CN108563675B - Electronic file automatic generation method and device based on target body characteristics - Google Patents
Electronic file automatic generation method and device based on target body characteristics Download PDFInfo
- Publication number
- CN108563675B CN108563675B CN201810166463.3A CN201810166463A CN108563675B CN 108563675 B CN108563675 B CN 108563675B CN 201810166463 A CN201810166463 A CN 201810166463A CN 108563675 B CN108563675 B CN 108563675B
- Authority
- CN
- China
- Prior art keywords
- target body
- target
- archive
- file
- features
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of data processing, in particular provides a method and a device for automatically generating an electronic file based on target body characteristics, and aims to solve the technical problem of how to quickly and accurately acquire file information of target personnel. For this purpose, the method for automatically generating the electronic file can store the file information of the targets belonging to the same file category into the file area corresponding to the target of the current category in the pre-constructed file library, so that the file information stored in each file area is the file information of the targets of the same file category, thereby realizing 'one-class one-file', and realizing 'one-person one-file' when the targets are human. Based on the dynamic portrait base generated by the automatic electronic file generation method, all file information of the target person can be obtained by searching the attribute information of the target person in the dynamic portrait base after the target person is determined without adopting a one-by-one searching mode. Meanwhile, the device in the invention can execute and realize the method.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an electronic file automatic generation method and device based on target body characteristics.
Background
The dynamic portrait library refers to an information database containing portrait information and accessory information thereof, wherein the accessory information comprises information such as hair style, clothes color and the like. The information data in the current dynamic portrait base is mainly stored according to the collection time and the collection place in a classified manner, so that the data information in the dynamic portrait base can be searched according to the following steps: firstly, a search time range and a search space range are determined according to a search target, then corresponding data information in a dynamic portrait base is searched one by one according to the determined time and space range, however, the data volume of the data information is often large, the search time is inevitably increased by adopting the one-by-one search mode, and meanwhile, the search accuracy is also reduced.
Disclosure of Invention
In order to solve the above problems in the prior art, that is, to solve the technical problem of how to quickly and accurately acquire the file information of the target person, the invention provides a method and a device for automatically generating an electronic file based on the characteristics of a target body.
In a first aspect, the method for automatically generating an electronic archive based on target characteristics in the present invention includes:
acquiring a target image by adopting a cross-camera target tracking algorithm;
identifying the obtained target body image to obtain target body characteristics, and performing archive classification on the target body according to the target body characteristics;
and storing the file information of the targets belonging to the same file type into a file area corresponding to the target of the current type in a pre-constructed file library.
Further, a preferred technical solution provided by the present invention is:
the target volume features include facial features; the step of performing archive classification on the target according to the target characteristics specifically comprises the following steps:
according to the features of the five sense organs, file information stored in the pre-constructed file library is subjected to traversal search, and first candidate file information is obtained;
screening the first candidate file information according to a preset target screening condition to obtain second candidate file information;
acquiring a facial organ model of a candidate target body corresponding to the second candidate file information;
matching the features of the five sense organs with the features of the five sense organs model, and performing weighted calculation on matching results to obtain a first similarity;
and comparing the first similarity with a preset threshold value, and determining the file type of the target body according to the comparison result.
Further, a preferred technical solution provided by the present invention is:
the target volume features further comprise hair styling/hair accessory features, apparel features, and/or vehicle features;
the step of comparing the first similarity with a preset threshold and determining the archival category of the target according to the comparison result specifically comprises:
if fSimi≤Thr_lowThen constructing a new file type as the file type of the target, wherein fSimiFor the first similarity, the preset threshold includes a low threshold Thr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_high；
If fSimi≥Thr_highThen the current first similarity fSimiTaking the file type of the corresponding candidate target body as the file type of the target body;
if Thr_low＜fSimi＜Thr_highThen, the following operations are performed:
constructing a first feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the target body, constructing a second feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the candidate target body, and performing correlation calculation on the first feature vector and the second feature vector to obtain a second similarity;
performing weighted calculation on the first similarity and the second similarity to obtain the overall similarity;
and taking the file type of the candidate target body corresponding to the full-face similarity with the maximum similarity value as the file type of the target body.
Further, a preferred technical solution provided by the present invention is:
the step of performing traversal search on the archive information stored in the pre-constructed archive library according to the features of the five sense organs to obtain first candidate archive information specifically comprises the following steps:
identifying whether the head of the target body in the target body image wears ornaments or not, and adjusting the weight of the characteristics of the five sense organs according to the identification result;
and matching the file information stored in the pre-constructed file library according to the characteristics of the five sense organs after the weight is adjusted to obtain first candidate file information.
Further, a preferred technical solution provided by the present invention is:
the method further comprises the following steps:
associating the pre-constructed archive library with a preset target body verification library;
acquiring verification data of the target body in the preset target body verification library;
and matching the acquired verification data with the file information in the pre-constructed file library, and determining the correct category of the target body according to the matching result.
