CN115470418A - Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography - Google Patents
Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography Download PDFInfo
- Publication number
- CN115470418A CN115470418A CN202211124537.XA CN202211124537A CN115470418A CN 115470418 A CN115470418 A CN 115470418A CN 202211124537 A CN202211124537 A CN 202211124537A CN 115470418 A CN115470418 A CN 115470418A
- Authority
- CN
- China
- Prior art keywords
- queuing
- target detection
- time
- detection point
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Evolutionary Biology (AREA)
- Physiology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of data prediction, in particular to a queuing point recommendation method and a system based on unmanned aerial vehicle aerial photography, through using a YOLOv5 algorithm as a real-time target detection algorithm, compared with other algorithms, as a single-stage algorithm, the running speed of the algorithm is higher, the average detection precision is higher, a target detection point queuing condition of a future time point is predicted through a Markov prediction model, the number of queuing people and the speed of the queuing people after a period of time from a user departure point to a target detection point through a certain travel mode can be predicted, the accuracy and the practicability of the embodiment can be improved, in the case of counting the number of queuing people with a plurality of target detection points nearby, the most preferable item can be selected for the user to select according to the existing data and historical data, and the user can conveniently make a travel plan.
Description
Technical Field
The invention relates to the technical field of data prediction, in particular to a queuing point recommendation method and system based on unmanned aerial vehicle aerial photography.
Background
In recent years, with the rapid development of target detection, data processing and real-time sharing technologies, technologies of shooting specific scenes such as classrooms, offices and construction site security detection and processing images by using a high-performance server through fixed shooting equipment are mature, for example, chinese patent CN202110301889.7 discloses a subway platform guidance queuing system based on YOLOv3 face detection, which obtains passenger queuing images of waiting areas of each platform of a subway through a face detection subsystem, performs face real-time detection by using a YOLOv3 algorithm, matches the information of the number of queuing people with each waiting area, and visually displays the number of queuing people corresponding to each matched waiting area of the platform, but the counting of outdoor scenes in the prior art is not enough due to the influence of multiple factors, for example, when needing to queue for nucleic acid detection, if residents can determine the time for themselves to go to nucleic acid detection points by knowing the number of queuing people at the nucleic acid detection points, a lot of time can be saved. For another example, in a stadium (track and field ground), a certain sports team committee wants to know the number of people on the ground, but for dense occasions, the naked-eye counting efficiency is extremely low, and the flexibility of a fixed station is poor, so that a method for counting unmanned aerial vehicles in real time is available today, for example, CN202110793315.6 is an unmanned aerial vehicle platform people counting method and system based on image deep learning, but a feasible method for predicting the number of people in line is still lacking.
Disclosure of Invention
In view of the above, the invention aims to provide a queuing point recommendation method and system based on unmanned aerial vehicle aerial photography, so as to solve the problem that the prior art lacks a method for predicting the number of queuing people in an outdoor scene.
Based on the aim, the invention provides a queuing point recommendation method based on unmanned aerial vehicle aerial photography, which comprises the following steps:
intercepting a picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
counting the crowd of the queuing images of the detection points by using a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
clustering the queuing data of each target detection point by using dynamic time warping as distance measurement;
inputting the clustered historical queuing data into a Markov original model, and training the original model to obtain a prediction model;
calculating the time when the user reaches the target detection point according to the time designated by the user, the designated position and the position of the target detection point;
inputting the current queuing data into a Markov prediction model, predicting the number of queuing people when the user reaches a target detection point, and predicting the time required by queuing;
calculating the estimated total time consumption according to the time required by queuing and the time spent by the user to come and go to the target detection point;
and comparing the predicted total time consumption of different target detection points selected by the user, and outputting the target detection point corresponding to the minimum predicted total time consumption as the optimal detection point.
Preferably, the method further comprises:
acquiring a preference factor of a user;
after calculating the estimated total elapsed time, the method is replaced by:
calculating recommendation degrees of different target detection points according to the preference factors of the user and the calculated predicted total time consumption;
and comparing the recommendation degrees of the target detection points, and outputting the target detection point with the highest recommendation degree as an optimal detection point.
Preferably, the acquiring of the preference factor of the user comprises:
and obtaining the travel willingness coefficients of the user under the conditions of different terrains, different weather and different time periods in a questionnaire survey or historical data analysis mode.
