CN109800673A

CN109800673A - A kind of archiving method and device

Info

Publication number: CN109800673A
Application number: CN201811626737.9A
Authority: CN
Inventors: 俞梦洁; 梁晓涛
Original assignee: Fujian Yi Tu Network Technology Co Ltd; Shanghai Is According To Figure Network Technology Co Ltd
Current assignee: Fujian Yi Tu Network Technology Co Ltd; Shanghai Is According To Figure Network Technology Co Ltd; Shanghai Yitu Network Science and Technology Co Ltd
Priority date: 2018-12-28
Filing date: 2018-12-28
Publication date: 2019-05-24

Abstract

The embodiment of the present application provides a kind of archiving method and device, it is related to technical field of image processing, this method comprises: when filing to multiple facial images to be archived, first it is directed to every facial image to be archived, using the spatial information of facial image to be archived as prior information, real name archives are filtered out from facial image database, are filed after being then compared facial image to be archived with real name archives.It will classify with the unmatched facial image to be archived of the real name archives of facial image database according to spatial information, then the facial image to be archived in each classification is clustered, it generates multiple first and files archives in advance, the multiple first pre- filing archives are subjected to the pre- filing archives of cluster determining second.Non- real name archives are filtered out from facial image database using the spatial information of the second pre- filing archives as prior information again later, are filed after the second pre- filing archives are compared with the non-real name archives filtered out, to improve the accuracy and efficiency of filing.

Description

A kind of archiving method and device

Technical field

The present embodiments relate to technical field of image processing more particularly to a kind of archiving methods and device.

Background technique

In today's society, for the needs of safeguard management, it is various public that monitoring camera has been covered with street, community, building etc. Occasion.When alert occurs, police track down and arrest suspect using monitoring camera.

However as the expansion of monitoring network size, video data magnanimity increases.When there is alert, based on suspect's It is more and more difficult that image obtains useful information or information from large nuber of images, not only low efficiency, while human cost is high.

Summary of the invention

Due in the prior art, when there is alert, useful letter being obtained from large nuber of images based on the image of suspect Breath or information are more and more difficult, and not only low efficiency, while the problem that human cost is high, the embodiment of the present application provides one kind and returns Shelves method and device.

On the one hand, the embodiment of the present application provides a kind of archiving method, comprising:

Obtain the spatial information of multiple facial images to be archived and multiple facial images to be archived in preset period of time；

For every facial image to be archived, according to the spatial information of the facial image to be archived, from facial image database Offline profile in obtain the real name archives of N priority；The offline profile of the facial image database includes at least a real name Archives, the archives that the offline profile is filed before being the preset period of time, the real name archives are with identification information Archives；The facial image to be archived does not find matched real name archives in the real name archives of preceding N-1 priority；N is big In 1, preceding N-1 priority is higher than N priority；

Exist and the matched real name archives of facial image to be archived in the real name archives for determining the N priority When, the facial image to be archived is included into the matched real name archives.

Optionally, further includes:

It is not present and the matched real name shelves of facial image to be archived in the real name archives for determining the N priority When case, the real name archives of N+1 priority are obtained from the offline profile of the facial image database；

Determining in the real name archives of the N+1 priority whether there is and the matched real name of facial image to be archived Archives, the real name archives up to finding matched real name archives or each priority, which compare, to be finished.

Optionally, further includes:

After determining that the real name archives of each priority compare, there is no matched with the facial image to be archived When real name archives, it is determined as the facial image to be archived not file facial image.

Optionally, further includes:

The facial image of not filing is classified according to the spatial information for not filing facial image；

Facial image of not filing in sorted each classification is clustered, the multiple first pre- filing archives are generated； The multiple first pre- filing archives are clustered, determine the second pre- filing archives；

For each second pre- filing archives, according to the spatial information of the described second pre- filing archives, from the face figure As library offline profile in obtain the non-real name archives of M priority；The offline profile of the facial image database includes at least one A non-real name archives, the non-real name archives are the archives without identification information；Described second pre- filing archives are preceding Matched non-real name archives are not found in the non-real name archives of M-1 priority；M is greater than 1, and it is excellent that preceding M-1 priority is higher than M First grade；

Exist in the non-real name archives for determining the M priority and the described second pre- filing matched non-real name of archives When archives, the described second pre- filing archives are included into the matched non-real name archives.

Optionally, further includes:

There is no file the matched non-reality of archives in advance with described second in the non-real name archives for determining the M priority When name archives, the non-real name archives of M+1 priority are obtained from the offline profile of the facial image database；

It determines in the non-real name archives of the M+1 priority with the presence or absence of matched non-with the described second pre- filing archives Real name archives, the non-real name archives up to finding matched non-real name archives or each priority, which compare, to be finished.

Optionally, further includes:

After determining that the non-real name archives of each priority compare, there is no match with the described second pre- filing archives Non- real name archives when, establish a new non-real name archives for the described second pre- filing archives.

On the one hand, the embodiment of the present application provides a kind of filing apparatus, comprising:

Module is obtained, for obtaining multiple facial images to be archived and multiple described face figures to be archived in preset period of time The spatial information of picture；

Processing module, for being directed to every facial image to be archived, according to the spatial information of the facial image to be archived, The real name archives of N priority are obtained from the offline profile of facial image database；The offline profile of the facial image database is at least Including a real name archives, the archives that the offline profile is filed before being the preset period of time, the real name archives be with The archives of identification information；The facial image to be archived does not find matched in the real name archives of preceding N-1 priority Real name archives；N is greater than 1, and preceding N-1 priority is higher than N priority；

Profiling module, for existing and the facial image to be archived in the real name archives for determining the N priority When matched real name archives, the facial image to be archived is included into the matched real name archives.