Further, a preferred technical solution provided by the present invention is:
the method also comprises a step of acquiring the target body image and an image recognition training sample thereof, a step of analyzing the behavior characteristics of the target body, a step of monitoring a specific target body and a step of pre-warning/alarming the specific target body;
the acquisition steps of the target body image and the image recognition training sample thereof comprise: screening archive information in the pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type;
the step of analyzing the target body behavior characteristics comprises the following steps: counting the archive information of the target body, generating a behavior track of the target body according to a counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in a current target body image according to the counting result, and further acquiring the behavior characteristics of the other target bodies according to the acquired archive information;
the monitoring of the specific target includes: monitoring the target body behavior characteristics of a specific target body in real time based on the file information in the pre-constructed file library;
the pre/alert step for the particular target volume includes: comparing a plurality of target body characteristics of a target body with a plurality of characteristics of a specific target body respectively, and carrying out early warning according to comparison results; and/or acquiring and analyzing the behavior track of a specific target body, so that a plurality of specific target bodies give an alarm when the specific target bodies appear in the same area in the same time period.
Further, a preferred technical solution provided by the present invention is:
the archival information of the target body comprises identity information, gender, age, race, appearance characteristics, an identity card image, a typical representative image and a snapshot image set of the target body; the snapshot image set comprises snapshot images of the target body acquired through the camera device at different time and areas.
In a second aspect, an apparatus for automatically generating an electronic archive based on a target volume feature according to the present invention includes:
the target body image acquisition module is configured to acquire a target body image by adopting a cross-camera target tracking algorithm;
the target body archive classifying module is configured to identify the acquired target body image to obtain target body characteristics, and perform archive classification on the target body according to the target body characteristics;
and the target body archive information storage module is configured to store the archive information of the target bodies belonging to the same archive type into an archive area corresponding to the target body of the current type in a pre-constructed archive library.
Further, a preferred technical solution provided by the present invention is:
the target volume features include facial features; the target profile classification module comprises:
the first candidate archive information acquisition sub-module is configured to perform traversal search on the archive information stored in the pre-constructed archive library according to the features of the five sense organs to obtain first candidate archive information;
the second candidate archive information acquisition sub-module is configured to screen the first candidate archive information according to preset target screening conditions to obtain second candidate archive information;
the facial features matching sub-module is configured to acquire a facial features model of a candidate target body corresponding to the second candidate archive information, perform feature matching on the facial features and the facial features model, and perform weighted calculation on matching results to obtain a first similarity;
and the archive category determination submodule is configured to compare the first similarity with a preset threshold value and determine the archive category of the target body according to a comparison result.
Further, a preferred technical solution provided by the present invention is:
the target volume features further comprise hair styling/hair accessory features, apparel features, and/or vehicle features; the profile category determination submodule includes:
a first file type determination unit configured to determine a first file type at fSimi≤Thr_lowIn case of (a), constructing a new archive class as the archive class of the object, wherein fSimiFor the first similarity, the preset threshold includes a low threshold Thr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_high；
A second gear class determination unit configured to determine a second gear class at fSimi≥Thr_highIn case of (2), the current first similarity f is setSimiTaking the file type of the corresponding candidate target body as the file type of the target body;
a third gear class determination unit configured to determine a third gear class at Thr_low＜fSimi＜Thr_highIn the case of (2), the following operations are performed:
constructing a first feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the target body, constructing a second feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the candidate target body, and performing correlation calculation on the first feature vector and the second feature vector to obtain a second similarity;
performing weighted calculation on the first similarity and the second similarity to obtain the overall similarity;
and taking the file type of the candidate target body corresponding to the full-face similarity with the maximum similarity value as the file type of the target body.
Further, a preferred technical solution provided by the present invention is:
the first candidate archive information acquisition sub-module includes:
the weight adjusting unit of the features of the five sense organs is configured to identify whether the head of the target body in the target body image wears the ornament or not and adjust the weight of the features of the five sense organs according to the identification result;
and the facial features matching unit is configured to match the stored archive information in the pre-constructed archive library according to the weight-adjusted facial features to obtain first candidate archive information.
Further, a preferred technical solution provided by the present invention is:
the device also comprises a file category verification module; the archive category verification module comprises:
the archive association submodule is configured to associate the pre-established archive with a preset target verification library;
the verification data acquisition sub-module is configured to acquire verification data of a target body in the preset target body verification library;
and the file type verification sub-module is configured to match the verification data acquired by the verification data acquisition sub-module with file information in the pre-constructed file library and determine the correct type of the target body according to a matching result.
Further, a preferred technical solution provided by the present invention is:
the device also comprises a target body image and image recognition training sample acquisition module thereof, a target body behavior characteristic analysis module, a specific target body monitoring module and a specific target body pre/alarm module;
the target body image and image recognition training sample acquisition module thereof is configured to: screening archive information in the pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type;
the target body behavior feature analysis module is configured to: counting the archive information of the target body, generating a behavior track of the target body according to a counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in a current target body image according to the counting result, and further acquiring the behavior characteristics of the other target bodies according to the acquired archive information;
the specific target monitoring module is configured to: monitoring the target body behavior characteristics of a specific target body in real time based on the file information in the pre-constructed file library;
the specific target body pre/alarm module is configured to: comparing a plurality of target body characteristics of a target body with a plurality of characteristics of a specific target body respectively, and carrying out early warning according to comparison results; and/or acquiring and analyzing the behavior track of a specific target body, so that a plurality of specific target bodies give an alarm when the specific target bodies appear in the same area in the same time period.
In a third aspect, the storage device in the present invention stores a plurality of programs, and the programs are suitable for being loaded and executed by a processor to realize the method for automatically generating the electronic file based on the target body characteristics according to the above technical scheme.
In a fourth aspect, a processing apparatus in the present invention comprises:
a processor adapted to execute various programs;
a storage device adapted to store a plurality of programs;
the program is suitable for being loaded and executed by a processor to realize the automatic generation method of the electronic file based on the target body characteristics.