Preferably, the formula for calculating the recommendation degrees of different target detection points is:
wherein m is a travel intention coefficient under the current weather, n is a travel intention coefficient of a corresponding time period at a specified time, and k 1 、k 2 ……k n Will coefficient of travel respectively for different terrains, d 1 、d 2 ……d n Respectively the path lengths of different terrains, d is the total path length between the designated position and the target detection point, t is the predicted total time consumption, r is the correction coefficient, r 1 、r 2 ……r n Respectively d/t corresponding to different sets A 1 、A 2 ……A n Correction factor of time.
Preferably, clustering the queuing data of the target detection points using dynamic time warping as the distance metric comprises:
calculating the intra-cluster and inter-cluster dynamic time warping distance ratio, comprising:
calculating the dynamic time warping distance of any two pieces of historical daily load data in the cluster, and averaging the dynamic time warping distance to obtain the intra-cluster dynamic time warping average distance;
calculating the dynamic time warping distance between any two central points between clusters, and averaging the dynamic time warping distances to obtain the inter-cluster dynamic time warping average distance;
and obtaining the intra-cluster inter-cluster dynamic time warping distance ratio according to the intra-cluster dynamic time warping average distance and the inter-cluster dynamic time warping average distance.
Preferably, the method further comprises:
the method for planning the path of the unmanned aerial vehicle at a plurality of target detection points based on the genetic algorithm comprises the following steps:
a1, adopting binary coding and real number coding, enabling each individual to be a path sequence of a group of detection points, enabling each detection point to be a chromosome, randomly initializing the path sequence by applying np.
A2, defining the fitness as the reciprocal of the distance, and calculating the fitness;
a3, selecting by adopting a roulette wheel, firstly determining the probability of each individual to be selected, dividing the fitness of each individual by the fitness of the total individual by a formula, namely converting each probability into an accumulated probability when p = fi/sigma fi is realized, then randomly generating a number in [0,1], and selecting a corresponding individual according to a relative interval falling in the accumulated probability;
a4, under real number coding, adopting partial mapping intersection to define a random number of [0,1], and when the random number is greater than the intersection probability cross-prob =0.9, intersecting part of chromosomes (a plurality of adjacent detection points), and exchanging upper and lower individual collision parts to solve collision;
a5, adopting flip variation to define a random number of [0,1], and performing variation when the random number is less than the variation probability mut-prob = 0.1;
repeating the steps A2-A5 until the termination condition is met: the fitness values of the optimal individuals in the population are continuous.
Preferably, in the population counting of the queued images of the inspection points, the neural network parameter obtaining method of the YOLOv5 algorithm includes:
preprocessing the input end of the neural network on the queuing image;
slicing the queuing image through a Focus structure of a backhaul, and performing convolution operation on the sliced image to obtain a double-sampling feature map under the condition of no information loss;
carrying out convolution processing on the two times of downsampling feature map through a CBL convolution module, and then carrying out feature learning on the model by utilizing a CSP structure;
processing the characteristic tensor through the FPN + PAN structure of the Neck;
the Head layer predicts the image characteristics, generates a bounding box and a prediction category, and outputs convolution neural network parameters of YOLOv 5.
Preferably, the preprocessing comprises Mosaic data enhancement, adaptive anchor frame calculation and adaptive picture scaling.
Preferably, the method further comprises: and feeding back the number of real-time queuing people, the predicted queuing time of each target detection point and the predicted total consumption time of each target detection point in real time.
This specification still provides a queuing point recommendation system based on unmanned aerial vehicle takes photo by plane, includes:
the image processing module is used for intercepting the picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
the counting module is used for counting the crowd of the queuing images of the detection points through a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
the prediction module is used for clustering queuing data of each target detection point by using dynamic time warping as distance measurement, inputting the clustered historical queuing data into a Markov original model, training the original model to obtain a prediction model, calculating the time when a user reaches the target detection point according to the time designated by the user, the designated position and the position of the target detection point, inputting the current queuing data into the Markov prediction model, predicting the number of queuing people when the user reaches the target detection point, predicting the time required by queuing, and calculating the predicted total consumed time according to the time required by queuing and the time spent by the user to go to and return from the target detection point;
and the recommending module is used for comparing the predicted total time consumption of different target detection points selected by the user and outputting the shortest target detection point of the predicted total time consumption as the optimal detection point.
