CN117611109B - Method and system for monitoring and managing illegal article delivery information - Google Patents
Method and system for monitoring and managing illegal article delivery information Download PDFInfo
- Publication number
- CN117611109B CN117611109B CN202410094823.9A CN202410094823A CN117611109B CN 117611109 B CN117611109 B CN 117611109B CN 202410094823 A CN202410094823 A CN 202410094823A CN 117611109 B CN117611109 B CN 117611109B
- Authority
- CN
- China
- Prior art keywords
- feature vector
- local
- feature vectors
- degree
- parameter
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000012544 monitoring process Methods 0.000 title claims abstract description 38
- 239000013598 vector Substances 0.000 claims abstract description 254
- 230000008859 change Effects 0.000 claims abstract description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 13
- 238000013507 mapping Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 abstract description 15
- 238000007726 management method Methods 0.000 abstract description 7
- 230000009467 reduction Effects 0.000 description 12
- 238000004458 analytical method Methods 0.000 description 9
- 238000012545 processing Methods 0.000 description 5
- 230000005856 abnormality Effects 0.000 description 4
- 230000000875 corresponding effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012937 correction Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000002955 isolation Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Tourism & Hospitality (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Marketing (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of illegal item delivery monitoring and management, in particular to a method and a system for monitoring and managing illegal item delivery information. Firstly, acquiring a feature vector of a put-in article; further acquiring local density parameters, local regularity and texture curves of each feature vector; further combining the local density parameter, the local regularity degree and the change characteristics of the texture curve of each feature vector to obtain the importance degree of each feature vector; further reducing the dimension of the feature vector to obtain main feature data; and finally, carrying out database matching, and monitoring and managing the put-in articles. According to the method, the influence of abnormal data in the collected data caused by pollution, damage, folds and the like in the delivered articles is reduced by analyzing the local density characteristics and the local rule characteristics of the delivered articles and combining the change characteristics of the texture curve, and the accuracy of monitoring and managing the delivery of illegal articles is improved.
Description
Technical Field
The invention relates to the technical field of illegal item delivery monitoring and management, in particular to a method and a system for monitoring and managing illegal item delivery information.
Background
With the deep propaganda of the national energy-saving and environment-friendly concepts, the environmental protection consciousness of residents is greatly enhanced. Domestic garbage of residents serves as a main source of garbage of a community, and intelligent garbage cans are arranged in the community and used for assisting garbage classification conditions of the residents. The intelligent garbage cabinet equipment matches garbage information with garbage types selected by residents in the intelligent garbage cabinet through garbage information placed by residents collected by the camera and the sensor, judges whether the garbage throwing behavior of the residents has violations according to the matching result, does not throw illegal articles, requests the residents to recognize again so as to achieve the purpose of garbage classification, and improves environmental awareness.
When the garbage information thrown by residents is matched, a large amount of redundant information exists in the collected garbage information, so that the accuracy of garbage type identification can be influenced, important data in collected garbage information data need to be extracted, and important features are identified and matched. However, in an actual scene, the acquired garbage data have problems of pollution, breakage, wrinkling and the like, so that the acquired feature vectors have extremely abnormal data, and the extremely abnormal data can interfere with the effect of data dimension reduction, so that the dimension-reduced data have certain deviation to the extremely abnormal data, and the accuracy of matching, identifying and classifying the subsequent garbage data is affected.
Disclosure of Invention
In order to solve the technical problem that the dimension reduction of the information data of the put-in articles is not accurate enough and the monitoring and management of the put-in illegal articles are affected, the invention aims to provide a method and a system for monitoring and managing the information of the put-in illegal articles, and the adopted technical scheme is as follows:
A method for monitoring and managing information of release of illegal articles, the method comprising:
Acquiring a feature vector of a put object in a garbage putting area;
According to the distribution characteristics of the characteristic vectors of all the articles put in the garbage putting area, acquiring the distance parameter between any two characteristic vectors; obtaining a local density parameter of each feature vector according to the distance parameter; according to the change characteristics of the feature vectors in different direction intervals in the local neighborhood of each feature vector, obtaining the local rule degree of each feature vector; judging whether the directions of the feature vectors are similar, wherein the adjacent feature vectors with similar directions form a texture curve;
combining the local density parameter, the local regularity degree and the change characteristic of the texture curve of each feature vector to obtain the importance degree of each feature vector;
According to the importance degree of all the feature vectors, reducing the dimension of the feature vectors to obtain main feature data; and matching the main characteristic data with the database, and monitoring and managing the put-in articles.