Compared with the closest prior art, the technical scheme at least has the following beneficial effects:
1. the electronic file automatic generation method based on the object body characteristics can store the file information of the object bodies belonging to the same file type into the file area corresponding to the object body of the current type in the pre-constructed file library, so that the file information stored in each file area is the file information of the object body of the same file type, one-class-one-file is realized, and one-person-one-file can be realized when the object body is a person. Meanwhile, based on the dynamic portrait base generated by the electronic archive generation method, all archive information of the target person can be obtained by searching attribute information of the target person, such as the number of the target person, in the dynamic portrait base after the target person is determined without adopting a one-by-one searching mode.
2. The electronic file automatic generation method based on the target body characteristics can also correlate the pre-constructed file library with a preset target body verification library, such as a target person file information library published by a public security organization, and verify the file information of the target body in the file library based on verification data in the target body verification library, so as to ensure that the file information of the target body is correctly filed.
3. The electronic file automatic generation method based on the target body characteristics can also perform statistical analysis on the file information of the target body to obtain the behavior characteristics of the target body, and further perform deep analysis on the target body according to the behavior characteristics. For example, based on the behavior trajectory of the target, the activity time and the activity area of the target in a specific area can be analyzed. The number of times that the target body appears in a specific place, such as a station and an internet cafe, can be obtained based on the number of times that the target body is captured by each camera device. Based on the archival categories of other objects in the object image, it can be found which objects still appear around the object. Based on the file category of the new target, the file repository operation and maintenance personnel can quickly search the new target.
Drawings
FIG. 1 is a schematic diagram of the main steps of a method for automatically generating an electronic file based on object features according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an apparatus for automatically generating an electronic file based on object features according to an embodiment of the present invention.
Detailed Description
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and are not intended to limit the scope of the present invention.
Referring to fig. 1, fig. 1 schematically illustrates the main steps of an automatic electronic archive generation method based on target characteristics in this embodiment. As shown in fig. 1, the method for automatically generating an electronic file based on target characteristics in this embodiment includes the following steps:
step S101: and acquiring a target image by adopting a cross-camera target tracking algorithm.
In the embodiment, the cross-camera target tracking algorithm is adopted to obtain the target body image, so that the image or video information of the target body with reasonable space range and consistent time sequence can be obtained.
Step S102: and identifying the acquired target body image to obtain target body characteristics, and classifying the target body according to the target body characteristics.
In particular, the target body features in this embodiment may include the five-sense-organ features, hair style/hair accessory features, apparel features, and/or vehicle features of the target body. The vehicle characteristics refer to characteristics such as the type and color of the vehicle used by the target.
In this embodiment, the target volume may be subjected to archive classification according to the following steps:
step S1021: according to the characteristics of the five sense organs, the file information stored in a pre-constructed file library is subjected to traversal search, and first candidate file information is obtained.
The step of acquiring the first candidate archive information in this embodiment specifically includes: firstly, whether the head of the target body in the target body image is worn with ornaments, such as glasses, a mask, a hat and the like is identified, and the weight of the characteristics of the five sense organs is adjusted according to the identification result. Then, according to the features of the five sense organs after the weight adjustment, the stored archive information in the pre-constructed archive library is matched to obtain first candidate archive information, that is, in this embodiment, the first candidate archive information is obtained by performing weighted matching on the features of the five sense organs.
Step S1022: and screening the first candidate file information according to a preset target screening condition to obtain second candidate file information.
The preset target screening condition in this embodiment may include that the gender of the target is a preset gender type, and/or the age of the target is in a preset age range, and/or the race of the target is a preset race type.
Step S1023: and acquiring a facial feature model of the candidate target body corresponding to the second candidate file information.
Step S1024: and performing feature matching on the features of the five sense organs and the five sense organ model, and performing weighted calculation on a matching result to obtain a first similarity.
Specifically, the method comprises the steps of firstly segmenting a five sense organ model, extracting the characteristics of each five sense organ in the five sense organ model, then matching the characteristics of the five sense organs of a target body with the extracted characteristics of the five sense organs, and finally performing weighted calculation on a matching result to obtain a first similarity.
Step S1025: and comparing the first similarity with a preset threshold value, and determining the file type of the target body according to the comparison result.
The preset threshold includes a low threshold T in this embodimenthr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_highBased on the first similarity fSimiThe comparison result with the preset threshold includes three results: f. ofSimi≤Thr_low，fSimi≥Thr_high，Thr_low＜fSimi＜Thr_high。
When f isSimi≤Thr_lowIf the target object is a file type which is not stored in the file library, that is, the file type to which the target object belongs is a new type, a new file type can be constructed as the file type of the target object.
When f isSimi＞Thr_highThe time interval indicates that the file type of the target is the same as that of the candidate target corresponding to the current first similarity, so that the file type of the candidate target corresponding to the current first similarity can be used as the file type of the target.
When T ishr_low＜fSimi＜Thr_highThe time-of-day indicates that the file category of the target body cannot be accurately identified under the condition of only being based on the characteristics of the five sense organs, and the purpose can be continuously judged on the basis of the hair style/hair accessory characteristics, the clothing characteristics and/or the vehicle characteristics of the target bodyThe file category of the mark body specifically comprises the following steps:
firstly, a first feature vector is constructed according to hair style/hair accessory features, clothing features and/or vehicle features of a target body, a second feature vector is constructed according to hair style/hair accessory features, clothing features and/or vehicle features of a candidate target body, and the first feature vector and the second feature vector are subjected to correlation calculation to obtain a second similarity.