Optionally, the processing module is also used to:

Optionally, further includes:

Categorization module, for carrying out the facial image of not filing according to the spatial information for not filing facial image Classification；

Cluster module generates multiple for clustering the facial image of not filing in sorted each classification One pre- filing archives；The multiple first pre- filing archives are clustered, determine the second pre- filing archives；

The processing module is also used to for each second pre- filing archives, according to the sky of the described second pre- filing archives Between information, the non-real name archives of M priority are obtained from the offline profile of the facial image database；The facial image database Offline profile includes at least a non-real name archives, and the non-real name archives are the archives without identification information；It is described Second pre- filing archives do not find matched non-real name archives in the non-real name archives of preceding M-1 priority；M is greater than 1, preceding M- 1 priority is higher than M priority；

The profiling module is also used to exist in the non-real name archives for determining the M priority with described second in advance When filing the matched non-real name archives of archives, the described second pre- filing archives are included into the matched non-real name archives.

Optionally, the processing module is also used to:

On the one hand, the embodiment of the present application provides a kind of archival device, including at least one processing unit and at least one A storage unit, wherein the storage unit is stored with computer program, when described program is executed by the processing unit, So that the processing unit executes the step of archiving method.

On the one hand, the embodiment of the present application provides a kind of computer-readable medium, and being stored with can be executed by archival device Computer program, when described program is run in archival device, so that the step of archival device holds archiving method.

In the embodiment of the present application, when being filed for multiple facial images to be archived obtained in preset period of time, first For every facial image to be archived, using the spatial information of facial image to be archived as prior information, from facial image database Real name archives are filtered out, are filed after being then compared facial image to be archived with real name archives.It will be with according to spatial information The unmatched facial image to be archived of the real name archives of facial image database is classified, then by the people to be archived in each classification Face image is clustered, and the multiple first pre- filing archives are generated, and the multiple first pre- filing archives are clustered, determine that second is pre- File archives.Therefore the facial image to be archived in the second pre- filing archives includes similar characteristics of image.It is pre- with second again later File the spatial information of archives as prior information, non-real name archives are filtered out from facial image database, are then returned second in advance Shelves archives are filed after being compared with the non-real name archives filtered out, on the one hand improve the accuracy of filing, on the other hand Effectively improve the efficiency of filing.Further, when alert occurs, police can be directly according to the image of suspect from people The archives of suspect are obtained in face image library, and obtain useful information or information from the archives of suspect, improve police service people Member obtains the efficiency of information, while having saved human cost.

Detailed description of the invention

To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this For the those of ordinary skill in field, without any creative labor, it can also be obtained according to these attached drawings His attached drawing.

Fig. 1 is the system architecture diagram that the embodiment of the present application is applicable in；

Fig. 2 is a kind of flow diagram of archiving method of the embodiment of the present application；

Fig. 3 is a kind of flow diagram of archiving method of the embodiment of the present application；

Fig. 4 is a kind of flow diagram of archiving method of the embodiment of the present application；

Fig. 5 is a kind of flow diagram of archiving method of the embodiment of the present application；

Fig. 6 is a kind of structural schematic diagram of filing apparatus of the embodiment of the present application；

Fig. 7 is a kind of structural schematic diagram of archival device of the embodiment of the present application.

Specific embodiment

In order to which the purpose of the present invention, technical solution and beneficial effect is more clearly understood, below in conjunction with attached drawing and implementation Example, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only used to explain this hair It is bright, it is not intended to limit the present invention.

Archiving method in the embodiment of the present application can be applied to security protection, using the facial image of monitoring device acquisition as base Plinth establishes facial image database, includes multiple face archives in facial image database, and an archives correspond to a people, wherein face shelves The archiving method of case can be divided into online elite archiving and offline filing according to archiving method.Online elite archiving is real-time to monitoring device The method that the facial image of acquisition is filed, offline filing are to be filed to the facial image acquired in set period of time Method.For convenience of description, the face archives in the embodiment of the present application are known as online archives during online elite archiving, and the application is real It applies the face archives in example and is known as offline profile in offline archiving process.Offline filing can be used for updating online elite archiving.? Include real name archives and non-real name archives in line archives, also includes real name archives and non-real name archives in offline profile, wherein real Include personal identification information in name archives, for example includes identity document in real name archives.

When alert occurs, police can with the face archives in the picture search facial image database of suspect, from The corresponding face archives of middle acquisition suspect, analyze suspect according to the facial image in the corresponding face archives of suspect later Activity trajectory, frequent activities the information such as regional, most emerging position, consequently facilitating police arrest suspect.

Fig. 1 illustrates the applicable system architecture of the embodiment of the present application, includes monitoring device in the system architecture 101, server 102.Acquisition video flowing takes monitoring device 101 then by the video stream of acquisition to server 102 in real time It is engaged in including filing apparatus in device 102, server 102 obtains facial image to be archived from video flowing, then will be to be archived Facial image is included into corresponding face archives.Monitoring device 101 is connect by wireless network with server 102, and monitoring device is The electronic equipment, such as camera, video camera, video recorder etc. for having acquisition image function.Server 102 be a server or The server cluster or cloud computing center of several servers composition.

Based on system architecture shown in FIG. 1, the embodiment of the present application provides a kind of process of archiving method, the stream of this method Journey can be executed by filing apparatus, and filing apparatus can be server 102 shown in FIG. 1, as shown in Figure 2, comprising the following steps:

Step 201, the space of multiple facial images to be archived and multiple facial images to be archived in preset period of time is obtained Information.

Preset period of time can be set according to the actual situation, for example preset period of time is one day.Facial image to be archived is The facial image of in running order monitoring device acquisition.Each facial image to be archived carries spatial information and time Information, wherein spatial information is mark and the position for the monitoring device for capturing facial image to be archived.Temporal information is monitoring Equipment captures the time of facial image to be archived.

Illustratively, the spatial information of facial image A to be archived and temporal information are as follows:

Spatial information: camera identification are as follows: CAMERA_A, camera position: Xuhui District of Shanghai.