Further, the method for acquiring the distance parameter comprises the following steps:
selecting any two feature vectors as target feature vector pairs;
Acquiring Euclidean distance between the target feature vector pairs and normalizing the Euclidean distance as a first distance subparameter;
Acquiring a second distance subparameter according to the angle difference and the amplitude difference between the target feature vector pairs;
And taking Euclidean distance between the first distance subparameter and the second distance subparameter of the target feature vector pair as a distance parameter between the target feature vector pair.
Further, the method for obtaining the second distance subparameter comprises the following steps:
Acquiring an angle value of an included angle between the target feature vector pair as an angle parameter; acquiring the absolute value of the difference value of the amplitude values between the target feature vector pairs as an amplitude parameter; and normalizing the product of the angle parameter and the amplitude parameter to be used as a second distance subparameter.
Further, the method for obtaining the local density parameter comprises the following steps:
Setting a K value of an LOF algorithm according to a preset nearest neighbor number, and selecting K nearest feature vectors as local neighbors forming each feature vector;
and according to the distance parameter, obtaining the local reachable density of each feature vector, and taking the local reachable density as a local density parameter.
Further, the method for obtaining the local regularity comprises the following steps:
Dividing the plane into different direction intervals according to preset dividing parameters;
Obtaining the local rule degree of each feature vector according to a local rule degree calculation formula; the local regularity calculation formula comprises:
; wherein/> Represents the/>Local degree of regularity of the individual feature vectors; /(I)Number indicating direction section,/>，/>Indicating the number of directional intervals; /(I)Expressed as natural constant/>An exponential function of the base; /(I)Representing an obtained minimum function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The variance of the feature vector is contained in each direction interval; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The number of feature vectors is contained in each direction interval; /(I)Represents the/>Variance of feature vectors in all direction intervals in local neighborhood of each feature vector.
Further, the method for obtaining the importance degree includes:
obtaining discrete credibility parameters of each feature vector according to the change characteristics of the texture curve where each feature vector is located; acquiring the outlier degree of each feature vector according to an outlier degree calculation formula; and mapping the outlier degree negative correlation to obtain the importance degree of each feature vector.
Further, the outlier degree calculation formula includes:
; wherein/> Represents the/>The degree of outlier of the individual feature vectors; /(I)Representing a normalization function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local density parameters of the individual feature vectors; /(I)Represents the/>The number of feature vectors within the local neighborhood of the individual feature vectors; /(I)Represents the/>Local density parameters of the individual feature vectors; /(I)Represents the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>The length of the texture curve in which the feature vectors are located; /(I)Represents the/>Variance of slope of texture curve where each feature vector is located; /(I)Represents the/>Discrete reliability parameters of the individual feature vectors.
Further, the method for acquiring the discrete credibility parameter comprises the following steps:
And taking the ratio of the variance of the slope of the texture curve where each feature vector is located to the curve length as the discrete credibility parameter of each feature vector.
Further, the method for monitoring and managing the delivered objects comprises the following steps:
And matching the main characteristic data with a database to obtain the matching degree of the released articles, and judging that the released articles have violations when the matching degree is lower than a preset matching threshold value, and sending out a violation release warning.
The invention also provides a system for monitoring and managing the illegal item throwing information, which comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor realizes the steps of any illegal item throwing information monitoring and managing method when executing the computer program.
The invention has the following beneficial effects:
Aiming at the technical problems that the dimension reduction of the information data of the released articles is inaccurate and the monitoring management of the released illegal articles is affected, the invention firstly obtains the feature vector of the released articles and provides a data basis for the subsequent analysis; further acquiring distance parameters between any two feature vectors, and preparing for next analysis of density features; further obtaining local density parameters of each feature vector, providing a basis for analyzing the degree of abnormality of the feature vector from the viewpoint of density features, and preparing for the subsequent analysis of the degree of importance of the feature vector; further acquiring the local degree of regularity of each feature vector, analyzing the abnormal features of the feature vectors from the distribution degree of regularity, carrying out certain supplementary correction on the local density parameters, and improving the reliability of the parameters according to the subsequent calculation of the degree of importance; further acquiring a texture curve, providing more basis for the subsequent analysis of the importance degree, improving the accuracy of the importance degree, and finally improving the accuracy of monitoring and managing the put-in articles; further combining the local density parameter, the local regularity degree and the change characteristic of the texture curve of each feature vector, and synthesizing various parameters to obtain the importance degree of each feature vector, so as to prepare for accurate dimension reduction of data; further reducing the influence of the abnormal data on the data dimension reduction process through the importance degree, reducing the dimension of the feature vector, obtaining accurate main feature data, and providing a data basis for finally carrying out database matching and judging whether the put-in articles violate rules or not; and finally, carrying out database matching, and monitoring and managing the put-in articles. According to the method, the influence of abnormal data in the collected data caused by pollution, damage, folds and the like in the delivered articles is reduced by analyzing the local density characteristics and the local rule characteristics of the delivered articles and combining the change characteristics of the texture curve, and the accuracy of monitoring and managing the delivery of illegal articles is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for monitoring and managing information of putting illegal articles according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of a method and a system for monitoring and managing information of illegal object delivery according to the invention, which are specific embodiments, structures, features and effects thereof, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the method and system for monitoring and managing illegal item delivery information provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a method for monitoring and managing information of putting illegal articles according to an embodiment of the present invention specifically includes:
step S1: and acquiring the characteristic vector of the put-in articles in the garbage putting area.