Secondly, for the first similarity fSimiAnd a second degree of similarity cSimiAnd performing weighted calculation to obtain the similarity of the whole appearance. The global similarity w in this embodimentSimiAs shown in the following formula (1):
wSimi＝α×fSimi+β×cSimi (1)
the meaning of each parameter in the formula (1) is as follows: alpha is a first degree of similarity fSimiBeta is the second degree of similarity cSimiAnd α + β is 1. In a preferred embodiment of this example, α is 0.6 and β is 0.4.
And finally, taking the file type of the candidate target body corresponding to the global similarity with the maximum similarity value as the file type of the target body.
Step S103: and storing the file information of the targets belonging to the same file type into a file area corresponding to the target of the current type in a pre-constructed file library.
The profile information of the target body in this embodiment includes identity information, gender, age, race, appearance characteristics, identification card image, typical representative image, and snapshot image set of the target body. The snapshot image set comprises snapshot images of the target body acquired through the camera device at different time and areas.
In this embodiment, based on step S103, all the archive information of the objects of the same archive type are stored in a specific archive region, so that the archive information stored in each archive region is the archive information of the objects of the same archive type, thereby implementing "one class of first file", and implementing "one person one file" when the objects are human. The dynamic portrait base generated by the electronic archive automatic generation method based on the target body characteristics does not need to adopt a one-by-one searching mode, and after the target person is determined, the attribute information of the target person, such as the number of the target person, is searched in the dynamic portrait base, so that all archive information of the target person can be obtained.
Further, in this embodiment, the method for automatically generating an electronic file based on target characteristics shown in fig. 1 may further check the file information in the archive based on a preset target check library to ensure that the target file is correctly classified, specifically including:
step S201: and associating the pre-constructed archive library with a preset target body verification library. In this embodiment, when the target is a person, the preset target verification library may be a target person archive information library published by a public security organization.
Step S202: and acquiring verification data of the target body in a preset target body verification library.
Step S203: and matching the acquired verification data with file information in a pre-constructed file library, and determining the correct category of the target body according to the matching result.
Specifically, when the verification data matches the profile information, i.e., the data information of the target belonging to the same profile type, the profile type of the current target is not modified. When the verification data is not matched with the file information, namely the data information of the target body which does not belong to the same file type, the file type of the target body is modified into the file type corresponding to the verification data.
For example, if the target is a person, if the verification data is the file information of the target person a, and the file information currently matching and determining with the verification data is also the file information of the target person a, the file type of the current target person does not need to be modified. When the target is a person, if the verification data is the file information of the target person A, and the current file information matched and judged with the verification data is also the file information of the target person B, the file type of the current target person is modified into the file type of the target person A, and the file information of the current target person is stored in a file area corresponding to the target person A in the file library.
Further, in this embodiment, the method for automatically generating an electronic file based on target characteristics shown in fig. 1 further includes a step of acquiring a target image and an image recognition training sample thereof, a step of analyzing target behavior characteristics, a step of monitoring a specific target, and a step of pre-warning/warning of the specific target.
Specifically, the step of acquiring the target volume image and the image recognition training sample thereof in this embodiment includes: and screening archive information in a pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type. The target body image and the training sample image for image recognition can be automatically collected based on the electronic file.
In this embodiment, the step of analyzing the behavior characteristics of the target includes: and counting the archive information of the target body, generating a behavior track of the target body according to the counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in the current target body image according to the counting result, and acquiring the behavior characteristics of other target bodies according to the acquired archive information. Meanwhile, after the file type of the target is determined to be a new file type, the behavior characteristics can also comprise the file type of the new target. Based on the behavior trace of the target body, the activity time and the activity area of the target body in a specific area can be analyzed. The number of times that the target body appears in a specific place, such as a station and an internet cafe, can be obtained based on the number of times that the target body is captured by each camera device. Based on the archival categories of other objects in the object image, it can be found which objects still appear around the object. Based on the file category of the new target, the file repository operation and maintenance personnel can quickly search the new target. For example, when the target is a person, the fellow passenger of the target person can be obtained based on the profile type of other target in the target image. When the target is a person, the operation and maintenance personnel can quickly search for strangers in a specific area based on the profile category of the new target.
The step of monitoring the specific target in this embodiment includes: and monitoring the target body behavior characteristics of the specific target body in real time based on the pre-constructed archive information in the archive library. Namely, the real-time monitoring of specific targets, such as dangerous persons, can be realized based on the electronic file.
The pre/alarm step for a specific target in this embodiment includes: the characteristics of a plurality of target bodies of one target body are respectively compared with the characteristics of a specific target body, and early warning is carried out according to the comparison result, namely whether the target body is a specific target body which is mainly monitored or not can be judged according to the comparison result of the characteristics, so that early warning is carried out according to the comparison result. Further, the embodiment may further obtain and analyze the behavior trace of the specific target, so that a plurality of specific targets may give an alarm when appearing in the same area in the same time period, that is, when the specific targets labeled in advance as having the same attribute have a group rendezvous activity, the alarm may be given in time.
Although the foregoing embodiments describe the steps in the above sequential order, those skilled in the art will understand that, in order to achieve the effect of the present embodiments, the steps may not be executed in such an order, and may be executed simultaneously (in parallel) or in an inverse order, and these simple variations are within the scope of the present invention.