The invention has the beneficial effects that: the YOLOv5 algorithm is a real-time target detection algorithm, compared with other algorithms, the single-stage algorithm has the advantages that the running speed of the single-stage algorithm is higher, the average detection precision is higher, the target detection point queuing condition of a future time point is predicted through a Markov prediction model, the number of queuing people and the speed after a period of time from a user starting point to a certain trip mode to reach the target detection point can be predicted, the accuracy and the practicability of the example can be improved, in the case of counting the number of queuing people with a plurality of target detection points nearby, the most preferable item can be selected for the user to select according to the existing data and historical data, and the user can conveniently make a journey plan.
Drawings
In order to more clearly illustrate the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a queuing point recommendation method according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method for recommending a queuing point based on a recommendation degree according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a system of an unmanned aerial vehicle, an upper computer and a server according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to specific embodiments below.
It is to be noted that technical terms or scientific terms used herein should have the ordinary meaning as understood by those having ordinary skill in the art to which the present invention belongs, unless otherwise defined. The use of "first," "second," and similar terms in the present application do not denote any order, quantity, or importance, but rather the terms are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", and the like are used only to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
As shown in fig. 1, an embodiment of the present specification provides a queuing point recommendation method based on unmanned aerial vehicle aerial photography, including the following steps:
intercepting a picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
counting the crowd of the queuing images of the detection points by using a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
for example, first, the model of the YOLOv5 algorithm can be basically divided into four parts: input end, backbone, tack, network architecture of the output end.
Input end: firstly, the image is processed (preprocessed) at the input end of the neural network: the method comprises the following steps of Mosaic data enhancement, self-adaptive anchor frame calculation and self-adaptive picture scaling:
the practical methods of use are referenced below:
and Mosaic data enhancement: firstly, 1 picture and three random pictures in a transmitted data set are gathered into 4 pictures, and the pictures are spliced in a random scaling, random cutting and random arrangement mode.
And (3) self-adaptive anchor frame calculation: the preset frames are first framed to target at approximately the possible locations and then adjusted based on these preset frames.
And calculating the position coordinates of the target and the height and width of the frame through the self-adaptive anchor frame. Adaptive picture scaling: and adopting a letterbox self-adaptive picture scaling technology to scale and fill the picture to a specified size. And then, performing feature extraction on the preprocessed tensor through a Focus structure and a CSP structure of the Backbone, and splitting a high-resolution picture (feature map) into a plurality of low-resolution pictures/feature maps by adopting a slicing operation, namely, alternate column sampling and splicing.
Firstly, before a picture enters a backbone in v5, a Focus module performs slicing operation on the picture, specifically, every other pixel in one picture is taken to have a value, which is similar to adjacent downsampling, so that four pictures are taken, the four pictures are complementary and almost as long as each other, but no information is lost, so that W and H information is concentrated in a channel space, an input channel is expanded by 4 times, namely, the spliced picture is changed into 12 channels relative to an original RGB three-channel mode, and finally, an obtained new picture is subjected to convolution operation, and finally, a double downsampling feature map under the condition of no information loss is obtained.
Then, the CBL convolution module: CBL is a convolutional block: consists of three network layers, conv, batch Normalization, leaky ReLU.
The Conv layer is a convolution layer, and the input image is processed by adopting a plurality of different convolution kernels to obtain different response characteristic diagrams.
The BN (Batch Normalization) layer is a Batch Normalization layer, which is a layer of the neural network and is used after the convolution layer before the activation function. When the number of the obtained feature maps is m and the size of the feature maps is w × h (namely, the number of image pixels), the data volume of the BN is m × w × h. The BN layer mainly comprises the steps of calculating the mean value and the variance of all batch data, calculating the difference between a pixel value and the mean value, dividing the difference by the variance for normalization, and adding an offset factor and a scale change factor to control the normalized value, wherein the value of the factor is obtained by learning in the training process of a neural network.
