Intelligent control system and method of music fountain based on smart tourism

Abstract

The present application discloses an intelligent control system and method for a music fountain based on smart tourism, which captures and learns the geographical distribution characteristics of tourists' locations by adopting a deep learning algorithm based on artificial intelligence and computer vision technology, and introduces a saliency enhancement module of a global context to enhance the expression of the geographical distribution characteristics, and then uses a reverse masking mechanism to cluster and fuse the enhanced geographical distribution characteristics of tourists, thereby forming a comprehensive representation of the geographical distribution characteristics of tourists, and finally uses a classifier to construct a performance mode determiner to intelligently determine the performance mode of the fountain according to the real-time geographical location information of tourists.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an intelligent control system and method for a music fountain based on smart tourism. Background Art

[0002] With the development of social economy and the improvement of people's living standards, tourism has become an important part of people's daily life. However, traditional tourism methods can no longer meet the high standards of modern tourists for tourism experience. Tourists are increasingly pursuing personalized and interactive tourism experiences. In order to meet this demand, smart tourism came into being. It integrates a variety of high-tech means to achieve efficient management of tourism resources and intelligent improvement of tourism services.

[0003] As a dynamic art form that integrates vision and hearing, music fountain has become an important part of many tourist attractions. However, traditional music fountain systems often lack interaction with tourists, have a single performance mode, and cannot meet the viewing needs of tourists. Therefore, an intelligent control system and method for music fountain based on smart tourism is expected. Summary of the invention

[0004] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides a music fountain intelligent control system and method based on smart tourism, which captures and learns the geographical distribution characteristics of tourist locations by adopting artificial intelligence-based deep learning algorithms and computer vision technology, and introduces a global context saliency enhancement module to enhance the expression of geographical distribution characteristics, and then uses the reverse mask mechanism to cluster and fuse the enhanced tourist geographical distribution characteristics, thereby forming a comprehensive representation of the tourist geographical distribution characteristics, and finally uses a classifier to construct a performance mode determiner to intelligently determine the performance mode of the fountain according to the real-time geographical location information of the tourists.

[0005] According to one aspect of the present application, a music fountain intelligent control system based on smart tourism is provided, which includes:

[0006] A location sensor module is used to collect the location information of tourists;

[0007] A fountain control module, used for determining a performance mode of the fountain based on the location information of the visitor;

[0008] A music synchronization playing device, used to play music based on the performance mode of the fountain;

[0009] The lighting control device is used to determine the lighting operation program based on the performance mode of the fountain and provide lighting coordination based on the lighting operation program.

[0010] According to another aspect of the present application, a method for intelligently controlling a music fountain based on smart tourism is provided, which includes:

[0011] Collecting visitors’ location information;

[0012] Determining a performance mode of the fountain based on the location information of the visitor;

[0013] Playing music based on the performance mode of the fountain;

[0014] A lighting operation program is determined based on the performance mode of the fountain, and lighting coordination is provided based on the lighting operation program.

[0015] Compared with the prior art, the present application provides a smart tourism-based music fountain intelligent control system and method, which captures and learns the geographical distribution characteristics of tourists' locations by adopting artificial intelligence-based deep learning algorithms and computer vision technology, and introduces a global context saliency enhancement module to enhance the expression of geographical distribution characteristics, and then uses the reverse masking mechanism to cluster and fuse the enhanced geographical distribution characteristics of tourists, thereby forming a comprehensive representation of the geographical distribution characteristics of tourists, and finally uses a classifier to construct a performance mode determiner to intelligently determine the performance mode of the fountain according to the real-time geographical location information of tourists. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0017] Figure 1 A block diagram of a music fountain intelligent control system based on smart tourism according to an embodiment of the present application;

[0018] Figure 2 A system architecture diagram of a music fountain intelligent control system based on smart tourism according to an embodiment of the present application;

[0019] Figure 3 A block diagram of a fountain control module in a music fountain intelligent control system based on smart tourism according to an embodiment of the present application;

[0020] Figure 4 A block diagram of a tourist geographic distribution feature processing unit in a music fountain intelligent control system based on smart tourism according to an embodiment of the present application;

[0021] Figure 5The present invention is a flowchart of an intelligent control method for a music fountain based on smart tourism according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0024] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0025] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0026] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0027] Traditional music fountain systems often lack interaction with tourists, have a single performance mode, and cannot meet tourists' viewing needs. Therefore, an intelligent control system and method for music fountains based on smart tourism is expected.

[0028] The present application provides a smart control system for a music fountain based on smart tourism. The smart control system of the music fountain uses a position sensing module to collect the location information of tourists, and uses a fountain control module to intelligently adjust the performance mode of the fountain based on the location information of tourists, so that the fountain performance is closer to tourists and enhances the tourists' sense of participation and experience. Among them, in a specific example of the present application, the position sensing module is further used to locate the tourists through a portable bracelet worn by the tourists when entering the scenic area. More specifically, the portable bracelet includes a GPS module. In addition, the music synchronization playback device plays the corresponding music according to the performance mode of the fountain, realizes the perfect fusion of music and fountain performance, and enhances the audio-visual enjoyment of tourists. The lighting control device intelligently adjusts the lighting effect according to the performance mode of the fountain and the rhythm of the music, creating a more colorful visual experience for tourists.