Temporal information: 2018-10-20 10:07:21.

Step 202, for every facial image to be archived, according to the spatial information of facial image to be archived, from face figure As library offline profile in obtain the real name archives of N priority.

Specifically, facial image database includes offline profile and online archives, filing before offline profile is preset period of time Archives, offline profile are divided into real name archives and non-real name archives, and real name archives are the archives with identification information, non-real name Archives are the archives without identification information.The offline profile of facial image database includes at least a real name archives, wait return Shelves facial image does not find matched real name archives in the real name archives of preceding N-1 priority；N is greater than 1, and preceding N-1 preferential Grade is higher than N priority.Identification presentation information in real name archives can be the relevant information of identity document, such as identification card number Code, identity card picture, identity card are issued licence address etc..

It is each preferential for the ease of being obtained from the offline profile of facial image database according to the spatial information of facial image to be archived An information record sheet is arranged in each real name archives, has recorded in real name archives in the record sheet for the real name archives of grade The spatial information and temporal information of facial image.

In a kind of possible embodiment, what the facial image being recorded in real name archives in information record sheet was related to is taken the photograph As leader know, each camera capture facial image quantity and the newest candid photograph of each camera facial image when Between.

Illustratively, setting in real name archives 1 includes 6 facial images, wherein 5 facial images are by camera CAMERA_A is captured, and in addition 1 facial image is captured by camera CAMERA_B, then the information record sheet of real name archives 1 is specific It is as shown in table 1:

Table 1.

Camera identification	The quantity of the facial image of candid photograph	The time of the facial image of newest candid photograph
			CAMERA_A	5	2018-10-23 10:05:15
CAMERA_B	1	2018-10-23 10:07:21

Record in real name archives that every facial image is corresponding takes the photograph in a kind of possible embodiment, in information record sheet As the mark of head, the position of the corresponding camera of every facial image, every facial image the candid photograph time.

Illustratively, setting in real name archives 2 includes 6 facial images, and wherein facial image a, facial image b are by imaging Head CAMERA_A is captured, and facial image c, facial image d capture by camera CAMERA_B, facial image e, facial image f by Camera CAMERA_C is captured, then the information record sheet of real name archives 2 is specifically as shown in table 2:

Table 2.

Facial image	Camera identification	The position of camera	The candid photograph time of facial image
				a	CAMERA_A	Xuhui District of Shanghai	2018-10-21 10:07:21
b	CAMERA_A	Xuhui District of Shanghai	2018-10-23 09:40:11
				c	CAMERA_B	Pudong New Area, Shanghai	2018-10-24 08:20:31
d	CAMERA_B	Pudong New Area, Shanghai	2018-10-24 09:40:51
				e	CAMERA_C	Hongkou district Shanghai	2018-10-23 07:40:14
f	CAMERA_C	Hongkou district Shanghai	2018-10-23 09:22:11

Step 203, exist and the matched real name archives of facial image to be archived in the real name archives for determining N priority When, facial image to be archived is included into matched real name archives.

Specifically, multiple screening strategies are determined according to spatial information, face figure to be archived is successively determined according to screening strategy The real name archives of each priority of picture, wherein the corresponding spatial dimension of screening strategy is smaller, and determining priority is higher.In root When screening real name archives according to screening strategy, matched as long as there are a facial images in real name archives with screening strategy, it should Real name archives are determined as will be with each sieve later further according to the priority of each screening strategy with the matched real name archives of screening strategy The real name archives of strategy matching are selected to be determined as the real name archives of each priority of the first facial image.One priority of every determination Facial image to be archived is compared, when in the real name archives of the priority by real name archives with the real name archives of the priority When in the presence of real name archives matched with facial image to be archived, facial image to be archived is included into matched real name archives.

Illustratively, the spatial information for setting facial image to be archived includes: camera identification as CAMERA_A, camera Position is Xuhui District of Shanghai.The sequence according to spatial dimension from small to large is set, determines the following sieve of facial image to be archived Choosing strategy:

Screening strategy 1: it is excellent that the 1st is filtered out from the offline profile of facial image database according to " camera identification CAMERA_A " The real name archives of first grade.

Screening strategy 2: the real name shelves of the 2nd priority are filtered out from the offline profile of facial image database according to " Xuhui District " Case.

Screening strategy 3: the real name shelves of the 3rd priority are filtered out from the offline profile of facial image database according to " Shanghai City " Case.

Screening strategy 4: real name archives of all real name archives as the 4th priority in the offline profile of facial image database.

It further, include camera identification in real name archives 2 shown in real name archives 1 and table 2 as shown in table 1 For the facial image of CAMERA_A, then the 1st priority is filtered out from the offline profile of facial image database according to screening strategy 1 Real name archives are real name archives 1 and real name archives 2, and facial image to be archived and real name archives 1 are compared, determined to be archived When facial image is matched with real name archives 1, facial image to be archived is included into real name archives 1.

Since the life range of many people is relatively fixed, a people may often be captured by the camera of certain areas, Therefore when filing to the facial image to be archived of candid photograph, first believed using the spatial information of facial image to be archived as priori Breath, filters out real name archives from the offline profile of facial image database, then by facial image to be archived and the real name filtered out Archives are filed after being compared, and can effectively improve the efficiency of real name filing.

Optionally, it in above-mentioned steps S203, is not present and face to be archived in the real name archives for determining N priority When the real name archives of images match, the real name archives of N+1 priority are obtained from the offline profile of facial image database.Determine In the real name archives of N+1 priority with the presence or absence of with the matched real name archives of facial image to be archived, until find matched reality Name archives or the real name archives of each priority are compared and are finished.