In the embodiment of the invention, the surface information data of the delivered object is considered to be high-dimensional, and the feature vector extraction can represent the surface information in a more compact form, so that the dimension of the data is reduced, the storage and calculation cost is reduced, and the algorithm efficiency is improved; the feature vector can capture important structure and texture information in the surface information of the put article, so that the algorithm has certain robustness to changes such as illumination, visual angle, partial shielding and the like, and the accuracy of identification and matching is improved; the feature vector of the surface information data of the article to be delivered is first acquired.
In one embodiment of the invention, when the HOG algorithm is considered to extract the feature vector, the gradient direction is mainly concerned, the influence of the change of the illumination intensity is smaller, the calculation is simpler, and less calculation resources are needed, so the HOG algorithm is utilized to obtain the feature vector; in other embodiments of the present invention, the practitioner may acquire the feature vector by using other methods capable of extracting the feature vector in the prior art, and the HOG algorithm is a technical means well known to those skilled in the art, and will not be described herein.
Step S2: according to the distribution characteristics of the characteristic vectors of all the articles put in the garbage putting area, acquiring the distance parameter between any two characteristic vectors; obtaining a local density parameter of each feature vector according to the distance parameter; according to the change characteristics of the feature vectors in different direction intervals in the local neighborhood of each feature vector, obtaining the local rule degree of each feature vector; and judging whether the directions of the feature vectors are similar, and forming a texture curve by the adjacent feature vectors with similar directions.
In the embodiment of the invention, the problems of pollution, breakage, wrinkling and the like in the collected data of the garbage are considered, the obtained feature vectors have extremely abnormal data, the extremely abnormal data can interfere the effect of data dimension reduction, so that the dimension-reduced data have certain deviation extremely abnormal data to influence the accuracy of matching, identifying and classifying the subsequent data of the garbage, the feature vectors of the placed objects are required to be analyzed, the importance degree of each feature vector in the dimension reduction of the data is adjusted, the influence caused by the abnormal data is reduced, and more accurate main feature data are obtained and are accurately matched.
In the analysis of the feature vectors, the distance parameters are acquired first, considering that the greater the degree of isolation of the feature vectors, the smaller the local density, and the more likely the abnormal feature vectors are, and the distance between the feature vectors needs to be acquired in the analysis of the density.
Preferably, in one embodiment of the present invention, the distance parameter is obtained using the euclidean distance, considering that the euclidean distance is a common distance measurement method; and because the difference between the feature vectors with lower angle and amplitude similarity is larger and the distance parameter is larger when the distance measurement between the feature vectors is carried out, the similarity condition between the feature vectors also needs to be considered, and the distance parameter between the feature vectors is comprehensively acquired. Based on the above, any two feature vectors are selected as target feature vector pairs;
acquiring Euclidean distance between target feature vector pairs and normalizing the Euclidean distance as a first distance subparameter;
According to the angle difference and the amplitude difference between the target feature vector pairs, acquiring an angle value of an included angle between the target feature vector pairs as an angle parameter; acquiring an absolute value of a difference value of the amplitude values between the target feature vector pairs as an amplitude parameter; normalizing the product of the angle parameter and the amplitude parameter to be used as a second distance subparameter;
The Euclidean distance between the first distance subparameter and the second distance subparameter of the target feature vector pair is used as the distance parameter between the target feature vector pair, and the distance parameter is expressed as follows by a formula:
wherein, Representing a distance parameter between a j-th target feature vector pair; /(I)A first distance subparameter representing the jth target feature vector pair; /(I)A second distance subparameter representing the jth target feature vector pair,，/>Representing a normalization function,/>An included angle value of the jth target feature vector pair is represented; representing the absolute value of the difference in magnitude of the j-th target feature vector pair.