[0029] In the technical solution of the present application, an intelligent control system for a music fountain based on smart tourism is proposed. Figure 1 It is a block diagram of the intelligent control system of the music fountain based on smart tourism according to the embodiment of the present application. Figure 2 : is a system architecture diagram of the music fountain intelligent control system based on smart tourism according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the intelligent control system 300 of the music fountain based on smart tourism includes: a position sensing module 310, which is used to collect the location information of tourists; a fountain control module 320, which is used to determine the performance mode of the fountain based on the location information of the tourists; a music synchronization playback device 330, which is used to play music based on the performance mode of the fountain; a lighting control device 340, which is used to determine the lighting operation program based on the performance mode of the fountain, and provide lighting coordination based on the lighting operation program.

[0030] In particular, the position sensing module 310 is used to collect the position information of the tourists. The position sensing module is further used to locate the tourists through the portable wristbands worn by the tourists when entering the scenic spot, wherein the portable wristbands include a GPS module.

[0031] In particular, the fountain control module 320 is used to determine the performance mode of the fountain based on the location information of the visitor. It should be understood that in the above-mentioned music fountain intelligent control system, the fountain control module is the premise and basis for the functions of the music synchronization playback device and the lighting control device. In particular, in a specific example of the present application, Figure 3As shown, the fountain control module 320 includes: a position marking unit 321, which is used to mark the location information of the tourist on the digital map to obtain a tourist location geographical distribution image; a superpixel segmentation unit 322, which is used to use the superpixel segmentation method to perform image segmentation on the tourist location geographical distribution image to obtain a sequence of tourist geographical distribution superpixel blocks; a tourist geographical distribution feature processing unit 323, which is used to extract tourist geographical distribution features and enhance tourist geographical distribution features from the sequence of tourist geographical distribution superpixel blocks to obtain a sequence of enhanced tourist geographical distribution superpixel block granularity feature vectors; a clustering analysis unit 324, which is used to input the sequence of enhanced tourist geographical distribution superpixel block granularity feature vectors into a feature distribution reverse mask clustering analysis network to obtain a tourist distribution aggregation representation feature vector; a performance mode determination unit 325, which is used to obtain a determination result based on the tourist distribution aggregation representation feature vector, and the determination result is used to represent the performance mode of the fountain.

[0032] Specifically, the location marking unit 321 is used to mark the location information of the tourists on the digital map to obtain a geographical distribution image of the tourist location. Among them, the digital map is a map form that uses computer technology to store, manage and display geographic spatial data. Compared with traditional paper maps, digital maps have the advantages of digital storage, strong interactivity, real-time update and multi-scale display. In particular, visually displaying the location information of tourists on the digital map can enable the model to intuitively understand the distribution and flow characteristics of tourists, thereby providing an important data basis for the subsequent adjustment of the fountain performance mode to adapt to the changes in tourists.

[0033] Specifically, the superpixel segmentation unit 322 is used to perform image segmentation on the tourist location geographic distribution image using the superpixel segmentation method to obtain a sequence of tourist geographic distribution superpixel blocks. The superpixel segmentation method is an image segmentation technology that aims to divide an image into multiple superpixel blocks, each of which has similar visual features such as color, texture, brightness, etc., and there are obvious boundaries between blocks. In particular, the superpixel segmentation method is different from ordinary pixel clustering, and it pays more attention to maintaining the boundary integrity of the image and the uniformity of the region. In the technical solution of the present application, the tourist location geographic distribution image can be divided into a series of more meaningful data units through superpixel segmentation. In other words, each of the tourist geographic distribution superpixel blocks may have more consistent tourist location information, which helps the system identify and respond to the geographic distribution pattern of tourists.

[0034] Specifically, the tourist geographic distribution feature processing unit 323 is used to extract tourist geographic distribution features and enhance tourist geographic distribution features from the sequence of tourist geographic distribution superpixel blocks to obtain a sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors. In particular, in a specific example of the present application, Figure 4 As shown, the tourist geographic distribution feature processing unit 323 includes: a hole encoding subunit 3231, which is used to pass each tourist geographic distribution superpixel block in the sequence of tourist geographic distribution superpixel blocks through a tourist geographic distribution feature extractor based on a hole convolutional neural network model to obtain a sequence of tourist geographic distribution superpixel block granularity feature vectors; a feature enhancement subunit 3232, which is used to input the sequence of tourist geographic distribution superpixel block granularity feature vectors into a saliency enhancement module based on a global context to obtain a sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors.