Illustratively, setting filters out the 1st priority according to above-mentioned screening strategy 1 from the offline profile of facial image database Real name archives be real name archives 1 and real name archives 2, by facial image to be archived respectively with real name archives 1 and real name archives 2 into Row compares, and determines that facial image to be archived is mismatched with real name archives 1 and real name archives 2, then according to screening strategy 2 from face The real name archives of the 2nd priority are filtered out in the offline profile of image library.The real name archives of the 2nd priority are set as real name archives 1, real name archives 2, real name archives 3, real name archives 4.By facial image to be archived respectively with real name archives 3 and real name archives 4 into Row compares, and facial image to be archived is compared with real name archives 1 and real name archives 2, therefore can no longer be compared at this time Compared with.When being matched there are a real name archives with facial image to be archived in real name archives 3 and real name archives 4, by people to be archived Face image is included into matched real name archives, when real name archives 3 and real name archives 4 are mismatched with facial image to be archived, root The real name archives of the 3rd priority are filtered out from the offline profile of facial image database according to screening strategy 3, and so on, Zhi Dao Found in the offline profile of facial image database with real name archives matched in facial image to be archived or facial image to be archived with The real name archives comparison of 4th priority finishes.

Further, it after determining that the real name archives of each priority compare, is not present and facial image to be archived When matched real name archives, it is determined as facial image to be archived not file facial image.

Illustratively, if after facial image to be archived is compared with the real name archives of above-mentioned 4th priority, do not find with When the matched real name archives of facial image to be archived, illustrate to be not present in the offline profile of facial image database and face figure to be archived As matched real name archives, then it is determined as facial image to be archived not file facial image.

Optionally, in above-mentioned steps S203, exist in the specific real name archives for determining N priority using following methods With the matched real name archives of facial image to be archived:

For any real name archives in the real name archives of N priority, according to the class center of real name archives with it is to be archived Similarity between the class of facial image determines the similarity of real name archives Yu facial image to be archived.

If similarity is greater than or equal to first threshold, it is determined that exist and people to be archived in the real name archives of N priority The matched real name archives of face image.

Specifically, for the real name archives of different priorities, the first threshold of setting is not also identical, wherein priority is got over Height, corresponding first threshold are smaller.Illustratively, the corresponding first threshold of the 1st priority is set as T0, and the 2nd priority is corresponding First threshold be T1, the corresponding first threshold of the 3rd priority be T2, the corresponding first threshold of the 4th priority be T3, then T0 < T1<T2<T3。

In a kind of possible embodiment, the class center of real name archives is one, such center is to scheme in real name archives The highest facial image of image quality amount.By similarity between such center and the class of facial image to be archived, be determined as real name archives with Similarity between facial image to be archived.

In specific implementation, when determining similarity between class center and the class of facial image to be archived, class can be determined first The corresponding first eigenvector in center and the corresponding second feature vector of facial image to be archived, then calculate fisrt feature to The distance between amount and second feature vector determine class center according to the distance between first eigenvector and second feature vector The similarity between the class of facial image to be archived.

In alternatively possible embodiment, the class centers of real name archives be it is multiple, determine facial image to be archived with Similarity between the class at all kinds of centers of real name archives, then according to the weight at all kinds of centers of real name archives and all kinds of centers Similarity between class determines the similarity of real name archives Yu facial image to be archived.

When determining similarity between facial image to be archived and the class at multiple class centers, people to be archived can be first determined respectively Similarity between face image and the class at each class center, then according to the weight at each class center by facial image to be archived with Similarity is merged between the class at each class center, determines the similarity of real name archives Yu facial image to be archived, class center Weight can be allocated according to picture quality, can also be allocated according to characteristics of image.

Since real name archives indicate archives with class center, therefore when filing to facial image to be archived, only needing will be to Filing facial image is compared with the class center of real name archives, the determining and matched real name archives of facial image to be archived, and It does not need for facial image to be archived to be compared with every facial image in real name archives, to improve filing efficiency.

Optionally, for not having in multiple facial images to be archived to carry out real name filing not file facial image, this Shen It please be filed by the way of the filing of non-real name in embodiment, specifically include following steps, as shown in Figure 3:

Step S301 will not file facial image according to the spatial information for not filing facial image and classify.

Specifically, classification policy, such as camera classification policy can be set previously according to spatial information, i.e., according to camera shooting Leader knows the strategy classified.Subregion classification policy, i.e., the strategy classified according to subregion.City-level territorial classification plan Slightly, the strategy classified according to city-level region.Provincial region classification policy, the strategy classified according to provincial region.

Be specifically described below for camera classification policy, setting obtain 100 do not file facial image and this 100 Open the spatial information for not filing facial image, wherein spatial information is the mark for the camera that facial image is not filed in acquisition.It presses Do not file the corresponding camera identification of facial image according to every and do not file facial image for this 100 and classifies.Assuming that this 100 It opens and does not file in facial image, having 50 not file the corresponding camera identification of facial image is CAMERA_A, and 20 are not filed The corresponding camera identification of facial image is CAMERA_B, and not filing the corresponding camera identification of facial image for 30 is CAMERA_C does not then file facial image for this 100 and is divided into above-mentioned 3 class.

It should be noted that according to subregion classification policy, city-level territorial classification strategy and provincial region classification policy To not filing, the process that facial image is classified is identical as the process classified according to camera classification policy, herein no longer It repeats.

Step S302 clusters the facial image of not filing in sorted each classification, and it is pre- to generate multiple first File archives, the multiple first pre- filing archives are clustered, determines the second pre- filing archives.

Specifically, do not file facial image in each sorted classification, using clustering algorithm by the category In facial image of not filing clustered, clustering algorithm include but is not limited to K-Means (K mean value) clustering algorithm, mean value drift Move clustering algorithm, density-based algorithms.