Based on the same technical concept as the method embodiment, the embodiment of the invention also provides an automatic electronic file generation device based on the target body characteristics. The automatic electronic file generation device based on the target body features will be specifically described below with reference to the accompanying drawings.
Referring to fig. 2, fig. 2 schematically shows the main structure of an automatic electronic file generation device based on object features in this embodiment. As shown in fig. 2, the automatic generation apparatus of an electronic archive based on the target volume characteristics in the present embodiment may include a target volume image acquisition module 11, a target volume archive classification module 12, and a target volume archive information storage module 13. Specifically, in this embodiment, the target volume image obtaining module 11 may be configured to obtain the target volume image by using a cross-camera target tracking algorithm. The target volume archive classifying module 12 may be configured to identify the target volume image acquired by the target volume image acquiring module 11, obtain target volume characteristics, and perform archive classification on the target volume according to the target volume characteristics. The object archive information storage module 13 may be configured to store the archive information of the objects belonging to the same archive category into the archive region corresponding to the object of the current category in the pre-constructed archive.
Further, in this embodiment, the target volume characteristics may include facial features, and the target volume archive classification module 12 shown in fig. 2 may include a first candidate archive information acquisition sub-module, a second candidate archive information acquisition sub-module, a facial feature matching sub-module, and an archive category determination sub-module. Specifically, in this embodiment, the first candidate archive information acquisition sub-module may be configured to perform traversal search on archive information stored in a pre-constructed archive library according to features of five sense organs, so as to obtain the first candidate archive information. The second candidate archive information acquisition sub-module may be configured to filter the first candidate archive information according to a preset target screening condition to obtain second candidate archive information. The feature matching sub-module of the five sense organs can be configured to obtain a model of the five sense organs of the candidate target body corresponding to the second candidate archive information, perform feature matching on the features of the five sense organs and the model of the five sense organs, and perform weighted calculation on matching results to obtain a first similarity. The profile type determination sub-module may be configured to compare the first similarity with a preset threshold and determine a profile type of the target body according to a comparison result.
Further, in this embodiment, the object characteristics may further include hair style/hair accessory characteristics, apparel characteristics, and/or vehicle characteristics, and the profile category determination sub-module may include a first profile category determination unit, a second profile category determination unit, and a third profile category determination unit. Specifically, the first profile category determination unit in the present embodiment may be configured to determine the first profile category at fSimi≤Thr_lowIn case of (2), a new archive class is constructed as the archive class of the target, wherein fSimiFor the first similarity, the preset threshold includes a low threshold Thr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_high. Second document type determination orderThe element may be configured at fSimi≥Thr_highIn case of (2), the current first similarity f is setSimiThe file type of the corresponding candidate target body is used as the file type of the target body. The third gear class determination unit may be configured to determine the third gear class at Thr_low＜fSimi＜Thr_highIn the case of (2), the following operations are performed: firstly, a first feature vector is constructed according to hair style/hair accessory features, clothing features and/or vehicle features of a target body, a second feature vector is constructed according to hair style/hair accessory features, clothing features and/or vehicle features of a candidate target body, and the first feature vector and the second feature vector are subjected to correlation calculation to obtain a second similarity. And secondly, performing weighted calculation on the first similarity and the second similarity to obtain the global similarity. And finally, taking the file type of the candidate target body corresponding to the global similarity with the maximum similarity value as the file type of the target body.
Further, in this embodiment, the first candidate archive information acquisition sub-module may include a facial feature weight adjustment unit and a facial feature matching unit. Specifically, the weight adjusting unit for the features of five sense organs in this embodiment may be configured to identify whether the head of the target body in the target body image wears the accessory, and adjust the weight of the features of five sense organs according to the identification result. The facial features matching unit may be configured to match stored archive information in a pre-constructed archive library according to the weighted facial features to obtain first candidate archive information.
Further, in this embodiment, the device for automatically generating an electronic file based on the target feature shown in fig. 2 may further include a file category verification module, where the file category verification module may include an archive association sub-module, a verification data acquisition sub-module, and a file category verification sub-module. Specifically, in this embodiment, the archive association sub-module may be configured to associate the pre-constructed archive with the preset target verification library. The verification data acquisition sub-module can be configured to acquire verification data of a target in a preset target verification library. The archive category verification sub-module can be configured to match the verification data acquired by the verification data acquisition sub-module with archive information in a pre-constructed archive library, and determine the correct category to which the target body belongs according to the matching result.
Further, the device for automatically generating an electronic file based on target characteristics shown in fig. 2 in this embodiment may further include a target image and image recognition training sample obtaining module, a target behavior characteristic analysis module, a specific target monitoring module, and a specific target pre-warning/warning module. .
Specifically, in this embodiment, the target volume image and the image recognition training sample acquiring module thereof may be configured to: and screening archive information in a pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type.
In this embodiment, the target behavior feature analysis module may be configured to: and counting the archive information of the target body, generating a behavior track of the target body according to the counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in the current target body image according to the counting result, and acquiring the behavior characteristics of other target bodies according to the acquired archive information.
The specific target monitoring module in this embodiment may be configured to: and monitoring the target body behavior characteristics of the specific target body in real time based on the pre-constructed archive information in the archive library.
The specific target body pre/alarm module in this embodiment may be configured to: comparing a plurality of target body characteristics of a target body with a plurality of characteristics of a specific target body respectively, and carrying out early warning according to comparison results; and/or acquiring and analyzing the behavior track of a specific target body, so that a plurality of specific target bodies give an alarm when the specific target bodies appear in the same area in the same time period.