The Leaky ReLU function is a variant of the ReLU function in which the learning speed of the ReLU may become slow when the input is negative, even making the neurons directly ineffective, because the input is less than zero and the gradient is zero, so that its weights cannot be updated, and will remain silent during the rest of the training. When the input value of the ReLU is negative, the output is always 0, and the first derivative thereof is also always 0, which may result in that the Neuron cannot update the parameter, i.e. the Neuron does not learn, and this phenomenon is called "Dead Neuron". To reduce the occurrence of silent neurons, allowing gradient-based learning (albeit slow), and to address this drawback of the ReLU function, a leak (leak) value is introduced in the negative half of the ReLU function, so called the leak ReLU function, whose output has a small slope to the negative input. Because the derivative is not zero, the problem that the neuron does not learn after the ReLU function enters a negative interval is solved.
In the CSP structure, the original input is divided into two branches, convolution operation is respectively carried out to reduce the number of channels by half, then a branch is carried out with Bottleneck N operation, and then two branches are concat, so that the input and the output of the BottleneckCSP are the same in size, and thus, a model can learn more characteristics.
And then, processing the feature tensor through the FPN + PAN structure of the Neck, combining the FPN + PAN to operate, wherein the FPN layer transmits strong semantic features from top to bottom, the feature pyramid comprises the PAN structure and transmits strong positioning features from bottom to top, and parameter aggregation is performed on different detection layers from different backbone layers.
The Head layer predicts the image characteristics, generates a boundary box and predicts the category
Yolov5 takes the GIOU _ Loss in the classification Loss function. IOU _ Loss: the overlapping area of the detection frame and the target frame is mainly considered.
GIOU _ Loss: on the basis of the IOU, the problem that the bounding boxes are not overlapped is solved. And NMS non-maximum suppression is employed. The GIoU focuses not only on the overlapping region but also on other non-overlapping regions, which can better reflect the contact ratio of the two, and can make the prediction frame and the real frame approach each other by minimizing the GIoU _ Loss.
Inputting: two arbitrary convex types A, B belongs to S n ，
And (3) outputting: GIoU, finding the smallest convex shape C containing a, B,
GIoU _ Loss function:
IoU function:
and clustering the queuing data of the target detection points by using the dynamic time warping as a distance measure, and comparing the similarity of the queuing data.
For example, calculating the intra-cluster and inter-cluster dynamic time warping distance ratio includes:
calculating the dynamic time warping distance of any two pieces of historical data in the cluster, and averaging the dynamic time warping distance to obtain the intra-cluster dynamic time warping average distance;
calculating the dynamic time warping distance between any two central points between clusters, and averaging the dynamic time warping distances to obtain the inter-cluster dynamic time warping average distance;
and obtaining the intra-cluster inter-cluster dynamic time warping distance ratio according to the intra-cluster dynamic time warping average distance and the inter-cluster dynamic time warping average distance.
Inputting the clustered historical queuing data into a Markov original model, and training the original model to obtain a prediction model M1; the construction of the prediction model of M1 mainly has the following characteristics:
the unmanned aerial vehicle comprises an acquisition module, a storage module and a display module, wherein an imaging system of the unmanned aerial vehicle is used for acquiring and outputting multiple pieces of historical queuing data;
the clustering module is used for clustering the historical daily load data by using dynamic time warping as distance measurement;
the training module is used for inputting the clustered historical data into a Markov chain original model and training the Markov chain original model to obtain a Markov chain prediction model;
and the prediction module is used for inputting the current daily load data into the Markov chain prediction model so as to predict the data of the next stage.
And calculating the time when the user reaches the target detection point according to the user-specified time, the specified position and the target detection point position, wherein the user-specified time can be the current time or a certain time point in the future, and the specified position can be the current position or a position selected by the user.
Inputting the current queuing data into a Markov prediction model, predicting the number of queuing people when a user reaches a target detection point, and predicting the time required by queuing;
calculating the estimated total time spent according to the time required for queuing and the time spent by the user to come and go to the target detection point;
and comparing the predicted total time consumed by the user for selecting different target detection points, and outputting the target detection point at the left end of the predicted total time consumed as the optimal detection point.
According to the queuing point recommendation method based on unmanned aerial vehicle aerial photography, provided by the embodiment of the description, the YOLOv5 algorithm is used as a real-time target detection algorithm, compared with other algorithms, the single-stage algorithm is faster in operation speed and higher in average detection precision, the target detection point queuing condition at a future time point is predicted through a Markov prediction model, the number and speed of queuing people after a period of time from a user departure point to a target detection point in a certain trip mode can be predicted, the accuracy and practicability of the embodiment can be improved, in the case of counting the number of queuing people with a plurality of target detection points nearby, the most preferable item can be selected according to the existing data and historical data for the user to select, and the user can conveniently make a trip plan.