Illustratively, setting does not file facial image according to 100 that camera identification will acquire and is divided into three classes, the first kind Do not file facial image including 50, the second class includes 20 and do not file facial image, and third class includes 30 and do not file face Image.Do not file facial image to 50 in the first kind first to cluster, not file facial image to gather be one by similar Class.Assuming that not filing facial image after end of clustering for 50 in the first kind and being polymerized to 3 subclasses, wherein including in the first subclass 20 are not filed facial image, in the second subclass are included 20 and are not filed facial image, include 10 in third subclass and do not file Facial image does not then file facial image for 20 in the first subclass and establishes one first pre- filing archives, for second 20 of subclass do not file facial image and establish one first pre- filing archives, do not file face figure for 10 of third subclass As establishing one first pre- filing archives.Then using same method to 20 in the second class do not file facial image and Do not file facial image for 30 in third class to be clustered and establish the corresponding first pre- filing archives.Further, by The multiple first pre- filing archives that one kind is generated into third class are clustered, and are generated second and are filed archives in advance.

Step S303, for each second pre- filing archives, according to the spatial information of the second pre- filing archives, from face figure As library offline profile in obtain the non-real name archives of M priority.

The offline profile of facial image database includes at least a non-real name archives, and non-real name archives are without identity The archives of information；Second pre- filing archives do not find matched non-real name archives in the non-real name archives of preceding M-1 priority； M is greater than 1, and preceding M-1 priority is higher than M priority.

Specifically, it is obtained from the offline profile of facial image database for the ease of filing the spatial information of archives in advance according to second Corresponding non-real name archives are taken, an information record sheet is set in each non-real name archives, is had recorded in the record sheet non- The spatial information and temporal information of facial image in real name archives.It should be noted that the information record sheet of non-real name archives with The form of the information record sheet of real name archives is identical, and details are not described herein again.

Step S304 exists and the second pre- filing matched non-reality of archives in the non-real name archives for determining M priority When name archives, the second pre- filing archives are included into matched non-real name archives.

Specifically, multiple screening strategies are determined according to spatial information and classification policy above-mentioned, according to screening strategy according to The non-real name archives of each priority of archives are filed in secondary determination second in advance, wherein the corresponding spatial dimension of screening strategy is smaller, really Fixed priority is higher.The non-real name archives of one priority of every determination, by the non-reality of the second pre- filing archives and the priority Name archives are compared, when existing in the non-real name archives of the priority and the second pre- filing matched non-real name archives of archives When, the second pre- filing archives are included into matched non-real name archives.

Illustratively, setting does not file facial image for 100 according to camera classification policy and is divided into three classes, first kind packet It including 50 and does not file facial image, corresponding camera identification is CAMERA_A, and the second class includes 20 and do not file facial image, Corresponding camera identification is CAMERA_B, and third class includes 30 and do not file facial image, and corresponding camera identification is CAMERA_C.50 in the first kind are not filed and obtain three first pre- filing archives after facial image is clustered, respectively For the first pre- filing archives A, the first pre- filing archives B and first, filing archives C, these three first pre- filing archives are corresponding in advance Camera identification be CAMERA_A.20 in second class are not filed and obtain two first after facial image is clustered in advance File archives, filing archives E, the two first pre- filing archives are corresponding in advance by the respectively first pre- filing archives D and first Camera identification is CAMERA_B.30 in third class are not filed and obtains two first after facial image is clustered and returns in advance Shelves archives, the respectively first pre- filing archives F and first filing archives G in advance, archives are corresponding takes the photograph for the two first pre- filings As head is identified as CAMERA_C.Further, by the first pre- filing archives A to first, filing archives G is clustered in advance, and setting is poly- It has been determined that four second are filed archives in advance after class, wherein the first pre- filing archives A, the first pre- archives D that files constitute second in advance File archives L, the first pre- filing archives C, the first pre- filing archives E constitute the second pre- filing archives M, the first pre- filing archives Filing archives F constitutes the second pre- filing archives N to B and first in advance, and the first pre- filing archives G is as the second pre- filing archives O.

According to spatial dimension sequence from small to large and classification policy, the following screening plan of the second pre- filing archives is determined Slightly:

Screening strategy 1: the non-of the 1st priority is filtered out from the offline profile of facial image database according to " camera identification " Real name archives.

Screening strategy 2: non-real name shelves of all non-real name archives as the 2nd priority in the offline profile of facial image database Case.

Further, for the second pre- filing archives L, the face for being CAMERA_A and CAMERA_B comprising camera identification Image.It sets and the non-of the 1st priority is filtered out as CAMERA_A from the offline profile of facial image database according to camera identification Real name archives are non-real name archives 1 and non-real name archives 2, are CAMERA_B from the offline of facial image database according to camera identification The non-real name archives that the 1st priority is filtered out in archives are non-real name archives 3.

Second pre- filing archives L is compared with non-real name archives 1, determines the second pre- filing archives L and non-real name shelves When case 1 matches, the second pre- filing archives L is included into non-real name archives 1.It can be to the out second pre- filing shelves based on same method Filing archives O is filed case M to second in advance, and details are not described herein again.

Since the life range of many people is relatively fixed, a people may often be captured by the camera of certain areas, Therefore when to not filing facial image and filing, first classify according to the spatial information for not filing facial image, it then will be each Facial image of not filing in classification is clustered, and the multiple first pre- filing archives are generated, will multiple first file in advance archives into Row cluster determines the second pre- filing archives, therefore the facial image of not filing in the second pre- filing archives includes that similar image is special Sign.It is screened from the offline profile of facial image database using the spatial information of the second pre- filing archives as prior information again later Non- real name archives out are filed after being then compared the second pre- filing archives with the non-real name archives filtered out, can effectively be mentioned The efficiency of high non-real name filing.

Optionally, in above-mentioned steps S304, there is no return in advance with second in the non-real name archives for determining M priority When the shelves matched non-real name archives of archives, the non-real name archives of M+1 priority are obtained from the offline profile of facial image database. It determines in the non-real name archives of M+1 priority and files the matched non-real name archives of archives in advance with the presence or absence of with second, until looking for Non- real name archives to matched non-real name archives or each priority are compared and are finished.