[0035] More specifically, the hole encoding subunit 3231 is used to pass each tourist geographic distribution superpixel block in the sequence of tourist geographic distribution superpixel blocks through a tourist geographic distribution feature extractor based on a hole convolutional neural network model to obtain a sequence of tourist geographic distribution superpixel block granularity feature vectors. In the technical solution of the present application, each tourist geographic distribution superpixel block in the sequence of tourist geographic distribution superpixel blocks is passed through a tourist geographic distribution feature extractor based on a hole convolutional neural network model to obtain a sequence of tourist geographic distribution superpixel block granularity feature vectors. That is, the feature learning ability of the hole convolutional neural network model is used to capture the geographical distribution characteristics in each tourist geographic distribution superpixel block. It is worth mentioning that the hole convolutional neural network is a convolutional neural network structure that increases the receptive field of the convolutional layer by introducing holes, thereby expanding the receptive field range of the network while keeping the number of parameters relatively small. Specifically, in traditional convolution operations, the sliding step between the convolution kernel and the input feature map is 1, while in dilated convolution, the convolution kernel contains intermittent zero-value elements, which are called "holes" or "voids". Through these holes, the receptive field can be expanded without increasing the number of parameters. The receptive field refers to the response range of each layer of neurons in a neural network to the local area of ​​the input. By increasing the dilation rate, dilated convolution can expand the receptive field, enabling the network to better capture the long-range dependencies between input features.

[0036] More specifically, the feature enhancement subunit 3232 is used to input the sequence of the tourist geographic distribution superpixel block granularity feature vectors into the saliency enhancement module based on the global context to obtain the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors. Considering that not all tourist geographic distribution superpixel blocks (there may be interference information such as background and noise) have the same importance for expressing the geographical distribution characteristics of tourists, therefore, in the technical solution of the present application, it is expected that the sequence of the tourist geographic distribution superpixel block granularity feature vectors is input into the saliency enhancement module based on the global context to obtain the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors. In particular, the global context-based saliency enhancement module calculates the maximum value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of tourist geographic distribution superpixel block granularity feature vectors as the saliency information of the sequence of tourist geographic distribution superpixel block granularity feature vectors, and calculates the average value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of tourist geographic distribution superpixel block granularity feature vectors as the global information of the sequence of tourist geographic distribution superpixel block granularity feature vectors, and then uses one-dimensional convolution processing and nonlinear activation processing to fuse the saliency information and the global information and convert them into weights for feature distribution adjustment and enhancement, thereby achieving global enhancement and saliency enhancement of the sequence of tourist geographic distribution superpixel block granularity feature vectors, so that the model can more accurately identify and understand the geographical distribution pattern of tourists.

[0037] Specifically, in an embodiment of the present application, the sequence of the tourist geographic distribution superpixel block granularity feature vectors is input into a global context-based saliency enhancement module to obtain the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors, including: calculating the maximum value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of the tourist geographic distribution superpixel block granularity feature vector to obtain a tourist geographic distribution superpixel block granularity significant feature vector composed of each of the maximum values; calculating the average value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of the tourist geographic distribution superpixel block granularity feature vector to obtain a tourist geographic distribution superpixel block granularity global feature vector composed of each of the average values; performing neighborhood encoding and activation on the tourist geographic distribution superpixel block granularity significant feature vector and the tourist geographic distribution superpixel block granularity global feature vector to obtain a tourist geographic distribution superpixel block granularity significant feature context-associated activation feature vector and a tourist geographic distribution superpixel block granularity global feature context-associated activation feature vector; fusing After obtaining the tourist geographic distribution superpixel block granularity significant feature context-associated activation feature vector and the tourist geographic distribution superpixel block granularity global feature context-associated activation feature vector to obtain the tourist geographic distribution superpixel block granularity significant-global feature context-associated activation feature vector, the tourist geographic distribution superpixel block granularity significant-global feature context-associated activation feature vector is respectively subjected to the tanh activation function and the sigmoid activation function, and the two obtained vectors are point multiplied to obtain the tourist geographic distribution superpixel block granularity significant-global feature context-associated weight feature vector; using the tourist geographic distribution superpixel block granularity significant-global feature context-associated weight feature vector as the weight vector, each tourist geographic distribution superpixel block granularity feature vector in the sequence of the tourist geographic distribution superpixel block granularity feature vector is point multiplied by position, and the sequence obtained by point multiplication by position is added element by element to the sequence of the tourist geographic distribution superpixel block granularity feature vector to obtain the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vector.

[0038] Among them, the significant feature vector of the superpixel block granularity of the tourist geographic distribution and the global feature vector of the superpixel block granularity of the tourist geographic distribution are subjected to neighborhood encoding and activation to obtain the significant feature context-associated activation feature vector of the superpixel block granularity of the tourist geographic distribution and the global feature context-associated activation feature vector of the superpixel block granularity of the tourist geographic distribution, including: performing one-dimensional convolution encoding, nonlinear activation and linear transformation on the significant feature vector of the superpixel block granularity of the tourist geographic distribution to obtain the significant feature context-associated activation feature vector of the superpixel block granularity of the tourist geographic distribution; performing one-dimensional convolution encoding, nonlinear activation and linear transformation on the global feature vector of the superpixel block granularity of the tourist geographic distribution to obtain the global feature context-associated activation feature vector of the superpixel block granularity of the tourist geographic distribution.