As an embodiment, the method further comprises:
acquiring a preference factor of a user;
after calculating the estimated total elapsed time, the method is replaced by:
calculating the recommendation degrees of different target detection points according to the preference factor of the user and the calculated estimated total time consumption;
and comparing the recommendation degrees of the target detection points, and outputting the target detection point with the highest recommendation degree as an optimal detection point.
If the obtaining of the preference factor of the user comprises:
obtaining travel willingness coefficients of the user under the conditions of different terrains, different weather and different time periods in a questionnaire survey or historical data analysis mode, and calculating the recommendation degrees of different target detection points according to the following formula:
wherein m is a travel desire coefficient under the current weather, and n is the time slot corresponding to the designated timeWillingness coefficient, k 1 、k 2 ……k n Will coefficient of travel respectively for different terrains, d 1 、d 2 ……d n Respectively the path lengths of different terrains, d is the total path length between the designated position and the target detection point, t is the predicted total time consumption, r is the correction coefficient, r 1 、r 2 ……r n Respectively d/t corresponding to different sets A 1 、A 2 ……A n The correction factor of time.
The method considers that the time spent by a user is generally not only taken as the only consideration factor when the user selects the target queuing point, when the time difference between two or more queuing points is not large, the factors such as weather conditions, road conditions and the like can also influence the selection of the user, if the weather is fine and the temperature is appropriate, the user may prefer to queue points which are slightly far away but have less total time spent, if the weather is severe and the road is muddy in the course, the user can more frequently spend slightly more time but queue points which are close, and even time factors such as before work, after work, on the weekend, in the evening can also influence the judgment of the user.
As an embodiment, the method further includes, performing path planning on the unmanned aerial vehicle at a plurality of target detection points based on a genetic algorithm, including:
a1, first, encoding is performed.
For the TSP problem, binary coding and real coding are usually used, real coding is more convenient (which is reflected in mutation and exchange later), each individual is a path sequence of a group of detection points, each detection point is a chromosome, the path sequence is randomly initialized by applying np.
A2, then, calculating the fitness
The higher the fitness to the environment, the easier it is to choose naturally, while in TSP the total path is the shortest, for convenience this project defines fitness as the reciprocal of distance, fit =1
A3, and then, selecting.
The project adopts roulette selection, firstly, the probability of each individual being selected is determined, the formula is that the fitness of each individual is divided by the fitness of the total individual, p = fi/Σ fi converts each probability into an accumulated probability when the probability is realized, then, numbers are randomly generated in [0,1], relative intervals of the accumulated probabilities are observed, and corresponding individuals are selected.
A4, and then, crossover.
Under real number coding, partial mapping intersection is adopted to define a random number of [0,1], and when the random number is greater than the intersection probability cross-prob =0.9, part of chromosomes (a plurality of adjacent detection points) are intersected, but in TSP, detection point paths are intersected to generate collision, and repeated individual staining may occur. The upper and lower individual conflict parts carry out exchange to resolve the conflict.
And A5, and finally, mutation.
Flip variation is taken to define a random number of [0,1] that is varied when it is less than the variation probability mut-prob = 0.1.
And A6, repeating A2-A5 until a termination condition is met.
And (4) termination conditions: there are generally three types: and when the maximum iteration frequency is reached, the accuracy is reached, the fitness value of the optimal individual in the population is continuous, and the third method is adopted in the project, namely, the iteration frequency is adjusted until the finally obtained fitness value is unchanged.
Through the flight path planning adopting the genetic algorithm, the efficiency of the unmanned aerial vehicle in image acquisition of a plurality of target detection points can be improved, the energy is saved, and the endurance of the unmanned aerial vehicle is improved.
The embodiment of this specification still provides a queuing point recommendation system based on unmanned aerial vehicle takes photo by plane, includes:
the image processing module is used for intercepting the picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
the counting module is used for counting the crowd of the queuing images of the detection points through a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
the prediction module is used for clustering queuing data of each target detection point by using dynamic time warping as distance measurement, inputting the clustered historical queuing data into a Markov original model, training the original model to obtain a prediction model, calculating the time when a user reaches the target detection point according to the time designated by the user, the designated position and the position of the target detection point, inputting the current queuing data into the Markov prediction model, predicting the number of queuing people when the user reaches the target detection point, predicting the time required by queuing, and calculating the predicted total consumed time according to the time required by queuing and the time spent by the user to go to and return from the target detection point;
and the recommending module is used for comparing the predicted total time consumption of different target detection points selected by the user and outputting the target detection point at the left end of the predicted total time consumption as the optimal detection point.