Those skilled in the art will appreciate that the above-described automatic generation device of an electronic archive based on object volume characteristics also includes some other well-known structures, such as processors, controllers, memories, etc., wherein the memories include, but are not limited to, random access memory, flash memory, read only memory, programmable read only memory, volatile memory, non-volatile memory, serial memory, parallel memory, registers, etc., and the processors include, but are not limited to, CPLD/FPGA, DSP, ARM processor, MIPS processor, etc., and these well-known structures are not shown in fig. 2 in order to unnecessarily obscure embodiments of the present disclosure.
It should be understood that the number of individual modules in fig. 2 is merely illustrative. The number of modules may be any according to actual needs.
Based on the above embodiment of the method for automatically generating an electronic archive based on target volume characteristics, an embodiment of the present invention further provides a storage device, where multiple programs are stored in the storage device, and the programs are suitable for being loaded and executed by a processor to implement the method for automatically generating an electronic archive based on target volume characteristics described in the above embodiment of the method.
Further, based on the above-mentioned embodiment of the method for automatically generating an electronic archive based on target volume characteristics, an embodiment of the present invention further provides a processing apparatus, where the processing apparatus includes a processor and a storage device, where the processor may be adapted to execute various programs, the storage device may be adapted to store a plurality of programs, and the programs may be adapted to be loaded and executed by the processor to implement the method for automatically generating an electronic archive based on target volume characteristics described in the above-mentioned embodiment of the method.
Those skilled in the art will appreciate that the modules in the devices of the embodiments can be adaptively changed and placed in one or more devices other than the embodiments. The modules or units in the embodiments may be combined into one module or unit, and furthermore, they may be divided into a plurality of sub-modules or sub-units. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the claims of the present invention, any of the claimed embodiments may be used in any combination.
The various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functionality of some or all of the components in a server, client, or the like, according to embodiments of the present invention. The present invention may also be embodied as an apparatus or device program (e.g., PC program and PC program product) for carrying out a portion or all of the methods described herein. Such a program implementing the invention may be stored on a PC readable medium or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed PC. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.
Claims (11)
1. An automatic generation method of electronic files based on target body features is characterized by comprising the following steps S1-S3:
step S1: acquiring a target image by adopting a cross-camera target tracking algorithm;
step S2: identifying the obtained target body image to obtain target body characteristics, and performing archive classification on the target body according to the target body characteristics, wherein the target body characteristics comprise facial features, behavior characteristics, hair style/hair ornament characteristics, clothing characteristics and/or vehicle characteristics; the behavior feature is determined according to a behavior track of the target body;
the archive classification of the target body according to the target body characteristics specifically comprises the following steps: traversing and searching the file information stored in a pre-constructed file library according to the facial features in the target body feature to obtain first candidate file information; screening the first candidate file information according to a preset target screening condition to obtain second candidate file information; acquiring a facial features model of a candidate target body corresponding to the second candidate file information, and segmenting the facial features model to extract facial features of each facial feature in the facial features model; performing feature matching on the features of the five sense organs in the target body features and the extracted features of the five sense organs, and performing weighted calculation on matching results to obtain a first similarity; comparing the first similarity with a preset threshold value, and determining the first similarity according to the comparison resultThe file type of the object includes fSimi≤Thr_lowThen constructing a new file type as the file type of the target, wherein fSimiFor the first similarity, the preset threshold includes a low threshold Thr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_high(ii) a If fSimi≥Thr_highThen the current first similarity fSimiTaking the file type of the corresponding candidate target body as the file type of the target body; if Thr_low＜fSimi＜Thr_highThen, the following operations are performed: constructing a first feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the target body, constructing a second feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the candidate target body, and performing correlation calculation on the first feature vector and the second feature vector to obtain a second similarity; performing weighted calculation on the first similarity and the second similarity to obtain the overall similarity; taking the file type to which the candidate target body corresponding to the full-face similarity with the maximum similarity value belongs as the file type of the target body;
step S3: and storing the file information of the targets belonging to the same file type into a file area corresponding to the target of the current type in a pre-constructed file library.
2. The method for automatically generating an electronic archive based on target features as claimed in claim 1, wherein the step of performing a traversal search on the archive information stored in the pre-constructed archive repository according to the features of the five sense organs in the target features to obtain the first candidate archive information specifically comprises:
identifying whether the head of the target body in the target body image wears ornaments or not, and adjusting the weight of the characteristics of the five sense organs according to the identification result;
and matching the file information stored in the pre-constructed file library according to the characteristics of the five sense organs after the weight is adjusted to obtain first candidate file information.
3. The method for automatically generating an electronic archive based on target volume characteristics according to claim 1 or 2, characterized in that the method further comprises:
associating the pre-constructed archive library with a preset target body verification library;
acquiring verification data of the target body in the preset target body verification library;
and matching the acquired verification data with the file information in the pre-constructed file library, and determining the correct category of the target body according to the matching result.