[0039] In summary, in the above embodiment, the sequence of the superpixel block granularity feature vectors of the geographic distribution of tourists is input into the saliency enhancement module based on the global context to obtain the sequence of the enhanced superpixel block granularity feature vectors of the geographic distribution of tourists, including: applying the saliency enhancement module based on the global context to enhance the sequence of the superpixel block granularity feature vectors of the geographic distribution of tourists with the following enhancement formula to obtain the sequence of the enhanced superpixel block granularity feature vectors of the geographic distribution of tourists; wherein the enhancement formula is:

[0040]

[0041]

[0042]

[0043]

[0044]

[0045] in, is the sequence of superpixel block granularity feature vectors of the tourist geographic distribution, is the global feature vector of the superpixel granularity of the geographical distribution of tourists, is the superpixel-granularity salient feature vector of the geographical distribution of tourists, It means taking the average value, Indicates taking the maximum value, represents a one-dimensional convolution operation, express Activation function, and are the first weight matrix and the second weight matrix for linear transformation, and They are the global feature context-related activation feature vector at the superpixel block granularity of the geographical distribution of tourists and the salient feature context-related activation feature vector at the superpixel block granularity of the geographical distribution of tourists, express function, express Activation function, represents dot product, is a sequence of super-pixel block granularity feature vectors of the enhanced tourist geographic distribution.

[0046] It is worth mentioning that in other specific examples of the present application, the sequence of tourist geographic distribution superpixel blocks can also be used to extract tourist geographic distribution features and enhance tourist geographic distribution features in other ways to obtain a sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors, for example: input the sequence of tourist geographic distribution superpixel blocks; perform feature extraction on each superpixel block, and various methods such as convolutional neural networks and recurrent neural networks can be used to extract the geographic distribution features of each superpixel block; organize the geographic distribution features of each superpixel block into a sequence to form a feature sequence, which represents the feature sequence of the entire geographic distribution; perform feature enhancement on the feature sequence, and techniques such as attention mechanism, graph convolutional neural network, etc. can be used to enhance the correlation and importance between features; the enhanced feature sequence is converted into a granularity feature vector sequence to obtain the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors.

[0047] Specifically, the cluster analysis unit 324 is used to input the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors into the feature distribution reverse mask cluster analysis network to obtain the tourist distribution aggregate representation feature vector. Among them, the feature distribution reverse mask cluster analysis network can enhance the model's recognition ability for key features by suppressing unimportant variability features, thereby improving the model's robustness to noise and outliers, and aggregate and interact important key features to form a comprehensive representation of tourist geographic distribution features. More specifically, the feature distribution reverse mask cluster analysis network clusters the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors to obtain a feature representation representing the overall tourist geographic distribution features, that is, the tourist geographic distribution superpixel block cluster center vector; then calculate the semantic correlation coefficient between each enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector to quantify the correlation and correlation between each tourist geographic distribution superpixel block and the cluster center, and suppress and eliminate the feature distribution with low correlation and correlation with the cluster center, thereby retaining the important feature information therein.

[0048] Specifically, in an embodiment of the present application, the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors is input into a feature distribution reverse mask clustering analysis network to obtain a tourist distribution aggregate representation feature vector, including: calculating the position mean vector of the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors to cluster the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors to obtain a tourist geographic distribution superpixel block cluster center vector; calculating the semantic correlation coefficient between each enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector to obtain a sequence of semantic correlation coefficients; taking the inverse of each semantic correlation coefficient in the sequence of the semantic correlation coefficient to obtain a sequence of semantic anti-correlation coefficients; using S The igmoid function normalizes the sequence of semantic anti-correlation coefficients to obtain a sequence of semantic anti-correlation weight coefficients; using each semantic anti-correlation weight coefficient in the sequence of semantic anti-correlation weight coefficients as a weight, weighting each enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors to obtain a sequence of tourist geographic distribution superpixel block suppression feature vectors; calculating the positional difference between the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors and the sequence of tourist geographic distribution superpixel block suppression feature vectors to obtain a sequence of cluster-optimized tourist geographic distribution superpixel block granularity feature vectors; calculating the positional mean vector of the sequence of cluster-optimized tourist geographic distribution superpixel block granularity feature vectors to obtain the tourist distribution aggregation representation feature vector.

[0049] Among them, the semantic correlation coefficient between each enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector is calculated to obtain a sequence of semantic correlation coefficients, including: dividing the dot product result between the transposed vector of the i-th enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector by the dot product result between the binary norm of the i-th enhanced tourist geographic distribution superpixel block granularity feature vector and the binary norm of the tourist geographic distribution superpixel block cluster center vector to obtain the i-th semantic correlation coefficient.