Further, this system can also be responsible for supervision unmanned aerial vehicle and synchronous real-time data through setting up host computer and server, as shown in fig. 3.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, is limited to those examples; within the idea of the invention, also technical features in the above embodiments or in different embodiments may be combined, steps may be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.
The present invention is intended to embrace all such alternatives, modifications and variances which fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalents, improvements, and the like that may be made without departing from the spirit or scope of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A queuing point recommendation method based on unmanned aerial vehicle aerial photography is characterized by comprising the following steps:
intercepting a picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
counting the crowd of the queuing images of the detection points by using a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
clustering the queuing data of each target detection point by using dynamic time warping as distance measurement;
inputting the clustered historical queuing data into a Markov original model, and training the original model to obtain a prediction model;
calculating the time when the user reaches the target detection point according to the time designated by the user, the designated position and the position of the target detection point;
inputting the current queuing data into a Markov prediction model, predicting the number of queuing people when the user reaches a target detection point, and predicting the time required by queuing;
calculating the estimated total time consumption according to the time required by queuing and the time spent by the user to come and go to the target detection point;
and comparing the predicted total time consumption of different target detection points selected by the user, and outputting the target detection point corresponding to the minimum predicted total time consumption as the optimal detection point.
2. The UAV-based queue point recommendation method of claim 1, further comprising:
acquiring a preference factor of a user;
after calculating the estimated total elapsed time, the method is replaced by:
calculating recommendation degrees of different target detection points according to the preference factors of the user and the calculated predicted total time consumption;
and comparing the recommendation degrees of the target detection points, and outputting the target detection point with the highest recommendation degree as an optimal detection point.
3. The method of claim 2, wherein the obtaining of the preference factor of the user comprises:
and obtaining the travel willingness coefficients of the user under the conditions of different terrains, different weather and different time periods in a questionnaire survey or historical data analysis mode.
4. The method for recommending queuing points based on unmanned aerial vehicle aerial photography according to claim 3, wherein the formula for calculating the recommendation degrees of different target detection points is as follows:
wherein m is a travel intention coefficient under the current weather, n is a travel intention coefficient of a corresponding time period at a specified time, and k 1 、k 2 ……k n Will coefficient of travel respectively for different terrains, d 1 、d 2 ……d n Respectively the lengths of the paths of different terrains, d is the total length of the path between the designated position and the target detection point, t is the predicted total time consumption, r is a correction coefficient, r is the total time consumption 1 、r 2 ……r n Respectively d/t corresponding to different sets A 1 、A 2 ……A n The correction factor of time.
5. The method of claim 1, wherein clustering the queuing data for each target detection point using dynamic time warping as a distance metric comprises:
calculating the intra-cluster and inter-cluster dynamic time warping distance ratio, comprising:
calculating the dynamic time regular distance of any two historical daily load data in the cluster, and averaging the dynamic time regular distances to obtain the dynamic time regular average distance in the cluster;
calculating the dynamic time warping distance between any two central points between clusters, and averaging the dynamic time warping distances to obtain the inter-cluster dynamic time warping average distance;
and obtaining the intra-cluster inter-cluster dynamic time warping distance ratio according to the intra-cluster dynamic time warping average distance and the inter-cluster dynamic time warping average distance.
6. The unmanned aerial vehicle aerial-based queuing point recommendation method of claim 1, further comprising:
the method for planning the path of the unmanned aerial vehicle at a plurality of target detection points based on the genetic algorithm comprises the following steps:
a1, binary coding and real number coding are adopted, each individual is a path sequence of a group of detection points, each detection point is a chromosome, the path sequence is randomly initialized in matlab by using np.array function, shuffle is carried out by using shuffle function, each individual is determined, and population initialization is realized;
a2, defining the fitness as the reciprocal of the distance, and calculating the fitness;
a3, selecting by adopting a roulette wheel, firstly determining the probability of each individual to be selected, dividing the fitness of each individual by the fitness of the total individual by a formula, namely converting each probability into an accumulated probability when p = fi/sigma fi is realized, then randomly generating a number in [0,1], and selecting a corresponding individual according to a relative interval falling in the accumulated probability;
a4, under real number coding, adopting partial mapping intersection to define a random number of [0,1], and when the random number is greater than the intersection probability cross-prob =0.9, intersecting part of chromosomes (a plurality of adjacent detection points), and exchanging upper and lower individual collision parts to solve collision;
a5, adopting flip variation to define a random number of [0,1], and performing variation when the random number is less than the variation probability mut-prob = 0.1;
repeating the steps A2-A5 until the termination condition is met: the fitness values of the best individuals in the population are continuous.