In the calculation formula of the distance parameter, the larger the first distance subparameter is, the larger the distance between the target feature vector pairs obtained from the spatial distance angle measurement is, and the larger the distance parameter is; the larger the second distance subparameter is, the larger the direction angle difference and the amplitude difference between the target feature vector pairs are, the lower the similarity is, and the larger the difference between the target feature vector pairs is, the larger the distance parameter is.
It should be noted that, calculating euclidean distance and normalizing are well known to those skilled in the art, and in other embodiments of the present invention, other basic mathematical operations or function mapping may be used to implement the related mapping, which are well known to those skilled in the art, and will not be described herein.
And changing the target feature vector pair to obtain a distance parameter between any two feature vectors. After the distance parameter is obtained, the local density characteristics of the characteristic vector can be analyzed, and a basis is provided for the importance degree of the subsequent analysis of the characteristic vector from the local density angle.
Preferably, in one embodiment of the present invention, considering that in the LOF algorithm, the local reachable density of the data can be obtained by setting the K value, and the smaller the local reachable density is, the more likely the data is an outlier, the greater the likelihood of abnormality, so the local density parameter of each feature vector is obtained by using the LOF algorithm.
Based on the above, setting a K value of an LOF algorithm according to a preset nearest neighbor number, and selecting K nearest feature vectors as local neighbors forming each feature vector;
and obtaining local reachable density of each feature vector according to the distance parameter, and taking the local reachable density as a local density parameter.
It should be noted that, in one embodiment of the present invention, the nearest neighbor number is preset to be 15, and the K value is set to be 15; the LOF algorithm is a well known technical means for those skilled in the art, and will not be described in detail herein; in other embodiments of the present invention, the implementer may also use a density-based clustering method to cluster the feature vectors, and use the density of the entire cluster in which the feature vectors are located as the local density parameter of the feature vectors.
In the embodiment of the invention, the regularity of the distribution characteristics of the feature vector is considered to reflect the abnormality degree of the feature vector, and the local density parameter can be subjected to certain supplementary correction, so that the reliability of the parameter according to which the importance degree of the feature vector is calculated later is improved, and the distribution regularity characteristics of the feature vector are analyzed in the local neighborhood of the feature vector.
Preferably, in one embodiment of the present invention, considering that the number of feature vectors and fluctuation characteristics of different direction intervals are different in the local neighborhood of the feature vector, the direction interval in which the overall fluctuation is most stable represents the direction interval in which the degree of change rule of the feature vector is highest in the whole local neighborhood of the feature vector, and the more severe the fluctuation, the lower the degree of change rule of the feature vector is reflected; in addition, the situation that the change rule of a single direction interval is obvious and the change rule of other direction intervals is low is considered to be easy to occur by analyzing the local rule degree only through one direction interval, so that the variance of the feature vector in all direction intervals in the local neighborhood of the feature vector is obtained, the stability and the credibility are weighted, the variance of the feature vector in all direction intervals is small, the higher the stability and the credibility of the change rule degree of the selected direction interval are, the more the local rule degree of the whole local neighborhood can be represented, the analysis is performed by combining the local feature and the whole feature, and the convincing power and the accuracy of the obtained local rule degree are improved.
Based on the above, dividing the plane into different direction intervals according to preset dividing parameters;
Obtaining the local rule degree of each feature vector according to a local rule degree calculation formula; the local rule degree calculation formula comprises:
wherein, Represents the/>Local degree of regularity of the individual feature vectors; /(I)Number indicating direction section,/>，Indicating the number of directional intervals; /(I)Expressed as natural constant/>An exponential function of the base; /(I)Representing an obtained minimum function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The variance of the feature vector is contained in each direction interval; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The number of feature vectors is contained in each direction interval; /(I)Represents the/>Variance of feature vectors in all direction intervals in local neighborhood of each feature vector.
It should be noted that, in other embodiments of the present invention, other basic mathematical operations or function mapping may be used to implement the correlation mapping, which is a technical means well known to those skilled in the art, and will not be described herein.