[0050] In summary, in the above embodiment, the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors is input into the feature distribution reverse mask clustering analysis network to obtain the tourist distribution aggregate representation feature vector, including: using the feature distribution reverse mask clustering analysis network to aggregate the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors according to the following formula to obtain the tourist distribution aggregate representation feature vector; wherein the formula is:

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059] in, is a sequence of superpixel-block granularity feature vectors that reinforce the geographical distribution of tourists, , , and are respectively the first enhanced tourist geographic distribution superpixel block granularity feature vector, the second enhanced tourist geographic distribution superpixel block granularity feature vector, the i-th enhanced tourist geographic distribution superpixel block granularity feature vector and the N-th enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vector, The value of is the number of enhanced tourist geographic distribution superpixel block granularity feature vectors in the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors, is the cluster center vector of the superpixel block of the tourist geographic distribution, represents the transpose of a vector, is the i-th semantic correlation coefficient, represents the two-norm of a vector, is the i-th semantic anti-correlation coefficient, is the i-th semantic anti-correlation weight coefficient, represents an exponential function with a natural constant as base, is a sequence of suppressed feature vectors of superpixel patches of tourists’ geographic distribution, , , and are respectively the first tourist geographic distribution superpixel block suppression feature vector, the second tourist geographic distribution superpixel block suppression feature vector, the i-th tourist geographic distribution superpixel block suppression feature vector and the N-th tourist geographic distribution superpixel block suppression feature vector in the sequence of the tourist geographic distribution superpixel block suppression feature vector, is a sequence of superpixel-block granularity feature vectors of clustering optimization of tourists’ geographic distribution, , , and They are the first clustering optimization tourist geographic distribution superpixel block granularity feature vector, the second clustering optimization tourist geographic distribution superpixel block granularity feature vector, the i-th clustering optimization tourist geographic distribution superpixel block granularity feature vector and the N-th clustering optimization tourist geographic distribution superpixel block granularity feature vector. is the aggregated representation feature vector of the visitor distribution.

[0060] Specifically, the performance mode determination unit 325 is used to obtain a determination result based on the tourist distribution aggregation representation feature vector, and the determination result is used to represent the performance mode of the fountain. In particular, in a specific example of the present application, the tourist distribution aggregation representation feature vector is input into a classifier-based performance mode determiner to obtain a determination result, and the determination result is used to represent the performance mode of the fountain.

[0061] In a preferred embodiment, taking into account the image semantic features of the local image semantic space domain of the tourist location geographic distribution image expressed by the sequence of superpixel block granularity feature vectors, after saliency enhancement and reverse mask clustering based on the local image semantic space domain context under the global image semantic space domain, the obtained tourist distribution aggregate representation feature vector will also lack the class regression semantic consistent aggregate expression relative to the source image semantic feature representation due to the difference in global context saliency weights and reverse mask clustering weights under the image semantic feature expression between local image semantic space domains, thereby affecting the accuracy of the classification results.

[0062] Based on this, the present application considers optimizing the tourist distribution aggregate representation feature vector when passing the tourist distribution aggregate representation feature vector through a performance mode determiner based on a classifier to obtain a determination result, that is, inputting the tourist distribution aggregate representation feature vector into a performance mode determiner based on a classifier to obtain a determination result, including

[0063] Perform feature clustering on all feature values ​​of the tourist distribution aggregate representation feature vector to obtain a feature set within the tourist distribution aggregate representation class And the tourist distribution aggregation represents the out-of-class feature set ,in, represents the distribution of tourists, and the aggregation represents the feature set within the class. The distribution of tourists is aggregated to represent the first eigenvalues, The distribution of tourists is aggregated to represent the out-of-class feature set. eigenvalues;

[0064] Calculate the ratio of the number of eigenvalues ​​in the feature set within the tourist distribution aggregate representation class to the number of eigenvalues ​​of all features of the tourist distribution aggregate representation feature vector ,in, represents the number of feature values ​​in the feature set within the cluster of tourist distribution. represents the number of eigenvalues ​​of all features of the feature vector representing the aggregate distribution of the tourists, It indicates the ratio;

[0065] Calculate the sum of the absolute values ​​of all eigenvalues ​​in the eigenvector within the cluster representing the distribution of tourists The power of the aggregated distribution of tourists represents the sum of the absolute values ​​of all eigenvalues ​​in the eigenvector The ratio of the powers is used to obtain the weight of the first visitor distribution aggregation representation;

[0066] Calculate the sum of squares of all eigenvalues ​​in the eigenvector of the clustered representation of the visitor distribution The power of the aggregated distribution of visitors represents the sum of the squares of all eigenvalues ​​in the eigenvector The ratio of the powers is used to obtain the weight of the second visitor distribution aggregation representation;

[0067] That is, the first visitor distribution aggregate represents the weight and the second visitor distribution aggregate represents the weight It is expressed as:

[0068]

[0069]

[0070] in, represents the eigenvalues ​​of the characteristic vector representing the aggregate distribution of tourists, represents the weight of the first visitor distribution aggregation, represents the weight of the second visitor distribution aggregation;