Illustratively, for the second pre- filing archives L, set according to camera identification as CAMERA_A and CAMERA_B from The non-real name archives that the 1st priority is filtered out in the offline profile of facial image database are non-real name archives 1, non-2 and of real name archives Non- real name archives 3 compare the second pre- filing archives L with non-real name archives 1, non-real name archives 2 and non-real name archives 3 respectively Compared with determining that the second pre- filing archives L is mismatched with non-real name archives 1, non-real name archives 2 and non-real name archives 3, then by face Non- real name archives of all non-real name archives as the 2nd priority in the offline profile of image library.Set the non-of the 2nd priority Real name archives are non-real name archives 1, non-real name archives 2, non-real name archives 3, non-real name archives 4.By the second pre- filing archives L with Non- real name archives 4 are compared, the second pre- filing archives L with non-real name archives 1, non-real name archives 2 and non-real name archives 3 It compares, therefore can no longer compare.When being matched in non-real name archives 4 with the second pre- filing archives L, second is returned in advance Shelves archives L is included into matched non-real name archives 4.

Further, after determining that the non-real name archives of each priority compare, there is no file shelves in advance with second When the matched non-real name archives of case, a new non-real name archives are established for the second pre- filing archives.

Illustratively, for the second pre- filing archives L, when non-real name archives 4 and the second pre- filing archives L mismatch, Illustrate there is no the matched non-real name archives of archives L are filed in advance with second in the non-real name archives of facial image database, then in face A new non-real name archives are established in image library, and the second pre- filing archives L is included into new non-real name archives.

Optionally, it in above-mentioned steps S304, is deposited in the specific non-real name archives for determining M priority using following methods Filing the matched non-real name archives of archives in advance with second:

For any non-real name archives in the non-real name archives of M priority, according to the class center of non-real name archives with The similarity at the class center of the second pre- filing archives determines the similarity of non-real name archives and the second pre- filing archives.If not real The similarity of name archives and the second pre- filing archives is greater than or equal to first threshold, it is determined that the non-real name archives of M priority It is middle to exist and the second pre- filing matched non-real name archives of archives.

Specifically, for the non-real name archives of different priorities, the first threshold of setting is not also identical, wherein priority Higher, corresponding first threshold is smaller.Illustratively, the corresponding first threshold of the 1st priority is set as T0, the 2nd priority pair The first threshold answered is T1, then T0 < T1.

In a kind of possible embodiment, the class center of non-real name archives is one, such center is non-real name archives The middle highest facial image of picture quality.The class center of second pre- filing archives is one, such center is the second pre- filing shelves The highest facial image of picture quality in case.The class center of non-real name archives is similar to the class center of the second pre- filing archives Degree, the similarity being determined as between non-real name archives and the second pre- filing archives.

Similarity in specific implementation, at the class center for determining non-real name archives and the class center of the second pre- filing archives When, the corresponding first eigenvector in class center of non-real name archives and the class center of the second pre- filing archives can be determined first Then corresponding second feature vector calculates the distance between first eigenvector and second feature vector, according to fisrt feature The distance between vector and second feature vector determine the class center and the class center of the second pre- filing archives of non-real name archives Similarity.

In alternatively possible embodiment, the class center of non-real name archives is multiple, multiple classes of non-real name archives Center can be determined according to the picture quality and characteristics of image of the facial image in non-real name archives.The class of second pre- filing archives Center is one, such center is the highest facial image of picture quality in the second pre- filing archives.Determine the second pre- filing shelves The similarity at the class center of case and all kinds of centers of non-real name archives, then according to the class center of the second pre- filing archives and non-reality The similarity at all kinds of centers of name archives determines the similarity of non-real name archives and the second pre- filing archives.

In the similarity at multiple class centers at the class center and non-real name archives for determining the second pre- filing archives, Ke Yixian The class center of the second pre- filing archives and the similarity at non-each class center of real name archives are determined respectively, then according to non-real name The weight at each class center of archives is similar to non-each class center of real name archives by the class center of the second pre- filing archives Degree is merged, and determines the similarity of non-real name archives and the second pre- filing archives, the power at non-each class center of real name archives Weight can be allocated according to picture quality, can also be allocated according to characteristics of image.

In alternatively possible embodiment, the class center of the second pre- filing archives is multiple, the second pre- filing archives Multiple class centers can be according to not filing the picture quality of facial image in the second pre- filing archives and characteristics of image determines.It is non- The class center of real name archives is one, such center is the highest facial image of picture quality in non-real name archives.Determine non-reality The similarity at the class center of name archives and all kinds of centers of the second pre- filing archives, then according to the class center of non-real name archives with The similarity at all kinds of centers of the second pre- filing archives determines the similarity of non-real name archives and the second pre- filing archives.

In the similarity at the class center for determining non-real name archives and all kinds of centers of the second pre- filing archives, can first divide The similarity for not determining the class center and second pre- filing each class center of archives of non-real name archives, then returns according to second in advance The phase at each class center of archives is filed at the class center of non-real name archives and second by the weight at each class center of shelves archives in advance It is merged like degree, determines the similarity of non-real name archives and the second pre- filing archives, second files in each class of archives in advance The weight of the heart can be allocated according to picture quality, can also be allocated according to characteristics of image.

In alternatively possible embodiment, the class center of non-real name archives is multiple, multiple classes of non-real name archives Center can be determined according to the picture quality and characteristics of image of the facial image in non-real name archives.The class of second pre- filing archives Center be it is multiple, multiple class centers of the second pre- filing archives can not file facial image according in the second pre- filing archives Picture quality and characteristics of image determine.Determine the phase at all kinds of centers and all kinds of centers of the second pre- filing archives of non-real name archives Like degree, then according to the similarity at all kinds of centers of non-real name archives and all kinds of centers of the second pre- filing archives, determine offline The similarity of archives and the second pre- filing archives.