In one embodiment of the present invention, the preset dividing parameter is 36, the plane is equally divided into 36 parts, from 0 deg., every 10 deg. is a direction interval,36. When no feature vector exists in a certain direction interval, ignoring the direction interval; in other embodiments of the invention, an implementer may divide the plane into other numbers of directional intervals.
In the embodiment of the invention, the characteristics of ductility or continuity of the feature vectors are considered, namely, the feature vectors with similar directions exist around the feature vectors, and the more adjacent and similar feature vectors are, the longer the texture curve formed, the more obvious the texture characteristics of the feature vectors are, the lower the abnormal possibility of the feature vectors is, the lower the possibility of interference is, the greater the importance degree is, so that the texture curve is obtained, more basis is provided for the importance degree of subsequent analysis, the accuracy of the importance degree is improved, and the accuracy of monitoring and managing the put articles is finally improved.
Preferably, in one embodiment of the present invention, the smaller the angle between the vectors, the higher the similarity is considered, so when the angle between two feature vectors is smaller than 10 °, the two feature vectors are considered to be similar in direction, and the feature vectors that are adjacent and similar in direction form a texture curve.
It should be noted that, in one embodiment of the present invention, when a neighborhood around a feature vector has a plurality of similar feature vectors, a feature vector corresponding to the feature vector with the smallest included angle is selected; in another embodiment of the present invention, all similar feature vectors may be initially reserved to form a texture curve with branches, and finally, the longest one of all curve branches is selected for reservation; in other embodiments of the present invention, the practitioner may set the threshold for the included angle when determining similarity.
When the texture curve is acquired, eight feature vectors are connected with one feature vector, so that eight adjacent domains are selected as feature vectors to find adjacent domains of similar feature vectors.
Step S3: and combining the local density parameter, the local regularity degree and the change characteristic of the texture curve of each feature vector to acquire the importance degree of each feature vector.
Through step S2, a plurality of calculation bases of importance degrees are obtained, so that the importance degrees of each feature vector are obtained by combining the plurality of calculation bases.
Preferably, in one embodiment of the present invention, the smoother the variation of the texture curve is considered, the longer the length is, the greater the possibility that the corresponding feature vector does not belong to an interference point is, the lower the outlier is, and the greater the importance level is; the larger the local density parameter of the feature vector is compared with the local density parameter mean value of other feature vectors in the local neighborhood, the larger the local density parameter of the feature vector is, the smaller the outlier degree of the feature vector is, the smaller the abnormality probability is, and the greater the importance degree is; the local density degree can be further corrected by considering the local rule degree, so the local density degree is weighted by the local rule degree, and the larger the local rule degree is, the smaller the outlier degree of the feature vector is, so the local rule degree and the local density degree are positively correlated.
Based on the above, according to the change characteristics of the texture curve where each feature vector is located, taking the ratio of the variance of the slope of the texture curve where each feature vector is located to the curve length as the discrete credibility parameter of each feature vector; acquiring the outlier degree of each feature vector according to an outlier degree calculation formula; mapping the outlier degree negative correlation to obtain the importance degree of each feature vector;
The outlier degree calculation formula includes:
wherein, Represents the/>The degree of outlier of the individual feature vectors; /(I)Representing a normalization function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local density parameters of the individual feature vectors; /(I)Represents the/>The number of feature vectors within the local neighborhood of the individual feature vectors; Represents the/> Local density parameters of the individual feature vectors; /(I)Represents the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>The length of the texture curve in which the feature vectors are located; /(I)Represents the/>Variance of slope of texture curve where each feature vector is located; Represents the/> Discrete reliability parameters of the individual feature vectors.
It should be noted that, when one feature vector is not similar to any adjacent feature vector, a texture curve cannot be constructed and analyzed, and the feature vector is considered to have a high degree of isolation and outlier, so that the discrete reliability parameter is 1.
In one embodiment of the invention, the importance level is obtained by directly subtracting the outlier level of the feature vector from 1 in consideration of the fact that the importance level and the outlier level have a negative correlation and the outlier level is subjected to normalization processing，Represents the/>The degree of importance of the feature vectors.
It should be noted that, in other embodiments of the present invention, when calculating the outlier degree and the importance degree, other basic mathematical operations or function mapping may be used to implement the related mapping, which is a technical means known to those skilled in the art, and will not be described herein.
Step S4: according to the importance degree of all the feature vectors, reducing the dimension of the feature vectors to obtain main feature data; and matching the main characteristic data with the database, and monitoring and managing the put-in articles.