4. The method for automatically generating the electronic file based on the target body characteristics as claimed in claim 1 or 2, wherein the method further comprises a step of acquiring a target body image and an image recognition training sample thereof, a step of analyzing the target body behavior characteristics, a step of monitoring a specific target body, and a step of pre/warning of the specific target body;
the acquisition steps of the target body image and the image recognition training sample thereof comprise: screening archive information in the pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type;
the step of analyzing the target body behavior characteristics comprises the following steps: counting the archive information of the target body, generating a behavior track of the target body according to a counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in a current target body image according to the counting result, and further acquiring the behavior characteristics of the other target bodies according to the acquired archive information;
the monitoring of the specific target includes: monitoring the target body behavior characteristics of a specific target body in real time based on the file information in the pre-constructed file library;
the pre/alert step for the particular target volume includes: respectively comparing a plurality of target body characteristics of a specific target body with a plurality of preset characteristics, and carrying out early warning according to comparison results; and/or acquiring and analyzing the behavior track of a specific target body, so that a plurality of specific target bodies give an alarm when the specific target bodies appear in the same area in the same time period.
5. The method for automatically generating an electronic file based on the target volume characteristics as claimed in claim 1 or 2,
the archival information of the target body comprises identity information, gender, age, race, appearance characteristics, an identity card image, a typical representative image and a snapshot image set of the target body; the snapshot image set comprises snapshot images of the target body acquired through the camera device at different time and areas.
6. An apparatus for automatically generating an electronic file based on target characteristics, the apparatus comprising:
the target body image acquisition module is configured to acquire a target body image by adopting a cross-camera target tracking algorithm;
the target body archive classifying module is configured to identify the acquired target body image to obtain target body characteristics, and perform archive classification on the target body according to the target body characteristics, wherein the target body characteristics comprise five sense organ characteristics, behavior characteristics, hair style/hair accessory characteristics, clothing characteristics and/or vehicle characteristics; the behavior feature is determined according to a behavior track of the target body;
the target body archive classification module comprises a first candidate archive information acquisition sub-module, a second candidate archive information acquisition sub-module, a facial features matching sub-module and an archive category determination sub-module; the first candidate archive information acquisition submodule is configured to perform traversal search on archive information stored in a pre-constructed archive library according to the facial features in the target body feature to obtain first candidate archive information; the second candidate archive information acquisition sub-module is configured to screen the first candidate archive information according to a preset target screening condition to obtain second candidate archive information; the features of the five sense organs are matchedThe submodule is configured to acquire a facial features model of a candidate target corresponding to the second candidate archive information, divide the facial features model to extract facial features of each facial feature in the facial features model, perform feature matching on the facial features in the target feature and the extracted facial features, and perform weighted calculation on matching results to obtain a first similarity; the archive category determination submodule is configured to compare the first similarity with a preset threshold value and determine the archive category of the target body according to a comparison result; the profile category determination submodule includes: a first file type determination unit configured to determine a first file type at fSimi≤Thr_lowIn case of (a), constructing a new archive class as the archive class of the object, wherein fSimiFor the first similarity, the preset threshold includes a low threshold Thr_lowAnd a high threshold value Thr_highAnd T ishr_low＜Thr_high(ii) a A second gear class determination unit configured to determine a second gear class at fSimi≥Thr_highIn case of (2), the current first similarity f is setSimiTaking the file type of the corresponding candidate target body as the file type of the target body; a third gear class determination unit configured to determine a third gear class at Thr_low＜fSimi＜Thr_highIn the case of (2), the following operations are performed: constructing a first feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the target body, constructing a second feature vector according to the hair style/hair accessory features, the clothing features and/or the vehicle features of the candidate target body, and performing correlation calculation on the first feature vector and the second feature vector to obtain a second similarity; performing weighted calculation on the first similarity and the second similarity to obtain the overall similarity; taking the file type to which the candidate target body corresponding to the full-face similarity with the maximum similarity value belongs as the file type of the target body;
and the target body archive information storage module is configured to store the archive information of the target bodies belonging to the same archive type into an archive area corresponding to the target body of the current type in a pre-constructed archive library.
7. The apparatus according to claim 6, wherein the first candidate archive information acquisition sub-module comprises:
the weight adjusting unit of the features of the five sense organs is configured to identify whether the head of the target body in the target body image wears the ornament or not and adjust the weight of the features of the five sense organs according to the identification result;
and the facial features matching unit is configured to match the stored archive information in the pre-constructed archive library according to the weight-adjusted facial features to obtain first candidate archive information.
8. The apparatus for automatically generating an electronic file based on the target body characteristics as claimed in claim 6 or 7, wherein the apparatus further comprises a file category verification module; the archive category verification module comprises:
the archive association submodule is configured to associate the pre-established archive with a preset target verification library;
the verification data acquisition sub-module is configured to acquire verification data of a target body in the preset target body verification library;
and the file type verification sub-module is configured to match the verification data acquired by the verification data acquisition sub-module with file information in the pre-constructed file library and determine the correct type of the target body according to a matching result.
9. The automatic generation device of the target body characteristic-based electronic file according to claim 6 or 7, characterized in that the device further comprises a target body image and image recognition training sample acquisition module thereof, a target body behavior characteristic analysis module, a specific target body monitoring module, and a specific target body pre/alarm module;
the target body image and image recognition training sample acquisition module thereof is configured to: screening archive information in the pre-constructed archive library according to a preset target body type to obtain an image and an image recognition training sample corresponding to the preset target body type;
the target body behavior feature analysis module is configured to: counting the archive information of the target body, generating a behavior track of the target body according to a counting result, acquiring the occurrence frequency of the target body in a specific area, acquiring the archive information of other target bodies in a current target body image according to the counting result, and further acquiring the behavior characteristics of the other target bodies according to the acquired archive information;
the specific target monitoring module is configured to: monitoring the target body behavior characteristics of a specific target body in real time based on the file information in the pre-constructed file library;
the specific target body pre/alarm module is configured to: comparing a plurality of target body characteristics of a target body with a plurality of characteristics of a specific target body respectively, and carrying out early warning according to comparison results; and/or acquiring and analyzing the behavior track of a specific target body, so that a plurality of specific target bodies give an alarm when the specific target bodies appear in the same area in the same time period.
