand layer, are loaded on a bus route network electronic map, the three layers need to be loaded simultaneously within each set time period, the arrival time is obtained according to route planning estimated travel time, a customized bus planned route travel demand thermodynamic diagram is formed on the basis of a departure demand thermodynamic diagram and an arrival demand thermodynamic diagram, an annotation method of the customized bus planned route travel demand thermodynamic diagram is decided, and an annotation method of a customized bus planned route arrival demand thermodynamic diagram is decided. Through establishing the customized bus planned route travel demand thermodynamic diagram, the method provided by the invention satisfies the route planning and bus scheduling requirements when encountering a large number of passengers, and fully considers different get-off requirements of fixed routes.

The invention is applicable to the technical field of vehicle-mounted radars and provides a method for detecting people in a vehicle, a radar and a vehicle. The method comprises the following steps of: acquiring a plurality of paths of target echo signals when people are in the vehicle, and preprocessing each path of target echo signals to obtain a plurality of paths of digital target signals; calculating a CAPON spectrum of the plurality paths of digital target signals to obtain target two-dimensional thermodynamic diagram information, and establishing an in-vehicle people identification model according to the target two-dimensional thermodynamic diagram information; acquiring multiple paths of echo signals to be detected in the vehicle, repeating the steps of preprocessing and calculating the CAPON spectrum to obtain the two-dimensional thermodynamic diagram information to be detected, and determining whether people are in the vehicle or not according to the in-vehicle people identification model and the two-dimensional thermodynamic diagram information to be detected. The embodiment of the invention is not influenced by the environment, can accurately detect whether a resident person exists in the vehicle, and ensures the personal safety of resident children or people who are inconvenient to move in the vehicle.

Embodiments of the invention provide an image processing method, device and system, and a storage medium. The image processing method comprises the steps of acquiring a human face image; performing human face semantic segmentation on the human face image, so as to obtain a human face thermodynamic diagram used for indicating the positions of human faces in the human face image and label information used for indicating whether different pixels in the human face image belong to the same human face or not; and distinguishing different human faces in the human face thermodynamic diagram based on the label information and the human face thermodynamic diagram. According to the image processing method, device and system, and the storage medium, when the human face image is subjected to the humanface semantic segmentation, the label information is obtained in addition to the human face thermodynamic diagram. The different human faces in the human face thermodynamic diagram can be distinguished on the basis of the label information. The position of each human face in the human face image can be recognized in combination with the human face thermodynamic diagram and the label information. Therefore, the problem that multiple human faces cannot be distinguished when being overlapped can be well solved according to the image processing method.

The invention relates to a method of a lightweight framework for generating a thermodynamic diagram according to a streaming data concept. According to the method of the lightweight framework for generating the thermodynamic diagram according to the streaming data concept, due to the facts that a traditional data collection framework is improved, and the high concurrency processing capacity and the good extensibility characteristic brought by a new technology are used, the data collection and cleaning process is simplified, effectiveness is improved, the data storage scale is decreased, the storage process is optimized according to the data characteristics of the thermodynamic diagram, and the thermodynamic diagram generation and showing process is formed finally. The high concurrency processing capacity of Node.js, caching of Redis and the data batch processing capacity of MySql are combined with the characteristics of the thermodynamic diagram to implement the method of the lightweight framework for generating the thermodynamic diagram according to the streaming data concept.

The invention discloses a malignant melanoma and non-malignant melanin nevus classification method based on deep learning. The method comprises the following steps: firstly, acquiring data; performingdata preprocessing, then, enhancing the processed data; then, performing the malignant and non-malignant tumor classification at a small sliding block level, generating probabilistic thermodynamic diagrams, machine learning feature extraction, and finally, performing random forest classification based on pathological image levels. According to the invention, a very good experiment result of the small slide block level of the pathological image of the eye superpixel tissue is obtained; more importantly, a whole set of process of automatically classifying tissue pathological images as malignantor non-malignant is realized, a computer aided diagnosis (CAD) system which can be directly applied to clinical aided diagnosis is constructed, and the accuracy is very high.

The invention discloses an Echarts-based hierarchical diagnosis and treatment analysis and data visualization system, comprising a data cleaning module used for acquiring information about patients, diseases and hospitals from medical record information, a patient analysis module used for analyzing medical treatment of the patients, transfer treatment, characteristics of the patients, influence factors of choices for the hospitals and the like from the levels of provinces, cities, districts and counties, a disease analysis module used for analyzing disease regional characteristics, disease seasonal characteristics, cross-level disease treatment characteristics, different place disease treatment characteristics, common disease treatment characteristics, slow disease treatment characteristics and the like from the levels of provinces, cities, districts and counties, a hospital analysis module used for analyzing sources of the patients clinically accepted by the hospitals, characteristics of the clinically accepted patients and structural characteristics of the clinical diseases from the levels of provinces, cities, districts and counties, and a data visualization module used for realizing analysis result visualization by means of drawing a map, a dynamic flow graph, a thermodynamic diagram, a pie chart or a histogram. After the Echarts-based hierarchical diagnosis and treatment analysis and data visualization system is adopted, a hierarchical diagnosis and treatment analysis process is more systematized, and data analyzing and processing efficiency is improved.

The invention provides a target tracking method and a device based on a thermodynamic diagram. The method comprises the following steps: obtaining a current frame image; inputting the current frame image to a retrained residual network model, and outputting a thermodynamic map of the current frame image, wherein the thermodynamic map displays a probability of a tracking target position in the image in a special highlight form; a position of a tracking target in the current frame image being determined according to a thermodynamic map of the current frame image. The invention provides a targettracking method and a device based on a thermodynamic diagram. By outputting a thermodynamic diagram form of a current frame image, the position of the tracking target is determined, the complexity ofa residual network is reduced, the tracking speed is improved under the premise of ensuring the accuracy of the target tracking, and the real-time performance of the target tracking is enhanced.

The invention discloses a link click rate display method of a webpage thermodynamic diagram and a link click rate display device thereof. The link click rate display method comprises the steps of: acquiring the click rate of each pixel involved in a target page; acquiring the thermodynamic diagram corresponding to the target page according to the click rate of each pixel; determining at least one program element involved in the target page, wherein the at least one program element comprises a target program element with multiple linking elements; receiving a link click rate triggering action, wherein the link click rate triggering action is used to trigger the click rate of displaying multiple linking elements; respectively displaying the click rates of the multiple linking elements. Through the method and the device, the link click rate of each page can be displayed by the webpage thermodynamic diagram, and then the click rate of each linking element is displayed by receiving the link click rate triggering action, so that the page link click rate in the webpage thermodynamic diagram can be displayed.