Archives are indicated since non-real name archives and the second pre- filing archives are all made of class center, therefore to the second pre- filing When archives carry out the filing of non-real name, the class center at the class center of the second pre- filing archives and non-real name archives need to only be compared Compared with, it is determining to file the matched non-real name archives of archives in advance with second, without filing archives and non-real name archives in advance for second In every facial image be compared, to improve non-real name filing efficiency.

Embodiment in order to preferably explain the present invention describes the embodiment of the present application below with reference to specific implement scene and provides A kind of archiving method, the archiving method include real name filing and non-real name file two parts, this method can be by filing apparatus It executes, the screening strategy of following real name archives is set previously according to spatial information:

First camera screening strategy: it is excellent that the 1st is filtered out from the offline profile of facial image database according to camera identification The real name archives of first grade.

First subregion screening strategy: the 2nd priority is filtered out from the offline profile of facial image database according to subregion Real name archives.

First city-level region screening strategy: it is excellent that the 3rd is filtered out from the offline profile of facial image database according to city-level region The real name archives of first grade.

First all screening strategies: reality of all real name archives as the 4th priority in the offline profile of facial image database Name archives.

Multiple facial images to be archived obtained for one day, filing apparatus is first against each facial image to be archived Real name filing is carried out, following steps are specifically included, as shown in Figure 4:

Step S401 filters out people to be archived based on the first camera screening strategy from the offline profile of facial image database The real name archives of 1st priority of face image.

Facial image to be archived is compared step S402 with the real name archives of the 1st priority.

Step S403 judges to whether there is and the matched real name of facial image to be archived in the real name archives of the 1st priority Archives, if so, S404 is thened follow the steps, it is no to then follow the steps S405.

Facial image to be archived is included into matched real name archives by step S404.

Step S405 filters out people to be archived based on the first subregion screening strategy from the offline profile of facial image database The real name archives of 2nd priority of face image.

Facial image to be archived is compared step S406 with the real name archives of the 2nd priority.

Step S407 judges to whether there is and the matched real name of facial image to be archived in the real name archives of the 2nd priority Archives, if so, S408 is thened follow the steps, it is no to then follow the steps S409.

Facial image to be archived is included into matched real name archives by step S408.

Step S409 is filtered out from the offline profile of facial image database to be archived based on the first city-level region screening strategy The real name archives of 3rd priority of facial image.

Facial image to be archived is compared step S410 with the real name archives of the 3rd priority.

Step S411 judges to whether there is and the matched real name of facial image to be archived in the real name archives of the 3rd priority Archives, if so, S412 is thened follow the steps, it is no to then follow the steps S413.

Facial image to be archived is included into matched real name archives by step S412.

Step S413 filters out face to be archived based on first all screening strategies from the offline profile of facial image database The real name archives of 4th priority of image.

Facial image to be archived is compared step S414 with the real name archives of the 4th priority.

Step S415 judges to whether there is and the matched real name of facial image to be archived in the real name archives of the 4th priority Archives, if so, S416 is thened follow the steps, it is no to then follow the steps S417.

Facial image to be archived is included into matched real name archives by step S416.

Step S417 is determined as facial image to be archived not file facial image.

Further, determine in multiple facial images to be archived obtained in one day it is all do not file facial image it Afterwards, non-real name filing is carried out to all facial images of not filing, determines classification policy for camera point previously according to spatial information Class strategy sets the screening strategy of following non-real name archives based on spatial information and classification policy:

Second camera screening strategy: it is excellent that the 1st is filtered out from the offline profile of facial image database according to camera identification The non-real name archives of first grade.

Second all screening strategies: all non-real name archives are as the 2nd priority in the offline profile of facial image database Non- real name archives.

Following steps are specifically included, as shown in Figure 5:

Step S501 is not filed facial image to multiple based on camera classification policy and classified.

Step S502 clusters the facial image of not filing in sorted each classification, and it is pre- to generate multiple first File archives, the multiple first pre- filing archives are clustered, determines the second pre- filing archives.

Step S503, for each second pre- filing archives, based on second camera screening strategy from facial image database The non-real name archives of the 1st priority of the second pre- filing archives are filtered out in offline profile.

Second pre- filing archives are compared step S504 with the non-real name archives of the 1st priority.

Step S505 judges in the non-real name archives of the 1st priority with the presence or absence of matched non-with the second pre- filing archives Real name archives, if so, S506 is thened follow the steps, it is no to then follow the steps S507.

Second pre- filing archives are included into matched non-real name archives by step S506.

Step S507 filters out the second pre- filing based on second all screening strategies from the offline profile of facial image database The non-real name archives of 2nd priority of archives.

Second pre- filing archives are compared step S508 with the non-real name archives of the 2nd priority.

Step S509 judges in the non-real name archives of the 2nd priority with the presence or absence of matched non-with the second pre- filing archives Real name archives, if so, S510 is thened follow the steps, it is no to then follow the steps S511.

Second pre- filing archives are included into matched non-real name archives by step S510.

Step S511 establishes a new non-real name archives for the second pre- filing archives.

In the embodiment of the present application, when being filed for multiple facial images to be archived obtained in preset period of time, first For every facial image to be archived, using the spatial information of facial image to be archived as prior information, from facial image database Real name archives are filtered out, are filed after being then compared facial image to be archived with real name archives.It will be with according to spatial information The unmatched facial image to be archived of the real name archives of facial image database is classified, then by the people to be archived in each classification Face image is clustered, and the multiple first pre- filing archives are generated, and the multiple first pre- filing archives are clustered, determine that second is pre- File archives, therefore the facial image to be archived in the second pre- filing archives includes similar characteristics of image.It is pre- with second again later File the spatial information of archives as prior information, non-real name archives are filtered out from facial image database, are then returned second in advance Shelves archives are filed after being compared with the non-real name archives filtered out, on the one hand improve the accuracy of filing, on the other hand Effectively improve the efficiency of filing.Further, when alert occurs, police can be directly according to the image of suspect from people The archives of suspect are obtained in face image library, and obtain useful information or information from the archives of suspect, improve police service people Member obtains the efficiency of information, while having saved human cost.