7. The unmanned aerial vehicle aerial photography-based queuing point recommendation method according to claim 1, wherein in the population counting of the queuing images of detection points, the neural network parameter obtaining method of the YOLOv5 algorithm comprises the following steps:
preprocessing the input end of the neural network on the queuing image;
slicing the queuing image through a Focus structure of a backhaul, and performing convolution operation on the sliced image to obtain a double-sampling feature map under the condition of no information loss;
performing convolution processing on the doubled down-sampled feature map through a CBL convolution module, and then performing feature learning on the model by utilizing the CSP structure;
processing the characteristic tensor through the FPN + PAN structure of the Neck;
the Head layer predicts the image characteristics, generates a bounding box and a prediction category, and outputs the convolution neural network parameters of YOLOv 5.
8. The UAV-based queue point recommendation method of claim 7, wherein the preprocessing comprises Mosaic data enhancement, adaptive anchor frame calculation, and adaptive picture scaling.
9. The unmanned aerial vehicle aerial-based queuing point recommendation method of claim 1, further comprising: and feeding back the number of real-time queuing people, the predicted queuing time of each target detection point and the predicted total time consumption of each target detection point in real time.
10. The utility model provides a queuing point recommendation system based on unmanned aerial vehicle takes photo by plane which characterized in that includes:
the image processing module is used for intercepting the picture captured by the unmanned aerial vehicle to obtain a queuing image of each target detection point;
the counting module is used for counting the crowd of the queuing images of the detection points through a YOLOv5 algorithm to obtain the real-time queuing number of people at each target detection point;
the prediction module is used for clustering queuing data of each target detection point by using dynamic time warping as distance measurement, inputting the clustered historical queuing data into a Markov original model, training the original model to obtain a prediction model, calculating the time when a user reaches the target detection point according to the time designated by the user, the designated position and the position of the target detection point, inputting the current queuing data into the Markov prediction model, predicting the number of queuing people when the user reaches the target detection point, predicting the time required by queuing, and calculating the predicted total consumed time according to the time required by queuing and the time spent by the user to go to and return from the target detection point;
and the recommending module is used for comparing the predicted total cost time of different target detection points selected by the user and outputting the target detection point corresponding to the minimum predicted total cost time as the optimal detection point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211124537.XA CN115470418B (en) | 2022-09-15 | 2022-09-15 | Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211124537.XA CN115470418B (en) | 2022-09-15 | 2022-09-15 | Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115470418A true CN115470418A (en) | 2022-12-13 |
| CN115470418B CN115470418B (en) | 2023-07-25 |
Family
ID=84333885
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211124537.XA Active CN115470418B (en) | 2022-09-15 | 2022-09-15 | Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115470418B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116483210A (en) * | 2023-06-25 | 2023-07-25 | 安徽大学 | Brain-controlled unmanned aerial vehicle method and system based on deep learning and sliding mode control |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108288498A (en) * | 2018-01-15 | 2018-07-17 | 浙江大学 | A kind of hospital intelligent attendance system based on machine vision |
| CN108629323A (en) * | 2018-05-11 | 2018-10-09 | 哈尔滨工业大学 | A kind of integrated providing method of scenic spot tourist chain type trip |
| CN111160631A (en) * | 2019-12-17 | 2020-05-15 | 西北工业大学 | Conflict detection and resolution method based on four-dimensional track operation |
| CN111160617A (en) * | 2019-12-06 | 2020-05-15 | 北京国电通网络技术有限公司 | Power daily load prediction method and device |
| CN112561142A (en) * | 2020-12-04 | 2021-03-26 | 电子科技大学成都学院 | Queuing information inquiry system |
| CN114399552A (en) * | 2022-03-24 | 2022-04-26 | 武汉视合远方科技有限公司 | Indoor monitoring environment behavior identification and positioning method |
