39 results about "Visibility graph" patented technology

The invention relates to a method for drawing an approximate soft shadow of a three-dimensional scene by utilizing position-sensitive visibility smooth filtering. The method comprises the steps of determining a visibility filtering space width corresponding to each visible scene point according to the positions of a directional point light source, the visible scene point and a shielding point, and parameters such as the size of a virtual surface light source, the size of a virtual shielding ball and the like; performing the position-sensitive visibility smooth filtering on a binary visibility graph; and performing modulation on a direct illumination value, not considering the shielding, of the visible scene point by using a filtering result to obtain an actual direct illumination value of the visible scene point. According to the method, the approximate soft shadow with the softness changed with position change of a shadow projection object and a shadow receiving object can be drawn and better conforms to daily cognition of people.

A deep learning model based on complex networks and its application in measurement signal analysis: set the principle of constructing complex networks for measurement signals through the method of visual graphs, and establish complex networks of visual graphs; use visual graphs for each channel measurement signal Methods The complex network was established, and for each network, the following indicators were extracted: node aggregation coefficient, node degree, node betweenness, node proximity centrality, and node eigenvector centrality; a deep learning model was established. The vertical oil-water two-phase flow experiment is carried out to obtain the measurement signal by using the circularly excited conductivity sensor; the ratio of the oil phase and the water phase is fixed, and the flow rate of the oil phase and the water phase is changed for the experiment. The present invention constructs a complex network for the measurement signal, uses a large number of indicators of the complex network to form samples, and uses a layer-by-layer optimization mechanism to combine supervised learning and unsupervised learning to obtain a deep learning model. It can be used to predict and classify measurement signals of unknown categories.

The invention relates to an epilepsy electroencephalogram signal identification method based on an optimal kernel time-frequency distribution visibility graph, which comprises: acquiring original electroencephalogram data, and calculating adaptive optimal kernel time-frequency distribution of a preprocessed electroencephalogram signal; carrying out classification of epilepsy states; for each obtained time energy sequence, carrying out visualization analysis, obtaining an energy time sequence visibility graph complex network, extracting to obtain feature indexes of the energy time sequence visibility graph complex network, extracting adaptive optimal kernel time-frequency distribution indexes, and classifying the epilepsy electroencephalogram (EEG) signal by combining a support vector machine. According to the epilepsy electroencephalogram signal identification method based on the optimal kernel time-frequency distribution visibility graph, which is disclosed by the invention, the energy time sequence visibility graph complex network is constructed by combining an adaptive optimal kernel time-frequency distribution principle and a visibility graph idea, the indexes of the complex network are extracted, and identification on the epilepsy electroencephalogram signal is implemented.

The invention relates to an environment modeling method, in particular to an environment modeling method based on a recursive reduced visibility graph. The environment modeling method includes the steps: (1) recording the position of a robot and the reaching position of the robot; (2) connecting S-H-G-L-S to form a quadrangle; (3) traversing each current edge of the quadrangle; (4) judging whether the current edges are intersected with obstacles or not; (5) determining a middle point; (6) turning edges of a newly enclosed polygon S-A-H-B-G-L-C-S to the step 3 for traversing; (7) finishing traversing the current edges of the polygon. A polygonal area is finally obtained, all obstacles outside the rear can be simplified, and only the obstacles inside the area need to be considered by robot route planning. The number of visibility lines can be effectively decreased, and the executing efficiency of a subsequent robot route planning algorithm is improved.

The invention provides a monitoring method for visibility range from eyes to a screen. The monitoring method includes presetting a safety threshold value of distance between two eyes of a human face image, performing image sampling according to a preset sampling period so as to determine actual distance between the two eyes of the human face image; determining M images during sampling, calculating average of the actual distance between two eyes of each of the M images, and prompting the users to regulate the visibility range from the eyes to the screen if the average is higher than the safety threshold valve; otherwise, returning to sampling. By the monitoring method, the visibility range of the user can be monitored in real time, the user can be prompted to keep moderated distance from the screen, health of the user is guaranteed, and intelligent and personalized service are provided for the users.

The invention discloses an analysis method and application of EEG signals based on a complex network. The analysis method of the EEG signals based on the complex network comprises the following steps: constructing multi-scale level limited penetrable visibility graph complex networks; calculating characteristic indexes of each multi-scale level limited penetrable visibility graph complex network; combining a support vector machine to classify the EEG signals, namely using a leave-one-out cross-validation and support vector machine classifier to classify all two-dimensional index vectors, and using a ten-fold cross-validation and support vector machine classifier to classify all the two-dimensional index vectors. According to the invention, multi-scale ideas and level limited penetrable visibility graph theories are combined to construct an EEG multi-scale level limited penetrable visibility graph complex network so as to extract complex network indexes, and the support vector machine classifier in machine learning is combined to realize high-accuracy classification for different EEG signals. The analysis method and application of the EEG signals based on the complex network can be applied to smart head-mounted wearable equipment, and sleep EEG signals are measured through analyzing the smart wearable equipment to monitor the brain state of a user, furthermore, necessary early warning can be provided.

The invention provides a terrain visibility analysis-oriented visibility graph network construction method. The method includes the following steps: (1) scanning DEM grid data, analyzing whether null data exist, and if yes, performing interpolation calculation, and supplementing data; (2) selecting a DEM grid unit neighbor mode of a constructed graph network, mode types including a 4-grid unit neighbor mode, a 8-grid unit neighbor mode and 16-grid unit neighbor mode; (3) calculating attribute values of vertices in the visibility graph network: based on a visible range analysis algorithm, calculating the number of visible grid points of a grid unit corresponding to each vertex in the network or the area of a visible range; (4) according to the selected neighbor mode type, calculating the weights of sides in the visibility graph network; and (5) saving visibility graph network data. The terrain visibility analysis-oriented visibility graph network construction method provided by the invention can be completely applied to graph network-based optimization modeling and solving of different fields of terrain visibility analysis of large-scale mass data, and processing efficiency can be improved.