10. A storage means having stored therein a plurality of programs, characterized in that said programs are adapted to be loaded and executed by a processor to implement the method for automatic generation of an electronic archive based on characteristics of an object as claimed in any one of claims 1 to 5.
11. A processing apparatus, comprising:
a processor adapted to execute various programs;
a storage device adapted to store a plurality of programs;
characterized in that said program is adapted to be loaded and executed by a processor to implement the method for automatic generation of an electronic archive based on characteristics of a target volume according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810166463.3A CN108563675B (en) | 2018-02-28 | 2018-02-28 | Electronic file automatic generation method and device based on target body characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810166463.3A CN108563675B (en) | 2018-02-28 | 2018-02-28 | Electronic file automatic generation method and device based on target body characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108563675A CN108563675A (en) | 2018-09-21 |
| CN108563675B true CN108563675B (en) | 2021-03-30 |
Family
ID=63531423
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810166463.3A Active CN108563675B (en) | 2018-02-28 | 2018-02-28 | Electronic file automatic generation method and device based on target body characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108563675B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008362A (en) * | 2019-03-05 | 2019-07-12 | 青岛海信网络科技股份有限公司 | A kind of case classifying method and device |
| CN110163137A (en) * | 2019-05-13 | 2019-08-23 | 深圳市商汤科技有限公司 | A kind of image processing method, device and storage medium |
| CN110390031A (en) * | 2019-06-28 | 2019-10-29 | 深圳市商汤科技有限公司 | Information processing method and device, vision facilities and storage medium |
| CN113408557B (en) * | 2020-03-17 | 2023-10-13 | 深圳云天励飞技术有限公司 | File merging method and device and electronic equipment |
| CN111479225A (en) * | 2020-03-17 | 2020-07-31 | 赵军 | Medical intelligent body temperature electronic chest card and method based on Internet of things |
| CN114186964A (en) * | 2021-12-03 | 2022-03-15 | 武汉大政科技有限责任公司 | Active engineering file implementation collecting method based on information difference comparison |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104143079B (en) * | 2013-05-10 | 2016-08-17 | 腾讯科技（深圳）有限公司 | The method and system of face character identification |
| CN105404860B (en) * | 2015-11-13 | 2019-01-11 | 北京旷视科技有限公司 | Method and apparatus for managing personal information of wandering away |
| CN106776619B (en) * | 2015-11-20 | 2020-09-04 | 百度在线网络技术（北京）有限公司 | Method and device for determining attribute information of target object |
| CN107590154B (en) * | 2016-07-08 | 2021-03-02 | 阿里巴巴集团控股有限公司 | Object similarity determination method and device based on image recognition |
-
2018
- 2018-02-28 CN CN201810166463.3A patent/CN108563675B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108563675A (en) | 2018-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108563675B (en) | Electronic file automatic generation method and device based on target body characteristics | |
| Kumar et al. | The p-destre: A fully annotated dataset for pedestrian detection, tracking, and short/long-term re-identification from aerial devices | |
| Klare et al. | Pushing the frontiers of unconstrained face detection and recognition: Iarpa janus benchmark a | |
| CN110428399B (en) | Method, apparatus, device and storage medium for detecting image | |
| Alharbi et al. | Automatic counting of wheat spikes from wheat growth images | |
| CN110781733B (en) | Image duplicate removal method, storage medium, network equipment and intelligent monitoring system | |
| CN111160275B (en) | Pedestrian re-recognition model training method, device, computer equipment and storage medium | |
| CN111209776A (en) | Method, device, processing server, storage medium and system for identifying pedestrians | |
| US20200387994A1 (en) | System and method for detecting potential fraud between a probe biometric and a dataset of biometrics | |
| Chandran et al. | Missing child identification system using deep learning and multiclass SVM | |
| WO2019138329A1 (en) | Method and system, based on the use of deep learning techniques, for the univocal biometric identification of an animal | |
| CN107533547B (en) | Product indexing method and system | |
| CN111814690A (en) | Target re-identification method and device and computer readable storage medium | |
| CN109902550A (en) | The recognition methods of pedestrian's attribute and device | |
| Iazzi et al. | Fall detection based on posture analysis and support vector machine | |
| Wang et al. | Deepfake detection: a comprehensive study from the reliability perspective | |
| US20160239682A1 (en) | Method and system of enforcing privacy policies for mobile sensory devices | |
| CN108875500A (en) | Pedestrian recognition methods, device, system and storage medium again | |
| Bhattacharya et al. | A deep neural network framework for detection and identification of bengal tigers | |
| Almeida et al. | Visual rhythm-based time series analysis for phenology studies | |
| CN114049608A (en) | Track monitoring method and device, computer equipment and storage medium | |
| CN116311036A (en) | Behavior recognition method, behavior recognition device, computer equipment and storage medium | |
| CN110751034B (en) | Pedestrian behavior recognition method and terminal equipment | |
| Tagore et al. | SMSNet: A Novel Multi-scale Siamese Model for Person Re-Identification. | |
| Tin | Age Dependent Face Recognition using Eigenface |
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 |