| CN114741618A (en) * | 2022-04-19 | 2022-07-12 | 中国工商银行股份有限公司 | Offline service point recommendation method, recommendation device, equipment and medium |
-
2022
- 2022-09-15 CN CN202211124537.XA patent/CN115470418B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108288498A (en) * | 2018-01-15 | 2018-07-17 | 浙江大学 | A kind of hospital intelligent attendance system based on machine vision |
| CN108629323A (en) * | 2018-05-11 | 2018-10-09 | 哈尔滨工业大学 | A kind of integrated providing method of scenic spot tourist chain type trip |
| CN111160617A (en) * | 2019-12-06 | 2020-05-15 | 北京国电通网络技术有限公司 | Power daily load prediction method and device |
| CN111160631A (en) * | 2019-12-17 | 2020-05-15 | 西北工业大学 | Conflict detection and resolution method based on four-dimensional track operation |
| CN112561142A (en) * | 2020-12-04 | 2021-03-26 | 电子科技大学成都学院 | Queuing information inquiry system |
| CN114399552A (en) * | 2022-03-24 | 2022-04-26 | 武汉视合远方科技有限公司 | Indoor monitoring environment behavior identification and positioning method |
| CN114741618A (en) * | 2022-04-19 | 2022-07-12 | 中国工商银行股份有限公司 | Offline service point recommendation method, recommendation device, equipment and medium |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116483210A (en) * | 2023-06-25 | 2023-07-25 | 安徽大学 | Brain-controlled unmanned aerial vehicle method and system based on deep learning and sliding mode control |
| CN116483210B (en) * | 2023-06-25 | 2023-09-08 | 安徽大学 | Brain-controlled unmanned aerial vehicle method and system based on deep learning and sliding mode control |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115470418B (en) | 2023-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111712830B (en) | Real-time object detection using depth sensors | |
| CN109241871A (en) | A kind of public domain stream of people's tracking based on video data | |
| CN112488210A (en) | Three-dimensional point cloud automatic classification method based on graph convolution neural network | |
| CN116258817B (en) | Automatic driving digital twin scene construction method and system based on multi-view three-dimensional reconstruction | |
| CN111640116B (en) | Aerial photography graph building segmentation method and device based on deep convolutional residual error network | |
| TWI643137B (en) | Object recognition method and object recognition system | |
| CN109272487A (en) | The quantity statistics method of crowd in a kind of public domain based on video | |
| CN107862702A (en) | A kind of conspicuousness detection method of combination boundary connected and local contrast | |
| CN111126459A (en) | Method and device for identifying fine granularity of vehicle | |
| CN115359366A (en) | Remote sensing image target detection method based on parameter optimization | |
| CN111414931A (en) | Multi-branch multi-scale small target detection method based on image depth | |
| CN111738327B (en) | Ultrashort-term irradiation prediction method based on typical cloud shielding irradiation difference | |
| CN116091946A (en) | Yolov 5-based unmanned aerial vehicle aerial image target detection method | |
| CN111860823A (en) | Neural network training method, neural network training device, neural network image processing method, neural network image processing device, neural network image processing equipment and storage medium | |
| CN115810149A (en) | High-resolution remote sensing image building extraction method based on superpixel and image convolution | |
| CN115470418A (en) | Queuing point recommendation method and system based on unmanned aerial vehicle aerial photography | |
| CN114373104A (en) | Three-dimensional point cloud semantic segmentation method and system based on dynamic aggregation | |
| CN116580322A (en) | Unmanned aerial vehicle infrared small target detection method under ground background | |
| Wan et al. | A geometry-aware attention network for semantic segmentation of MLS point clouds | |
| Alshehhi et al. | Detection of Martian dust storms using mask regional convolutional neural networks | |
| CN117557983A (en) | Scene reconstruction method and driving assistance system based on depth forward projection and query back projection | |
| CN111476226B (en) | Text positioning method and device and model training method | |
| KR102014288B1 (en) | Development pressure prediction method based on artificial intelligence using drone | |
| CN116665091A (en) | Real-time multi-camera video analysis method under intelligent traffic scene | |
| CN116912486A (en) | Target segmentation method based on edge convolution and multidimensional feature fusion and electronic device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |