CN114282623A - Analysis method based on illegal boarding and alighting of passenger vehicles - Google Patents
Analysis method based on illegal boarding and alighting of passenger vehicles Download PDFInfo
- Publication number
- CN114282623A CN114282623A CN202111636282.0A CN202111636282A CN114282623A CN 114282623 A CN114282623 A CN 114282623A CN 202111636282 A CN202111636282 A CN 202111636282A CN 114282623 A CN114282623 A CN 114282623A
- Authority
- CN
- China
- Prior art keywords
- violation
- vehicle
- information
- data
- passenger
- 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.)
- Pending
Links
Images
Abstract
The invention provides a passenger getting-on and getting-off analysis method based on passenger vehicle violation, which comprises the following steps: step S01: vehicle violation screening is carried out on passenger transport monitoring data, and violation boarding and alighting characteristic values are obtained; step S02: according to the characteristic values of the illegal passengers getting on and getting off, characteristic analysis is carried out, and a vehicle illegal model is constructed; step S03: generating output value similarity by performing model comparison calculation on the vehicle violation model; step S04: according to the similarity of the output values, deviation identification is carried out to obtain information of the vehicle to be detected; step S05: carrying out violation detection on the vehicle to be detected according to the information of the vehicle to be detected, and acquiring violation information of the vehicle; by establishing the vehicle violation model, the analysis efficiency of the violation behaviors of the passenger vehicle is improved, the similarity of the output values is calculated, the accuracy of the violation analysis of the passenger vehicle is enhanced, the invalid detection is reduced, and the violation detection time is saved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an analysis method for getting on and off passengers in violation of passenger vehicles.
Background
At present, the problems of uneven driver quality and illegal boarding and alighting occur frequently, so that the development order of the industry is influenced, in addition, passengers getting off the bus can randomly walk on a highway, traffic accidents are very easy to occur, the passengers getting on and off the bus stop on a traffic lane, and law enforcement departments can not control the bus with unclear license plate numbers and vehicle characteristics; the problem that more traveling vehicles are not standard due to the influence of the environment exists, and unnecessary loss is caused because passengers are not loaded or unloaded according to the regulations in the traveling process; in the existing method for monitoring vehicles, except for the existing roadside common camera, the method for judging the traffic violation of the long-distance passenger vehicle based on the traffic information of the toll gate, which is disclosed as application number '201410692387.1', the method for comparing the spatial overlapping of the long-distance passenger vehicles from 2 to 5 in the morning by using the road gate is mentioned, and whether the violation exists is judged according to the vehicle attribute; the technical scheme is limited to be only suitable for being used in a specific time period, and only the vehicle is subjected to reference analysis, so that the passengers getting on and off the vehicle cannot be contrasted; the invention provides a method for constructing multi-algorithm capability based on an AI artificial intelligence technology, which realizes the behavior judgment of passengers getting on and off by analyzing the data of the passenger car and dynamically analyzing the passenger flow near the passenger car by performing frame extraction and algorithm analysis on the real-time video, and can quickly and accurately discover illegal behaviors by continuously collecting the data and continuously analyzing the behaviors of people.
Disclosure of Invention
The invention provides an analysis method for illegal boarding and alighting based on passenger vehicles, which is used for solving the problem.
The invention provides a passenger getting-on and getting-off analysis method based on passenger vehicle violation, which comprises the following steps:
step S01: vehicle violation screening is carried out on passenger transport monitoring data, and violation boarding and alighting characteristic values are obtained;
step S02: according to the characteristic values of the illegal passengers getting on and getting off, characteristic analysis is carried out, and a vehicle illegal model is constructed;
step S03: generating output value similarity by performing model comparison calculation on the vehicle violation model;
step S04: according to the similarity of the output values, deviation identification is carried out to obtain information of the vehicle to be detected;
step S05: and carrying out violation detection on the vehicle to be detected according to the information of the vehicle to be detected to obtain violation information of the vehicle
As an embodiment of the present invention, the step S01 includes:
the method comprises the steps of classifying passenger transport monitoring data according to data functions to obtain monitoring data categories; wherein the content of the first and second substances,
the passenger traffic monitoring data comprises: bayonet data and video structured data;
the monitoring data categories include: number category, speed category, time category, behavior category;
and performing category feature extraction on each detection data category to generate a category feature value of each category, and performing passenger getting-on and getting-off analysis on the category feature value to obtain a violation passenger getting-on and getting-off feature value.
As an embodiment of the present invention, the step S02 includes:
modeling and screening the characteristic values of the illegal upper and lower passengers to obtain the characteristic information of the illegal model; wherein the content of the first and second substances,
the violation model feature information includes: position information, node information, edge information and interval information;
and constructing a main body model according to the violation model characteristic information to generate a vehicle violation model.
As an embodiment of the present technical solution, the model comparison calculation includes the following steps:
the method comprises the following steps: normalization data is obtained by performing normalization processing on the vehicle default scale type data; wherein the content of the first and second substances,
the normalization processing maps the conversion result to a preset interval by carrying out dispersion standardization on the vehicle violation model data; wherein the content of the first and second substances,
the dispersion normalization includes: performing linear transformation on the vehicle violation model data through a preset conversion function;
step two: transmitting the normalized data to a twin network for mapping analysis processing to generate twin characteristic space data; wherein the content of the first and second substances,
the twin network includes: extracting dual-input data of the normalized data, inputting the dual-input data into a neural network, and performing mapping processing; wherein the content of the first and second substances,
the dual input data includes: front frame data and rear frame data;
step three: performing output calculation on the twin characteristic space data to obtain output value similarity, and judging; wherein the content of the first and second substances,
when the similarity of the output values is smaller than or equal to a preset threshold value, the vehicle behavior is normal;
and when the output value similarity is larger than a preset threshold value, judging that the vehicle behavior is abnormal.
As an embodiment of the present invention, the step S04 includes:
matching a corresponding calculation method according to the similarity of the output values, performing deviation calculation, acquiring a deviation value, and judging; wherein the content of the first and second substances,
the deviation calculation includes: according to the corresponding calculation method, comparing and calculating the output value with a preset output comparison data set to obtain a comparison deviation value;
when the deviation value is larger than a preset threshold value, determining the deviation value is an overlarge deviation, determining the vehicle to be detected, and acquiring information of the vehicle to be detected;
and when the deviation value is less than or equal to a preset threshold value, determining that the deviation value is a normal deviation.
As an embodiment of the present invention, the step S05 includes:
the method comprises the steps that key data screening is conducted on information of a vehicle to be detected, and first screening information is obtained; wherein the content of the first and second substances,
the key data screening comprises the following steps: screening the moving state and screening the path;
extracting vehicle information from the first screening information to obtain a first vehicle to be detected, and carrying out violation detection on the first vehicle to be detected to obtain vehicle violation information; wherein the content of the first and second substances,
the violation detection includes: passenger car detection and passenger car license plate identification.
As an embodiment of the technical solution, the passenger car detection comprises the following steps:
step S10: the method comprises the steps that characteristic layering is conducted on vehicle information to be detected, and an information characteristic layer is obtained; wherein the content of the first and second substances,
the feature layering includes: scale layering, cost layering, training layering;
step S20: respectively extracting feature information of each information feature layer according to preset multi-scale feature learning, and generating information feature nodes corresponding to the information feature layers;
step S30: performing subspace mapping on the information characteristic layer and the corresponding information characteristic nodes to obtain a mapping result, and judging the mapping result; wherein the content of the first and second substances,
when the mapping result is within a preset threshold range, the passenger car is in compliance;
and when the mapping result is not within the preset threshold range, acquiring the illegal passenger car information for the illegal passenger car, and performing preset violation coping processing.
As an embodiment of the technical scheme, the passenger vehicle license plate recognition comprises the following steps:
step S100: extracting convolution characteristics through a preset detection convolution neural network to obtain a characteristic diagram; wherein the content of the first and second substances,
the convolution feature extraction includes: convolution, activating function and constructing a pooling layer;
step S200: carrying out preset feature network analysis processing according to the feature map to generate a passenger vehicle detection frame; wherein the content of the first and second substances,
the feature network analysis processing includes: carrying out regional proposal network processing and full connection layer processing; wherein the content of the first and second substances,
the area proposal network processing comprises: obtaining nodes of a region to be selected by generating the region to be selected, judging the node attributes of the region to be selected, and performing bounding box regression processing according to the node attributes to determine a first region to be selected; wherein the content of the first and second substances,
the node attributes include: foreground nodes and background nodes;
the full connection layer processing comprises: according to the first region to be selected, extracting comprehensive information to generate a feature map of the region to be selected, and performing frame regression calculation according to the feature map of the region to be selected to obtain a passenger car detection frame;
step S300: and carrying out passenger vehicle violation detection according to the passenger vehicle detection frame, and acquiring passenger vehicle violation information.
As an embodiment of the present invention, the violation detection includes:
the method comprises the steps of acquiring segmented detection information by carrying out segmented detection on the passenger vehicle; wherein the content of the first and second substances,
the segment detection includes: time subsection detection and route subsection detection;
according to the segmentation detection information, carrying out segmentation similarity analysis to obtain a segmentation similarity value, and judging; wherein the content of the first and second substances,
the segmentation similarity analysis is to carry out similarity calculation analysis on each piece of information and the previous piece of information;
when the segment similarity value is within a preset threshold range, the segment is a normal segment;
and when the segment similarity value is not in the preset threshold range, performing abnormal segmentation and performing abnormal analysis processing.
As an embodiment of the present invention, the violation detection further includes:
detecting the portrait violation of the vehicle violation information to obtain portrait violation data; wherein the content of the first and second substances,
the portrait violation detection includes: detecting getting-on behaviors and getting-off behaviors;
performing data comparison processing according to the portrait violation data and a preset violation database to obtain violation levels; wherein the content of the first and second substances,
the violation levels include: minor, common, severe violations;
performing corresponding safety early warning according to the violation level to acquire early warning information; wherein the content of the first and second substances,
the safety precaution includes: the vehicle where the illegal portrait is located is warned in a networking mode according to the violation level, and warning information is obtained;
the illegal vehicle carries out illegal correction processing according to the warning information to generate illegal correction information;
the early warning information includes: warning information, violation correction information, and positioning information.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of an analysis method for getting on and off passengers based on violation of passenger vehicles in an embodiment of the invention;
FIG. 2 is a flowchart of model comparison calculation in an analysis method based on illegal boarding and alighting of passenger vehicles according to an embodiment of the present invention;
fig. 3 is a flow chart of passenger vehicle detection in an analysis method based on illegal boarding and alighting of passenger vehicles in the embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
It will be understood that when an element is referred to as being "secured to" or "disposed on" another element, it can be directly on the other element or be indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly or indirectly connected to the other element.
It will be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like, as used herein, refer to an orientation or positional relationship indicated in the drawings that is solely for the purpose of facilitating the description and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and is therefore not to be construed as limiting the invention.
Moreover, it is noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions, and "a plurality" means two or more unless specifically limited otherwise. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
The embodiment of the invention provides an analysis method for illegal boarding and alighting based on passenger vehicles, which comprises the following steps:
step S01: vehicle violation screening is carried out on passenger transport monitoring data, and violation boarding and alighting characteristic values are obtained;
step S02: according to the characteristic values of the illegal passengers getting on and getting off, characteristic analysis is carried out, and a vehicle illegal model is constructed;
step S03: generating output value similarity by performing model comparison calculation on the vehicle violation model;
step S04: according to the similarity of the output values, deviation identification is carried out to obtain information of the vehicle to be detected;
step S05: carrying out violation detection on the vehicle to be detected according to the information of the vehicle to be detected, and acquiring violation information of the vehicle;
the working principle of the technical scheme is as follows: in the prior art, the control method and system for intelligent monitoring of the passenger flow of the road passenger transport system analyzes getting-on and getting-off behaviors of the passenger transport vehicle, for example, 201911312032.4 detects the getting-on/getting-off behaviors of passengers on the vehicle, judges whether the passenger transport vehicle is overloaded or not, locks the vehicle by monitoring the station and in the way, registers the illegal vehicle into a warehouse for punishment, and analyzes the getting-on and getting-off behaviors of the passengers to judge whether the passenger transport vehicle has the illegal behaviors, particularly analyzes the passenger flow, mainly improves the safe driving of the vehicle owner and the driving safety of the passengers, and the technical scheme starts from the driving angle and the portrait behavior angle, analyzes and judges the illegal getting-on and getting-off behaviors of the passenger transport vehicle, and firstly screens the monitored data in violation, obtaining the characteristic values of illegal boarding and alighting passengers, then carrying out characteristic analysis to construct an illegal model, carrying out model comparison calculation on the model, calculating the similarity of output values, carrying out deviation identification according to the similarity, obtaining the information of the vehicle to be detected, carrying out illegal detection on the vehicle to be detected, and finally obtaining the illegal information of the vehicle;
the beneficial effects of the above technical scheme are: in the actual use process, when the passenger vehicle normally runs, violation detection and screening are carried out on the data by obtaining the data of the passenger vehicle, for example, the conditions of stopping at a section where the passenger vehicle is forbidden, running a traffic light and the like occur, the data are uploaded, the analysis efficiency of the violation behaviors of the passenger vehicle is greatly improved by screening, the accuracy of the violation analysis of the passenger vehicle is enhanced by establishing a vehicle violation model and calculating the similarity of output values, the vehicle to be detected is determined firstly, then the violation detection is carried out on the vehicle to be detected, the invalid detection is reduced, the violation detection efficiency is improved, and the violation detection time is saved.
In one embodiment, the step S01 includes:
the method comprises the steps of classifying passenger transport monitoring data according to data functions to obtain monitoring data categories; wherein the content of the first and second substances,
the passenger traffic monitoring data comprises: bayonet data and video structured data;
the monitoring data categories include: number category, speed category, time category, behavior category;
performing category feature extraction on each detection data category to generate a category feature value of each category, and performing passenger getting-on and getting-off analysis on the category feature value to obtain a violation passenger getting-on and getting-off feature value;
the working principle of the technical scheme is as follows: in the prior art, passenger transport vehicle data are generally directly detected to judge whether violation occurs, but the violation detection is not carried out when the violation occurs after many times of detection, so that the violation detection efficiency is reduced, and in the above technical scheme, passenger transport monitoring data are classified according to the data action, category feature extraction is carried out on each detected data category to generate a category feature value of each category, and passenger getting-on and getting-off analysis is carried out to obtain a violation passenger getting-on and getting-off feature value;
the beneficial effects of the above technical scheme are: the passenger vehicle monitoring data are classified, so that the feature extraction efficiency is greatly improved, and the class characteristic value is generated by extracting the features, so that the data judgment is more visual, and the data analysis speed is improved.
In one embodiment, the step S02 includes:
modeling and screening the characteristic values of the illegal upper and lower passengers to obtain the characteristic information of the illegal model; wherein the content of the first and second substances,
the violation model feature information includes: position information, node information, edge information and interval information;
constructing a main body model according to the violation model characteristic information to generate a vehicle violation model;
the working principle of the technical scheme is as follows: different from the single screening of screening the violation data according to the data in the prior art, the technical scheme carries out modeling screening processing through the violation passenger characteristic values to obtain the violation model characteristic information, and then carries out model construction to generate the vehicle violation model;
the beneficial effects of the above technical scheme are: modeling and screening are carried out according to the characteristic values of the passengers on the.
In one embodiment, the model comparison calculation comprises the steps of:
the method comprises the following steps: normalization data is obtained by performing normalization processing on the vehicle default scale type data; wherein the content of the first and second substances,
the normalization processing maps the conversion result to a preset interval by carrying out dispersion standardization on the vehicle violation model data; wherein the content of the first and second substances,
the dispersion normalization includes: performing linear transformation on the vehicle violation model data through a preset conversion function; wherein the content of the first and second substances,
obtaining a model data set { x1,x2,…,xtH, the transfer function yiComprises the following steps:
wherein, yiConversion function, x, for the ith data in the model data setiIs the ith data, x, in the model data setminFor a model data set { x1,x2,…,xtMinimum data in (x)maxFor a model data set { x1,x2,…,xtThe maximum data in the data, i is variable, i is more than or equal to 1 and less than or equal to t, t is constant and 1<t, mapping result is [0,1 ]]Within the range;
step two: transmitting the normalized data to a twin network for mapping analysis processing to generate twin characteristic space data; wherein the content of the first and second substances,
the twin network includes: extracting dual-input data of the normalized data, inputting the dual-input data into a neural network, and performing mapping processing; wherein the content of the first and second substances,
the dual input data includes: front frame data and rear frame data;
step three: obtaining a mapping data set y1,y2,…,ytCarrying out output calculation on the twin characteristic space data, and calculating the similarity D (X, Y) of output values;
wherein, yiConversion function, x, for the ith data in the model data setiFor the ith data in the model data group, p is a constant, and p is taken as 2;
when D (X, Y) is less than or equal to 0.95, the threshold value is normal vehicle behavior;
when D (X, Y) >0.95, the abnormal vehicle behavior is judged;
the working principle of the technical scheme is as follows: firstly, carrying out dispersion standardization on vehicle default scale data, mapping a conversion result to a preset interval, and acquiring normalized data, wherein the dispersion standardization comprises the following steps: linearly transforming vehicle violation model data through a preset conversion function, transmitting the normalized data to a twin network for mapping analysis processing to generate twin characteristic space data, and finally, performing output calculation on the twin characteristic space data to obtain output value similarity, wherein the output value similarity is normal vehicle behavior when the output value similarity is smaller than or equal to a preset threshold value, and the output value similarity is abnormal vehicle behavior when the output value similarity is larger than the preset threshold value;
the beneficial effects of the above technical scheme are: by carrying out normalization processing on the model data and mapping the result to a preset interval, the data accuracy is improved, by generating twin feature space data, the feature pertinence is improved, and by calculating the similarity of output values, whether the vehicle behavior violates rules or not is judged more intuitively, and the judgment accuracy is improved.
In one embodiment, the step S04 includes:
matching a corresponding calculation method according to the similarity of the output values, performing deviation calculation, acquiring a deviation value, and judging; wherein the content of the first and second substances,
the deviation calculation includes: according to the corresponding calculation method, comparing and calculating the output value with a preset output comparison data set to obtain a comparison deviation value;
the deviation calculation further comprises the steps of:
step S101: obtaining an output data set { alpha }1,α2,…,αn}, output data influencing parameter { mu1,μ2,…,μnRespectively calculating expected values delta of the output datad：
Wherein alpha istFor the t-th output data in the output data set, mutThe influence parameter of the t-th output data in the output data group is t which is a variable, t is more than or equal to 1 and less than or equal to d, d is a variable, d is more than or equal to 1 and less than or equal to n, n is the number of the output data in the output data group, and n is more than or equal to 1;
step S102: obtaining a control data set [ beta ]1,β2,…,βn}, reference data optimal parameter { epsilon1,ε2,…,εnThe maximum impact factor
Calculating expected values of control data τ respectivelyd：
Wherein, betalIs the first control data in the control data set,. epsilonlFor the optimal parameters of the ith control data in the control data set,
is the maximum influence factor of the first comparison data in the comparison data group, wherein l is a variable, and l is more than or equal to 1 and less than or equal to d;
step S103: according to the expected value delta of the output datadAnd expected value of control data τdEstablishing an equation set, and respectively calculating a data deviation value thetad：
Wherein, mun+1For the maximum influencing parameter of the data deviation, αuFor the u-th output data in the output data set, θuIs the data deviation value of the u-th output data in the output data group, u is a variable, u is more than or equal to 1 and less than or equal to n-1, munAn output data influence parameter for the nth output data in the output data set;
when the deviation value is larger than a preset threshold value, determining the deviation value is an overlarge deviation, determining the vehicle to be detected, and acquiring information of the vehicle to be detected;
when the deviation value is smaller than or equal to a preset threshold value, the deviation value is a normal deviation;
the working principle of the technical scheme is as follows: compared with the prior art, the obtained vehicle data is directly judged according to the set data, the output value and a preset output contrast data group are compared and calculated through a calculation method corresponding to the similarity matching of the output value in the technical scheme, a contrast deviation value is obtained, when the deviation value is larger than a preset threshold value, the deviation value is an overlarge deviation, the vehicle is determined to be a vehicle to be detected, the information of the vehicle to be detected is obtained, and when the deviation value is smaller than or equal to the preset threshold value, the deviation value is a normal deviation;
the beneficial effects of the above technical scheme are: by calculating the deviation value, the speed of determining the vehicle to be detected is improved, and meanwhile, the error rate of detecting the vehicle in a non-violation mode is reduced.
In one embodiment, the step S05 includes:
the method comprises the steps that key data screening is conducted on information of a vehicle to be detected, and first screening information is obtained; wherein the content of the first and second substances,
the key data screening comprises the following steps: screening the moving state and screening the path;
extracting vehicle information from the first screening information to obtain a first vehicle to be detected, and carrying out violation detection on the first vehicle to be detected to obtain vehicle violation information; wherein the content of the first and second substances,
the violation detection includes: passenger vehicle detection and license plate identification of the passenger vehicle;
the working principle of the technical scheme is as follows: firstly, a detection object is determined to be vehicle information to be detected, after moving state screening and path screening are carried out, first screening information is determined, then vehicle information is extracted to obtain a first vehicle to be detected, passenger vehicle detection and passenger vehicle license plate recognition are carried out, and vehicle violation information is obtained;
the beneficial effects of the above technical scheme are: by screening key data of the vehicle to be detected, the vehicle violation detection efficiency is improved, and vehicle determination is performed from the screening information, so that the richness of vehicle information and the accuracy of violation vehicles are improved.
In one embodiment, the passenger vehicle detection comprises the steps of:
step S10: the method comprises the steps that characteristic layering is conducted on vehicle information to be detected, and an information characteristic layer is obtained; wherein the content of the first and second substances,
the feature layering includes: scale layering, cost layering, training layering;
step S20: respectively extracting feature information of each information feature layer according to preset multi-scale feature learning, and generating information feature nodes corresponding to the information feature layers;
step S30: performing subspace mapping on the information characteristic layer and the corresponding information characteristic nodes to obtain a mapping result, and judging the mapping result; wherein the content of the first and second substances,
when the mapping result is within a preset threshold range, the passenger car is in compliance;
when the mapping result is not within the preset threshold range, acquiring the illegal passenger car information for the illegal passenger car, and performing preset violation coping processing;
the working principle of the technical scheme is as follows: different from the prior art in which video analysis and identification and fixed-point image analysis are performed on vehicles, the above technical scheme performs characteristic layering on the information of the vehicle to be detected, and obtains an information characteristic layer, including: the method comprises the steps of carrying out scale layering, cost layering and training layering, then respectively extracting characteristic information of each information characteristic layer according to preset multi-scale characteristic learning, generating information characteristic nodes corresponding to the information characteristic layers, carrying out subspace mapping on the information characteristic layers and the corresponding information characteristic nodes, obtaining a mapping result, and when the mapping result is within a preset threshold range, determining that the passenger car is in compliance; when the mapping result is not within the preset threshold range, acquiring the information of the illegal passenger car for the illegal passenger car, and performing preset violation coping processing;
the beneficial effects of the above technical scheme are: by means of characteristic layering and subdivision of passenger car detection modes, passenger car detection accuracy and detection efficiency are improved.
In one embodiment, the passenger vehicle license plate identification comprises the steps of:
step S100: extracting convolution characteristics through a preset detection convolution neural network to obtain a characteristic diagram; wherein the content of the first and second substances,
the convolution feature extraction includes: convolution, activating function and constructing a pooling layer;
step S200: carrying out preset feature network analysis processing according to the feature map to generate a passenger vehicle detection frame; wherein the content of the first and second substances,
the feature network analysis processing includes: carrying out regional proposal network processing and full connection layer processing; wherein the content of the first and second substances,
the area proposal network processing comprises: obtaining nodes of a region to be selected by generating the region to be selected, judging the node attributes of the region to be selected, and performing bounding box regression processing according to the node attributes to determine a first region to be selected; wherein the content of the first and second substances,
the node attributes include: foreground nodes and background nodes;
the full connection layer processing comprises: according to the first region to be selected, extracting comprehensive information to generate a feature map of the region to be selected, and performing frame regression calculation according to the feature map of the region to be selected to obtain a passenger car detection frame;
step S300: carrying out passenger vehicle violation detection according to the passenger vehicle detection frame, and acquiring passenger vehicle violation information;
the working principle of the technical scheme is as follows: with the special scene among the prior art scheme, license plate number discernment is difficult, can't solve the unclear condition of license plate, among the above-mentioned technical scheme, through carrying out convolution characteristic extraction to detecting convolution neural network, obtain the characteristic map, carry out predetermined characteristic network analysis according to the characteristic map and handle, generate passenger car and detect the frame, characteristic network analysis and handle includes: the method comprises the steps of regional proposal network processing and full-connection layer processing, wherein the regional proposal network processing acquires nodes of a region to be selected by generating the region to be selected, judges the node attributes of the region to be selected, performs bounding box regression processing, determines a first region to be selected, performs comprehensive information extraction according to the first region to be selected by full-connection layer processing, generates a feature map of the region to be selected, performs bounding box regression calculation, acquires a passenger car detection bounding box, and finally performs passenger car violation detection to acquire passenger car violation information;
the beneficial effects of the above technical scheme are: the passenger car violation detection frame is determined by analyzing the area to be selected, so that the passenger car violation detection accuracy is improved.
In one embodiment, the violation detection comprises:
the method comprises the steps of acquiring segmented detection information by carrying out segmented detection on the passenger vehicle; wherein the content of the first and second substances,
the segment detection includes: time subsection detection and route subsection detection;
according to the segmentation detection information, carrying out segmentation similarity analysis to obtain a segmentation similarity value, and judging; wherein the content of the first and second substances,
the segmentation similarity analysis is to carry out similarity calculation analysis on each piece of information and the previous piece of information;
when the segment similarity value is within a preset threshold range, the segment is a normal segment;
when the segment similarity value is not within the preset threshold range, performing abnormal analysis processing on the abnormal segment;
the working principle of the technical scheme is as follows: firstly, segmented detection is carried out on the passenger vehicles to obtain segmented detection information, and the segmented detection comprises the following steps: time segmentation detection and route segmentation detection, then, carrying out similarity calculation analysis on each segment of information and the previous segment of information to obtain a segmentation similarity value, and when the segmentation similarity value is within a preset threshold range, carrying out normal segmentation; when the segment similarity value is not within the preset threshold range, performing abnormal analysis processing for abnormal segments;
the beneficial effects of the above technical scheme are: by means of subsection detection of the passenger vehicles and calculation of subsection similarity values, accuracy and detection efficiency of violation detection are improved, abnormal conditions can be found as early as possible, exception handling is conducted, and driving safety of the passenger vehicles is improved.
In one embodiment, the violation detection further comprises:
detecting the portrait violation of the vehicle violation information to obtain portrait violation data; wherein the content of the first and second substances,
the portrait violation detection includes: detecting getting-on behaviors and getting-off behaviors;
performing data comparison processing according to the portrait violation data and a preset violation database to obtain violation levels; wherein the content of the first and second substances,
the violation levels include: minor, common, severe violations;
performing corresponding safety early warning according to the violation level to acquire early warning information; wherein the content of the first and second substances,
the safety precaution includes: the vehicle where the illegal portrait is located is warned in a networking mode according to the violation level, and warning information is obtained;
the illegal vehicle carries out illegal correction processing according to the warning information to generate illegal correction information;
the early warning information includes: warning information, violation correction information and positioning information;
the working principle of the technical scheme is as follows: with carrying out vehicle detection to the vehicle among the prior art scheme to it is different to upload vehicle information in violation, carry out portrait violation detection to vehicle violation information among the above-mentioned technical scheme, obtain portrait violation data, portrait violation detection includes: the method comprises the following steps of detecting getting-on behavior and getting-off behavior, carrying out data comparison processing according to a preset violation database, and acquiring violation levels, wherein the steps comprise: slight violation, common violation and serious violation, corresponding safety early warning is carried out according to the violation level, and early warning information is obtained, wherein the early warning information comprises the following steps: warning information, violation correction information and positioning information;
the beneficial effects of the above technical scheme are: the detection precision of getting on and off the bus violation of the passenger car is improved by comprehensively detecting the portrait, and the safety of the handling scheme of the violation of the passenger car is improved by detecting the portrait violation and detecting the violation level.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. An analysis method for illegal boarding and alighting based on passenger vehicles comprises the following steps:
step S01: vehicle violation screening is carried out on passenger transport monitoring data, and violation boarding and alighting characteristic values are obtained;
step S02: according to the characteristic values of the illegal passengers getting on and getting off, characteristic analysis is carried out, and a vehicle illegal model is constructed;
step S03: generating output value similarity by performing model comparison calculation on the vehicle violation model;
step S04: according to the similarity of the output values, deviation identification is carried out to obtain information of the vehicle to be detected;
step S05: and carrying out violation detection on the vehicle to be detected according to the information of the vehicle to be detected, and acquiring violation information of the vehicle.
2. The method for analyzing passengers getting on or off based on passenger vehicle violation according to claim 1, wherein said step S01 comprises:
the method comprises the steps of classifying passenger transport monitoring data according to data functions to obtain monitoring data categories; wherein the content of the first and second substances,
the passenger traffic monitoring data comprises: bayonet data and video structured data;
the monitoring data categories include: number category, speed category, time category, behavior category;
and performing category feature extraction on each detection data category to generate a category feature value of each category, and performing passenger getting-on and getting-off analysis on the category feature value to obtain a violation passenger getting-on and getting-off feature value.
3. The method for analyzing passengers getting on or off based on passenger vehicle violation according to claim 1, wherein said step S02 comprises:
modeling and screening the characteristic values of the illegal upper and lower passengers to obtain the characteristic information of the illegal model; wherein the content of the first and second substances,
the violation model feature information includes: position information, node information, edge information and interval information;
and constructing a main body model according to the violation model characteristic information to generate a vehicle violation model.
4. The passenger vehicle violation boarding and disembarking based analysis method as claimed in claim 1, wherein said model comparison calculation comprises the steps of:
the method comprises the following steps: normalization data is obtained by performing normalization processing on the vehicle default scale type data; wherein the content of the first and second substances,
the normalization processing maps the conversion result to a preset interval by carrying out dispersion standardization on the vehicle violation model data; wherein the content of the first and second substances,
the dispersion normalization includes: performing linear transformation on the vehicle violation model data through a preset conversion function;
step two: transmitting the normalized data to a twin network for mapping analysis processing to generate twin characteristic space data; wherein the content of the first and second substances,
the twin network includes: extracting dual-input data of the normalized data, inputting the dual-input data into a neural network, and performing mapping processing; wherein the content of the first and second substances,
the dual input data includes: front frame data and rear frame data;
step three: performing output calculation on the twin characteristic space data to obtain output value similarity, and judging; wherein the content of the first and second substances,
when the similarity of the output values is smaller than or equal to a preset threshold value, the vehicle behavior is normal;
and when the output value similarity is larger than a preset threshold value, judging that the vehicle behavior is abnormal.
5. The method for analyzing passengers getting on or off based on passenger vehicle violation according to claim 1, wherein said step S04 comprises:
matching a corresponding calculation method according to the similarity of the output values, performing deviation calculation, acquiring a deviation value, and judging; wherein the content of the first and second substances,
the deviation calculation includes: according to the corresponding calculation method, comparing and calculating the output value with a preset output comparison data set to obtain a comparison deviation value;
when the deviation value is larger than a preset threshold value, determining the deviation value is an overlarge deviation, determining the vehicle to be detected, and acquiring information of the vehicle to be detected;
and when the deviation value is less than or equal to a preset threshold value, determining that the deviation value is a normal deviation.
6. The method for analyzing passengers getting on or off based on passenger vehicle violation according to claim 1, wherein said step S05 comprises:
the method comprises the steps that key data screening is conducted on information of a vehicle to be detected, and first screening information is obtained; wherein the content of the first and second substances,
the key data screening comprises the following steps: screening the moving state and screening the path;
extracting vehicle information from the first screening information to obtain a first vehicle to be detected, and carrying out violation detection on the first vehicle to be detected to obtain vehicle violation information; wherein the content of the first and second substances,
the violation detection includes: passenger car detection and passenger car license plate identification.
7. The passenger vehicle violation boarding and disembarking based analysis method according to claim 6, wherein the passenger vehicle detection comprises the steps of:
step S10: the method comprises the steps that characteristic layering is conducted on vehicle information to be detected, and an information characteristic layer is obtained; wherein the content of the first and second substances,
the feature layering includes: scale layering, cost layering, training layering;
step S20: respectively extracting feature information of each information feature layer according to preset multi-scale feature learning, and generating information feature nodes corresponding to the information feature layers;
step S30: performing subspace mapping on the information characteristic layer and the corresponding information characteristic nodes to obtain a mapping result, and judging the mapping result; wherein the content of the first and second substances,
when the mapping result is within a preset threshold range, the passenger car is in compliance;
and when the mapping result is not within the preset threshold range, acquiring the illegal passenger car information for the illegal passenger car, and performing preset violation coping processing.
8. The passenger vehicle violation boarding and disembarking based analysis method according to claim 6, wherein the passenger vehicle license plate recognition comprises the steps of:
step S100: extracting convolution characteristics through a preset detection convolution neural network to obtain a characteristic diagram; wherein the content of the first and second substances,
the convolution feature extraction includes: convolution, activating function and constructing a pooling layer;
step S200: carrying out preset feature network analysis processing according to the feature map to generate a passenger vehicle detection frame; wherein the content of the first and second substances,
the feature network analysis processing includes: carrying out regional proposal network processing and full connection layer processing; wherein the content of the first and second substances,
the area proposal network processing comprises: obtaining nodes of a region to be selected by generating the region to be selected, judging the node attributes of the region to be selected, and performing bounding box regression processing according to the node attributes to determine a first region to be selected; wherein the content of the first and second substances,
the node attributes include: foreground nodes and background nodes;
the full connection layer processing comprises: according to the first region to be selected, extracting comprehensive information to generate a feature map of the region to be selected, and performing frame regression calculation according to the feature map of the region to be selected to obtain a passenger car detection frame;
step S300: and carrying out passenger vehicle violation detection according to the passenger vehicle detection frame, and acquiring passenger vehicle violation information.
9. The passenger vehicle violation boarding and disembarking based analysis method as claimed in claim 1, wherein said violation detection comprises:
the method comprises the steps of acquiring segmented detection information by carrying out segmented detection on the passenger vehicle; wherein the content of the first and second substances,
the segment detection includes: time subsection detection and route subsection detection;
according to the segmentation detection information, carrying out segmentation similarity analysis to obtain a segmentation similarity value, and judging; wherein the content of the first and second substances,
the segmentation similarity analysis is to carry out similarity calculation analysis on each piece of information and the previous piece of information;
when the segment similarity value is within a preset threshold range, the segment is a normal segment;
and when the segment similarity value is not in the preset threshold range, performing abnormal segmentation and performing abnormal analysis processing.
10. The passenger vehicle violation boarding and disembarking based analysis method as set forth in claim 1, wherein the violation detection further comprises:
detecting the portrait violation of the vehicle violation information to obtain portrait violation data; wherein the content of the first and second substances,
the portrait violation detection includes: detecting getting-on behaviors and getting-off behaviors;
performing data comparison processing according to the portrait violation data and a preset violation database to obtain violation levels; wherein the content of the first and second substances,
the violation levels include: minor, common, severe violations;
performing corresponding safety early warning according to the violation level to acquire early warning information; wherein the content of the first and second substances,
the safety precaution includes: the vehicle where the illegal portrait is located is warned in a networking mode according to the violation level, and warning information is obtained;
the illegal vehicle carries out illegal correction processing according to the warning information to generate illegal correction information;
the early warning information includes: warning information, violation correction information, and positioning information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111636282.0A CN114282623A (en) | 2021-12-29 | 2021-12-29 | Analysis method based on illegal boarding and alighting of passenger vehicles |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111636282.0A CN114282623A (en) | 2021-12-29 | 2021-12-29 | Analysis method based on illegal boarding and alighting of passenger vehicles |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114282623A true CN114282623A (en) | 2022-04-05 |
Family
ID=80877940
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111636282.0A Pending CN114282623A (en) | 2021-12-29 | 2021-12-29 | Analysis method based on illegal boarding and alighting of passenger vehicles |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114282623A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311633A (en) * | 2022-10-11 | 2022-11-08 | 深圳市旗扬特种装备技术工程有限公司 | Method and device for detecting illegal boarding and alighting of vehicle, electronic equipment and storage medium |
-
2021
- 2021-12-29 CN CN202111636282.0A patent/CN114282623A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311633A (en) * | 2022-10-11 | 2022-11-08 | 深圳市旗扬特种装备技术工程有限公司 | Method and device for detecting illegal boarding and alighting of vehicle, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11074813B2 (en) | Driver behavior monitoring | |
| WO2020052436A1 (en) | Vehicle overload alarming method and apparatus, electronic device, and storage medium | |
| CN111583639B (en) | Road traffic jam early warning method and system | |
| CN106652468A (en) | Device and method for detection of violation of front vehicle and early warning of violation of vehicle on road | |
| Peng et al. | Uncertainty evaluation of object detection algorithms for autonomous vehicles | |
| WO2017123665A1 (en) | Driver behavior monitoring | |
| CN111597905A (en) | Highway tunnel parking detection method based on video technology | |
| CN104821025A (en) | Passenger flow detection method and detection system thereof | |
| CN114282623A (en) | Analysis method based on illegal boarding and alighting of passenger vehicles | |
| Mokhtarimousavi et al. | A temporal investigation of crash severity factors in worker-involved work zone crashes: Random parameters and machine learning approaches | |
| CN114694060B (en) | Road casting detection method, electronic equipment and storage medium | |
| CN113610014A (en) | System and method for detecting freight vehicle with shielding number plate exceeding limit | |
| Doycheva et al. | Computer vision and deep learning for real-time pavement distress detection | |
| Chen et al. | The impact of truck proportion on traffic safety using surrogate safety measures in China | |
| CN114627643B (en) | Highway accident risk prediction method, device, equipment and medium | |
| CN111027365B (en) | Positioning method based on human face object position analysis | |
| CN113158922A (en) | Traffic flow statistical method, device and equipment based on YOLO neural network | |
| CN114141022A (en) | Emergency lane occupation behavior detection method and device, electronic equipment and storage medium | |
| CN113034895A (en) | ETC portal system, and expressway fatigue driving early warning method and device | |
| CN110718064A (en) | Vehicle violation detection method and device | |
| Jain et al. | A computational model for driver risk evaluation and crash prediction using contextual data from on-board telematics | |
| KR102484407B1 (en) | Taxi dispatch system using AI | |
| CN116343484B (en) | Traffic accident identification method, terminal and storage medium | |
| CN111341153B (en) | Early warning method, device, server and medium for travel driving | |
| CN117746632A (en) | Bus driving risk assessment and early warning system and method based on intelligent network |
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 |