After the processing of the step S3, the importance degree of each feature vector is obtained, the importance degree of the feature vector can be combined, the interference of abnormal data is reduced, the feature vector is subjected to accurate dimension reduction, data matching is performed, whether the delivered object is illegal or not is judged, and monitoring and management are performed on the delivered object.
Preferably, in one embodiment of the present invention, considering that the PCA dimension reduction algorithm is a commonly used dimension reduction method, and has a fast processing speed, and is applicable to monitoring and management of put-in article information, the PCA dimension reduction algorithm is selected to reduce the dimension of the feature vector; in other embodiments of the present invention, the practitioner may select other data dimension reduction methods such as LLE algorithm, UMAP algorithm, etc. to reduce the dimension of the feature vector, which are all technical means known to those skilled in the art, and will not be described herein.
Preferably, in one embodiment of the present invention, considering that the K nearest neighbor algorithm does not need training for data matching, the method is a simpler and intuitive algorithm and has a certain adaptability, so that the K nearest algorithm is selected for matching main feature data with the database to obtain the matching degree of the delivered item, and when the matching degree is lower than a preset matching threshold, it is determined that the delivered item has a violation, and a violation delivery warning is issued.
It should be noted that, the K nearest neighbor algorithm is a technical means well known to those skilled in the art, and will not be described herein again; in one embodiment of the present invention, the preset matching threshold is 0.65, and in other embodiments of the present invention, an implementer may set other matching thresholds, and replace the K nearest neighbor algorithm with a decision tree to obtain the matching degree.
In another embodiment of the invention, considering the possible situation of entrainment during garbage throwing, one garbage wraps other garbage, so that a gravity sensor is also designed to acquire the quality information of garbage throwing, the quality information of various garbage is supplemented in a database, the ratio of the garbage throwing quality to the garbage in the database is taken as a quality deviation parameter, when the quality deviation parameter is more than 2, the problem of inclusion in garbage throwing is considered, and residents are requested to throw in again after processing.
In summary, the feature vector of the delivered item is firstly obtained according to the technical problems that the dimension reduction of the information data of the delivered item is not accurate enough and the monitoring and management of the delivery of the illegal item are affected; further acquiring local density parameters, local regularity and texture curves of each feature vector; further combining the local density parameter, the local regularity degree and the change characteristics of the texture curve of each feature vector to obtain the importance degree of each feature vector; further reducing the dimension of the feature vector to obtain main feature data; and finally, carrying out database matching, and monitoring and managing the put-in articles. According to the method, the influence of abnormal data in the collected data caused by pollution, damage, folds and the like in the delivered articles is reduced by analyzing the local density characteristics and the local rule characteristics of the delivered articles and combining the change characteristics of the texture curve, and the accuracy of monitoring and managing the delivery of illegal articles is improved.
An embodiment of the present invention provides a system for monitoring and managing information of putting in illegal articles, which includes a memory, a processor and a computer program, wherein the memory is used for storing a corresponding computer program, the processor is used for running the corresponding computer program, and the computer program can implement a method for monitoring and managing information of putting in illegal articles described in steps S1 to S4 when running in the processor.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (3)
1. A method for monitoring and managing information of release of illegal articles, the method comprising:
Acquiring a feature vector of a put object in a garbage putting area;
according to the distribution characteristics of the characteristic vectors of all the articles put in the garbage putting area, acquiring distance parameters between any two characteristic vectors; obtaining a local density parameter of each feature vector according to the distance parameter; according to the change characteristics of the feature vectors in different direction intervals in the local neighborhood of each feature vector, obtaining the local regularity of each feature vector; judging whether the directions of the feature vectors are similar, wherein the feature vectors which are adjacent and similar in direction form a texture curve;
combining the local density parameter, the local regularity degree and the change characteristics of the texture curve of each feature vector to obtain the importance degree of each feature vector;
according to the importance degrees of all the feature vectors, reducing the dimension of the feature vectors to obtain main feature data; according to the main characteristic data, matching with a database, and monitoring and managing the put-in articles;