[0071] Then, for each feature value in the feature set within the tourist distribution aggregate representation class, calculate the product of the feature value and the weight of the second tourist distribution aggregate representation plus the weight of the first tourist distribution aggregate representation to obtain the optimized feature value ,in, represents the first feature set within the class based on the distribution of tourists Optimized eigenvalues;

[0072] For each feature value in the out-of-class feature set of the tourist distribution aggregate representation, calculate the product of the feature value and the weight of the first tourist distribution aggregate representation to obtain the optimized feature value ,in, represents the first out-of-class feature set based on the tourist distribution aggregation Optimized eigenvalues;

[0073] Combining the optimized feature values ​​based on the feature set within the tourist distribution aggregation representation class and the feature set outside the tourist distribution aggregation representation class into an optimized tourist distribution aggregation representation feature vector;

[0074] The optimized visitor distribution aggregation representation feature vector is input into a classifier-based performance mode determiner to obtain a determination result.

[0075] That is, while the feature vectors representing the aggregated representation of the tourist distribution are subjected to feature clustering, the interaction description of the key feature information of the feature vectors representing the aggregated representation of the tourist distribution in the clustering process is used to construct a geometric equivariant topology of the features through feature low-rank harmonic modulation based on the equivariance of the clustering features of the feature vectors representing the aggregated representation of the tourist distribution and the features as a whole, so as to obtain the graphical distribution translation and rotation symmetry of the clustering features of the feature vectors representing the aggregated representation of the tourist distribution relative to the features as a whole, thereby introducing geometric message passing into the feature expression of the feature vectors representing the aggregated representation of the tourist distribution, and realizing the cluster mapping symmetry of the feature vectors representing the aggregated representation of the tourist distribution through manipulation of irreducible low-rank coefficients, thereby improving the consistency of the clustering-based feature representation of the feature vectors representing the aggregated representation of the tourist distribution, thereby improving the accuracy of the determination result obtained by inputting the feature vectors representing the aggregated representation of the tourist distribution into the performance mode determiner based on the classifier.

[0076] In particular, the music synchronization playback device 330 and the lighting control device 340 are used to play music based on the performance mode of the fountain; and determine the lighting operation program based on the performance mode of the fountain, and provide lighting coordination based on the lighting operation program. It should be understood that through the music synchronization playback device, it is possible to select music suitable for the fountain performance according to the determined performance mode, taking into account the rhythm, beat and audio effects to ensure that the music is coordinated with the fountain performance action. Among them, the music synchronization playback device can be a plurality of audio players or a specific music playback system to ensure that the device can accurately and synchronously play the audio. Further, the lighting operation program is determined based on the performance mode of the fountain, and lighting coordination is provided based on the lighting operation program. In a specific example, factors such as the brightness, color, and change mode of the light can be considered to ensure that the light can perfectly coordinate with the music and the fountain action.

[0077] As described above, the music fountain intelligent control system 300 based on smart tourism according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a music fountain intelligent control algorithm based on smart tourism. In a possible implementation, the music fountain intelligent control system 300 based on smart tourism according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the music fountain intelligent control system 300 based on smart tourism can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the music fountain intelligent control system 300 based on smart tourism can also be one of the many hardware modules of the wireless terminal.

[0078] Alternatively, in another example, the smart tourism-based music fountain intelligent control system 300 and the wireless terminal may also be separate devices, and the smart tourism-based music fountain intelligent control system 300 may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0079] Furthermore, a method for intelligent control of a music fountain based on smart tourism is also provided.

[0080] Figure 5 Flow chart of the intelligent control method of the music fountain based on smart tourism according to the embodiment of the present application. Figure 5 As shown, the intelligent control method of the music fountain based on smart tourism according to the embodiment of the present application includes the following steps: S1, collecting the location information of tourists; S2, determining the performance mode of the fountain based on the location information of the tourists; S3, playing music based on the performance mode of the fountain; S4, determining the lighting operation program based on the performance mode of the fountain, and providing lighting coordination based on the lighting operation program.

[0081] In summary, the intelligent control method of a music fountain based on smart tourism according to the embodiment of the present application is explained, which captures and learns the geographical distribution characteristics of tourists' locations by adopting a deep learning algorithm based on artificial intelligence and computer vision technology, and introduces a global context saliency enhancement module to enhance the expression of geographical distribution characteristics, and then uses the reverse masking mechanism to cluster and fuse the enhanced geographical distribution characteristics of tourists, thereby forming a comprehensive representation of the geographical distribution characteristics of tourists, and finally uses a classifier to construct a performance mode determiner to intelligently determine the performance mode of the fountain according to the real-time geographical location information of tourists.

[0082] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A music fountain intelligent control system based on smart tourism, characterized in that: include: A location sensor module is used to collect the location information of tourists; A fountain control module, used for determining a performance mode of the fountain based on the location information of the visitor; A music synchronization playing device, used to play music based on the performance mode of the fountain; A lighting control device, used to determine a lighting operation program based on the performance mode of the fountain, and provide lighting coordination based on the lighting operation program; Wherein, the fountain control module includes: A location marking unit, used for marking the location information of the tourists on a digital map to obtain a geographical distribution image of the tourists' locations; A superpixel segmentation unit, configured to perform image segmentation on the tourist location geographic distribution image using a superpixel segmentation method to obtain a sequence of tourist geographic distribution superpixel blocks; A tourist geographic distribution feature processing unit, used for extracting tourist geographic distribution features and enhancing tourist geographic distribution features from the sequence of tourist geographic distribution superpixel blocks to obtain a sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors; A clustering analysis unit, used for inputting the sequence of superpixel block granularity feature vectors of the enhanced tourist geographic distribution into a feature distribution reverse mask clustering analysis network to obtain a tourist distribution aggregate representation feature vector; a performance mode determination unit, configured to obtain a determination result based on the visitor distribution aggregation representation feature vector, wherein the determination result is used to represent the performance mode of the fountain; Wherein, the performance mode determination unit is used to: Inputting the tourist distribution aggregate representation feature vector into a classifier-based performance mode determiner to obtain the determination result; The tourist distribution aggregation representation feature vector is input into a classifier-based performance mode determiner to obtain a determination result, including: Perform feature clustering on all feature values ​​of the tourist distribution aggregate representation feature vector to obtain a feature set within the tourist distribution aggregate representation class And the tourist distribution aggregation represents the out-of-class feature set in, represents the distribution of tourists, and the aggregation represents the feature set within the class. The distribution of tourists is aggregated to represent the first eigenvalues, The distribution of tourists is aggregated to represent the out-of-class feature set. eigenvalues; Calculate the ratio of the number of eigenvalues ​​in the feature set within the tourist distribution aggregate representation class to the number of eigenvalues ​​of all features of the tourist distribution aggregate representation feature vector in, represents the number of feature values ​​in the feature set within the cluster of tourist distribution. represents the number of eigenvalues ​​of all features of the feature vector representing the aggregate distribution of the tourists, It indicates the ratio; Calculate the sum of the absolute values ​​of all eigenvalues ​​in the eigenvector within the cluster representing the distribution of tourists The power of the aggregated distribution of tourists represents the sum of the absolute values ​​of all eigenvalues ​​in the eigenvector The ratio of the powers is used to obtain the weight of the first visitor distribution aggregation representation; Calculate the sum of squares of all eigenvalues ​​in the eigenvector of the clustered representation of the visitor distribution The power of the aggregated distribution of visitors represents the sum of the squares of all eigenvalues ​​in the eigenvector The ratio of the powers is used to obtain the weight of the second visitor distribution aggregation representation; That is, the first visitor distribution aggregate represents the weight and the second visitor distribution aggregate represents the weight It is expressed as: in, represents the eigenvalues ​​of the characteristic vector representing the aggregate distribution of tourists, represents the weight of the first visitor distribution aggregation, represents the weight of the second visitor distribution aggregation; Then, for each feature value in the feature set within the tourist distribution aggregate representation class, calculate the product of the feature value and the weight of the second tourist distribution aggregate representation plus the weight of the first tourist distribution aggregate representation to obtain the optimized feature value in, represents the first feature set within the class based on the distribution of tourists Optimized eigenvalues; For each feature value in the out-of-class feature set of the tourist distribution aggregate representation, calculate the product of the feature value and the weight of the first tourist distribution aggregate representation to obtain the optimized feature value in, represents the first out-of-class feature set based on the tourist distribution aggregation Optimized eigenvalues; Composing an optimized tourist distribution aggregation representation feature vector based on the optimized feature values ​​of the tourist distribution aggregation representation class feature set and the tourist distribution aggregation representation class feature set; The optimized visitor distribution aggregate representation feature vector is input into a classifier-based performance mode determiner to obtain a determination result.

2. According to the intelligent control system of music fountain based on smart tourism in claim 1, it is characterized in that: The position sensing module is further used to locate the tourist through a portable bracelet worn by the tourist when entering the scenic area, wherein the portable bracelet includes a GPS module.

3. The intelligent control system of the music fountain based on smart tourism according to claim 2 is characterized in that: The tourist geographic distribution feature processing unit comprises: A hole encoding subunit, used for respectively passing each tourist geographic distribution superpixel block in the sequence of tourist geographic distribution superpixel blocks through a tourist geographic distribution feature extractor based on a hole convolutional neural network model to obtain a sequence of tourist geographic distribution superpixel block granularity feature vectors; The feature enhancement subunit is used to input the sequence of the super-pixel block granularity feature vectors of the geographic distribution of tourists into the saliency enhancement module based on the global context to obtain the sequence of enhanced super-pixel block granularity feature vectors of the geographic distribution of tourists.

4. The intelligent control system of the music fountain based on smart tourism according to claim 3 is characterized in that: The feature enhancement subunit comprises: A secondary subunit for extracting significant features is used to calculate the maximum value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of the tourist geographic distribution superpixel block granularity feature vector to obtain a tourist geographic distribution superpixel block granularity significant feature vector composed of each of the maximum values; A global feature extraction secondary subunit is used to calculate the average value of each tourist geographic distribution superpixel block granularity feature vector in the sequence of the tourist geographic distribution superpixel block granularity feature vector to obtain a tourist geographic distribution superpixel block granularity global feature vector composed of each of the average values; A neighborhood coding activation secondary subunit is used to perform neighborhood coding and activation on the tourist geographic distribution superpixel block granularity significant feature vector and the tourist geographic distribution superpixel block granularity global feature vector to obtain the tourist geographic distribution superpixel block granularity significant feature context-related activation feature vector and the tourist geographic distribution superpixel block granularity global feature context-related activation feature vector; A weight generation secondary subunit is used to fuse the tourist geographic distribution superpixel block granularity significant feature context association activation feature vector and the tourist geographic distribution superpixel block granularity global feature context association activation feature vector to obtain the tourist geographic distribution superpixel block granularity significant-global feature context association activation feature vector, and then respectively pass the tourist geographic distribution superpixel block granularity significant-global feature context association activation feature vector through a tanh activation function and a sigmoid activation function, and perform dot multiplication of the two obtained vectors to obtain a tourist geographic distribution superpixel block granularity significant-global feature context association weight feature vector; The weight application and enhancement secondary sub-unit is used to use the tourist geographic distribution superpixel block granularity significance-global feature context association weight feature vector as the weight vector, perform positional dot multiplication on each tourist geographic distribution superpixel block granularity feature vector in the sequence of tourist geographic distribution superpixel block granularity feature vectors, and add the sequence obtained by positional dot multiplication to the sequence of tourist geographic distribution superpixel block granularity feature vectors element by element to obtain the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors.

5. The intelligent control system of the music fountain based on smart tourism according to claim 4 is characterized in that: The neighborhood coding activates the secondary subunit, which is used to: Performing one-dimensional convolution coding, nonlinear activation and linear transformation on the super-pixel block granularity significant feature vector of the tourist geographic distribution to obtain the context-related activation feature vector of the super-pixel block granularity significant feature of the tourist geographic distribution; One-dimensional convolution coding, nonlinear activation and linear transformation are performed on the global feature vector of the superpixel block granularity of the tourist geographic distribution to obtain the global feature context-associated activation feature vector of the superpixel block granularity of the tourist geographic distribution.

6. The intelligent control system of music fountain based on smart tourism according to claim 5 is characterized in that: The cluster analysis unit comprises: A cluster center representation subunit is used to calculate the position mean vector of the sequence of the super pixel block granularity feature vectors of the enhanced tourist geographic distribution to cluster the sequence of the super pixel block granularity feature vectors of the enhanced tourist geographic distribution to obtain the tourist geographic distribution super pixel block cluster center vector; A semantic correlation quantization subunit, used for calculating the semantic correlation coefficient between each enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector to obtain a sequence of semantic correlation coefficients; an anti-correlation coefficient calculation subunit, used for taking the reciprocal of each semantic correlation coefficient in the sequence of semantic correlation coefficients to obtain a sequence of semantic anti-correlation coefficients; A normalization subunit, used for normalizing the sequence of semantic anti-correlation coefficients using a Sigmoid function to obtain a sequence of semantic anti-correlation weight coefficients; An information suppression subunit is used to weight each enhanced tourist geographic distribution superpixel block granularity feature vector in the sequence of enhanced tourist geographic distribution superpixel block granularity feature vectors by using each semantic anti-correlation weight coefficient in the sequence of semantic anti-correlation weight coefficients as a weight to obtain a sequence of tourist geographic distribution superpixel block suppression feature vectors; A reverse optimization subunit, used for calculating the position difference between the sequence of the enhanced tourist geographic distribution superpixel block granularity feature vectors and the sequence of the tourist geographic distribution superpixel block suppression feature vectors to obtain a sequence of cluster-optimized tourist geographic distribution superpixel block granularity feature vectors; The aggregation representation subunit is used to calculate the position mean vector of the sequence of super-pixel block granularity feature vectors of the clustering optimization tourist geographic distribution to obtain the tourist distribution aggregation representation feature vector.

7. The intelligent control system of music fountain based on smart tourism according to claim 6 is characterized in that: The semantic relevance quantification subunit is used to: The i-th semantic correlation coefficient is obtained by dividing the dot product result between the transposed vector of the i-th enhanced tourist geographic distribution superpixel block granularity feature vector and the tourist geographic distribution superpixel block cluster center vector by the dot product result between the bi-norm of the i-th enhanced tourist geographic distribution superpixel block granularity feature vector and the bi-norm of the tourist geographic distribution superpixel block cluster center vector.

8. A method for intelligently controlling a music fountain based on smart tourism, using the intelligent control system for a music fountain based on smart tourism according to any one of claims 1 to 7, characterized in that: include: Collecting visitors’ location information; Determining a performance mode of the fountain based on the location information of the visitor; Playing music based on the performance mode of the fountain; A lighting operation program is determined based on the performance mode of the fountain, and lighting coordination is provided based on the lighting operation program.

Images