the distance parameter obtaining method comprises the following steps:
selecting any two feature vectors as target feature vector pairs;
Acquiring Euclidean distance between the target feature vector pairs and normalizing the Euclidean distance as a first distance subparameter;
Acquiring a second distance subparameter according to the angle difference and the amplitude difference between the target feature vector pairs;
taking Euclidean distance between the first distance subparameter and the second distance subparameter of the target feature vector pair as a distance parameter between the target feature vector pair;
the method for acquiring the second distance subparameter comprises the following steps:
acquiring an angle value of an included angle between the target feature vector pair as an angle parameter; acquiring the absolute value of the difference value of the amplitude values between the target feature vector pairs as an amplitude parameter; normalizing the product of the angle parameter and the amplitude parameter to be used as a second distance subparameter;
the method for acquiring the local density parameter comprises the following steps:
Setting a K value of an LOF algorithm according to a preset nearest neighbor number, and selecting K nearest feature vectors as local neighbors forming each feature vector;
according to the distance parameters, obtaining local reachable density of each feature vector, and taking the local reachable density as a local density parameter;
The method for acquiring the local regularity comprises the following steps:
Dividing the plane into different direction intervals according to preset dividing parameters;
Obtaining the local rule degree of each feature vector according to a local rule degree calculation formula; the local regularity calculation formula comprises:
; wherein/> Represents the/>Local degree of regularity of the individual feature vectors; /(I)Number indicating direction section,/>，/>Indicating the number of directional intervals; /(I)Expressed as natural constant/>An exponential function of the base; /(I)Representing an obtained minimum function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The variance of the feature vector is contained in each direction interval; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>The number of feature vectors is contained in each direction interval; /(I)Represents the/>In the local neighborhood of each feature vector, the variance of the feature vector in all direction intervals;
the method for acquiring the importance degree comprises the following steps:
Obtaining discrete credibility parameters of each feature vector according to the change characteristics of the texture curve where each feature vector is located; acquiring the outlier degree of each feature vector according to an outlier degree calculation formula; mapping the outlier degree negative correlation to obtain the importance degree of each feature vector;
The outlier degree calculation formula comprises:
; wherein/> Represents the/>The degree of outlier of the individual feature vectors; /(I)Representing a normalization function; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>Within the local neighborhood of the individual feature vectors, the/>Local density parameters of the individual feature vectors; /(I)Represents the/>The number of feature vectors within the local neighborhood of the individual feature vectors; /(I)Represents the/>Local density parameters of the individual feature vectors; /(I)Represents the/>Local degree of regularity of the individual feature vectors; /(I)Represents the/>The length of the texture curve in which the feature vectors are located; /(I)Represents the/>Variance of slope of texture curve where each feature vector is located; /(I)Represents the/>Discrete reliability parameters of the individual feature vectors;
the method for acquiring the discrete credibility parameter comprises the following steps:
And taking the ratio of the variance of the slope of the texture curve where each feature vector is located to the curve length as the discrete credibility parameter of each feature vector.
2. The method for monitoring and managing information of putting illegal objects according to claim 1, wherein the method for monitoring and managing the putting objects comprises the following steps:
And matching the main characteristic data with a database to obtain the matching degree of the released articles, and judging that the released articles have violations when the matching degree is lower than a preset matching threshold value, and sending out a violation release warning.
3. A system for monitoring and managing information of putting in illegal articles, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the steps of the method for monitoring and managing information of putting in illegal articles according to any one of claims 1-2 are realized when the processor executes the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410094823.9A CN117611109B (en) | 2024-01-24 | 2024-01-24 | Method and system for monitoring and managing illegal article delivery information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410094823.9A CN117611109B (en) | 2024-01-24 | 2024-01-24 | Method and system for monitoring and managing illegal article delivery information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117611109A CN117611109A (en) | 2024-02-27 |
| CN117611109B true CN117611109B (en) | 2024-04-26 |
Family
ID=89944670
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410094823.9A Active CN117611109B (en) | 2024-01-24 | 2024-01-24 | Method and system for monitoring and managing illegal article delivery information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117611109B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104780386A (en) * | 2015-03-19 | 2015-07-15 | 内蒙古民族大学 | Video-tampering detection method based on LOF-Co-Forest algorithm |
| WO2021038328A1 (en) * | 2019-08-27 | 2021-03-04 | Coupang Corp. | Computer-implemented method for detecting fraudulent transactions using locality sensitive hashing and locality outlier factor algorithms |
| CN112464983A (en) * | 2020-10-28 | 2021-03-09 | 吉林大学 | Small sample learning method for apple tree leaf disease image classification |
| CN113177484A (en) * | 2021-04-30 | 2021-07-27 | 洛阳师范学院 | Intelligent mechanical fault diagnosis method based on LOF self-encoding |
| CN113213017A (en) * | 2021-06-08 | 2021-08-06 | 苏州市伏泰信息科技股份有限公司 | Intelligent garbage classification propaganda and education and delivery supervision system based on visual AI technology |
| CN113762280A (en) * | 2021-04-23 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Image category identification method, device and medium |
| CN116977679A (en) * | 2023-08-09 | 2023-10-31 | 王子静 | Image acquisition method and system based on image recognition |
| CN117272210A (en) * | 2023-11-20 | 2023-12-22 | 湖南省瑞大工程建设有限公司 | Method and system for detecting abnormal hidden danger data of building construction |
| CN117421692A (en) * | 2023-12-18 | 2024-01-19 | 深圳火眼智能有限公司 | Garbage illegal delivery identification method, device and equipment for garbage delivery station |
-
2024
- 2024-01-24 CN CN202410094823.9A patent/CN117611109B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104780386A (en) * | 2015-03-19 | 2015-07-15 | 内蒙古民族大学 | Video-tampering detection method based on LOF-Co-Forest algorithm |
| WO2021038328A1 (en) * | 2019-08-27 | 2021-03-04 | Coupang Corp. | Computer-implemented method for detecting fraudulent transactions using locality sensitive hashing and locality outlier factor algorithms |
| CN112464983A (en) * | 2020-10-28 | 2021-03-09 | 吉林大学 | Small sample learning method for apple tree leaf disease image classification |
| CN113762280A (en) * | 2021-04-23 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Image category identification method, device and medium |
| CN113177484A (en) * | 2021-04-30 | 2021-07-27 | 洛阳师范学院 | Intelligent mechanical fault diagnosis method based on LOF self-encoding |
| CN113213017A (en) * | 2021-06-08 | 2021-08-06 | 苏州市伏泰信息科技股份有限公司 | Intelligent garbage classification propaganda and education and delivery supervision system based on visual AI technology |
| CN116977679A (en) * | 2023-08-09 | 2023-10-31 | 王子静 | Image acquisition method and system based on image recognition |
| CN117272210A (en) * | 2023-11-20 | 2023-12-22 | 湖南省瑞大工程建设有限公司 | Method and system for detecting abnormal hidden danger data of building construction |
| CN117421692A (en) * | 2023-12-18 | 2024-01-19 | 深圳火眼智能有限公司 | Garbage illegal delivery identification method, device and equipment for garbage delivery station |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117611109A (en) | 2024-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11288313B2 (en) | Method and apparatus for multi-dimensional content search and video identification | |
| CN110907762B (en) | Non-invasive load matching identification method | |
| CA2717612C (en) | Fingerprint representation using gradient histograms | |
| CN110750524A (en) | Method and system for determining fault characteristics of active power distribution network | |
| WO2014054042A1 (en) | Device and method for detecting plantation rows | |
| CN110826618A (en) | Personal credit risk assessment method based on random forest | |
| CN112926592B (en) | Trademark retrieval method and device based on improved Fast algorithm | |
| CN110084169A (en) | A kind of architecture against regulations object recognition methods based on K-Means cluster and profile topological constraints | |
| CN110737788B (en) | Rapid three-dimensional model index establishing and retrieving method | |
| Zheng et al. | The Credit Card Anti-Fraud Detection Model in the Context of Dynamic Integration Selection Algorithm | |
| CN110458094B (en) | Equipment classification method based on fingerprint similarity | |
| CN111709668A (en) | Power grid equipment parameter risk identification method and device based on data mining technology | |
| CN117611109B (en) | Method and system for monitoring and managing illegal article delivery information | |
| CN110334508A (en) | A kind of host sequence intrusion detection method | |
| CN111428064B (en) | Small-area fingerprint image fast indexing method, device, equipment and storage medium | |
| Gao et al. | An improved XGBoost based on weighted column subsampling for object classification | |
| CN106326927A (en) | Shoeprint new class detection method | |
| Zhang et al. | Saliency detection via image sparse representation and color features combination | |
| CN116821087A (en) | Power transmission line fault database construction method, device, terminal and storage medium | |
| CN106611418A (en) | Image segmentation algorithm | |
| CN108415958A (en) | The weight processing method and processing device of index weight VLAD features | |
| CN113127464B (en) | Agricultural big data environment feature processing method and device and electronic equipment | |
| CN109739840A (en) | Data processing empty value method, apparatus and terminal device | |
| CN115393631A (en) | Hyperspectral image classification method based on Bayesian layer graph convolution neural network | |
| Kumar et al. | Automatic feature weight determination using indexing and pseudo-relevance feedback for multi-feature content-based image retrieval |
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 |