Based on the same technical idea, the embodiment of the invention provides a kind of filing apparatus, as shown in fig. 6, the device 600 Include:

Module 601 is obtained, for obtaining multiple facial images to be archived and multiple described people to be archived in preset period of time The spatial information of face image；

Processing module 602 is believed for being directed to every facial image to be archived according to the space of the facial image to be archived Breath, obtains the real name archives of N priority from the offline profile of facial image database；The offline profile of the facial image database is extremely It less include a real name archives, the archives that the offline profile is filed before being the preset period of time, the real name archives are tool There are the archives of identification information；The facial image to be archived does not find matching in the real name archives of preceding N-1 priority Real name archives；N is greater than 1, and preceding N-1 priority is higher than N priority；

Profiling module 603, for existing and the face figure to be archived in the real name archives for determining the N priority When as matched real name archives, the facial image to be archived is included into the matched real name archives.

Optionally, the processing module 602 is also used to:

Optionally, further includes:

Categorization module 604, for not filing facial image for described according to the spatial information for not filing facial image Classify；

Cluster module 605 generates multiple for clustering the facial image of not filing in sorted each classification First pre- filing archives cluster the multiple first pre- filing archives, determine the second pre- filing archives；

The processing module 602 is also used to for each second pre- filing archives, according to the described second pre- filing archives Spatial information obtains the non-real name archives of M priority from the offline profile of the facial image database；The facial image database Offline profile include at least a non-real name archives, the non-real name archives be the archives without identification information；Institute It states the second pre- filing archives and does not find matched non-real name archives in the non-real name archives of preceding M-1 priority；M is preceding greater than 1 M-1 priority is higher than M priority；

The profiling module 603 is also used to exist and described second in the non-real name archives for determining the M priority When the matched non-real name archives of pre- filing archives, the described second pre- filing archives are included into the matched non-real name archives.

Optionally, the processing module 602 is also used to:

Optionally, the processing module 603 is also used to:

Based on the same technical idea, the embodiment of the present application provides a kind of archival device, as shown in fig. 7, comprises at least One processor 701, and the memory 702 connecting at least one processor do not limit processor in the embodiment of the present application Specific connection medium between 701 and memory 702 is connected as in Fig. 7 by bus between processor 701 and memory 702 Example.Bus can be divided into address bus, data/address bus, control bus etc..

In the embodiment of the present application, memory 702 is stored with the instruction that can be executed by least one processor 701, at least The instruction that one processor 701 is stored by executing memory 702, can execute step included in aforementioned archiving method.

Wherein, processor 701 is the control centre of archival device, can use various interfaces and connection archival device Various pieces, by running or executing the instruction being stored in memory 702 and calling is stored in number in memory 702 According to realize filing.Optionally, processor 701 may include one or more processing units, and processor 701 can integrate application Processor and modem processor, wherein the main processing operation system of application processor, user interface and application program etc., Modem processor mainly handles wireless communication.It is understood that above-mentioned modem processor can not also be integrated into In processor 701.In some embodiments, processor 701 and memory 702 can be realized on the same chip, in some realities It applies in example, they can also be realized respectively on independent chip.

Processor 701 can be general processor, such as central processing unit (CPU), digital signal processor, dedicated integrated Circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array or other can Perhaps transistor logic, discrete hardware components may be implemented or execute the application implementation for programmed logic device, discrete gate Each method, step and logic diagram disclosed in example.General processor can be microprocessor or any conventional processor Deng.The step of method in conjunction with disclosed in the embodiment of the present application, can be embodied directly in hardware processor and execute completion, Huo Zheyong Hardware and software module combination in processor execute completion.

Memory 702 is used as a kind of non-volatile computer readable storage medium storing program for executing, can be used for storing non-volatile software journey Sequence, non-volatile computer executable program and module.Memory 702 may include the storage medium of at least one type, It such as may include flash memory, hard disk, multimedia card, card-type memory, random access storage device (Random Access Memory, RAM), static random-access memory (StaticRandom Access Memory, SRAM), may be programmed read-only deposit Reservoir (Programmable Read OnlyMemory, PROM), read-only memory (Read Only Memory, ROM), electrification Erasable Programmable Read Only Memory EPROM (Electrically Erasable Programmable Read-Only Memory, EEPROM), magnetic storage, disk, CD etc..Memory 702 can be used for carrying or storing have instruction or data The desired program code of structure type and can by any other medium of computer access, but not limited to this.The application is real Applying the memory 702 in example can also be circuit or other devices that arbitrarily can be realized store function, for storing program Instruction and/or data.

Based on the same inventive concept, the embodiment of the present application provides a kind of computer-readable medium, and being stored with can be by returning The computer program that shelves equipment executes, when described program is run in archival device, so that the archival device executes filing The step of method.

It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method or computer program product. Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the present invention Form.It is deposited moreover, the present invention can be used to can be used in the computer that one or more wherein includes computer usable program code The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) Formula.

The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates, Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or The function of being specified in multiple boxes.

These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one The step of function of being specified in a box or multiple boxes.

Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as It selects embodiment and falls into all change and modification of the scope of the invention.

Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies Within, then the present invention is also intended to include these modifications and variations.

Claims

1. a kind of archiving method characterized by comprising

For every facial image to be archived, according to the spatial information of the facial image to be archived, from facial image database from The real name archives of N priority are obtained in line archives；The offline profile of the facial image database includes at least a real name archives, The archives that the offline profile is filed before being the preset period of time, the real name archives are the shelves with identification information Case；The facial image to be archived does not find matched real name archives in the real name archives of preceding N-1 priority；N is greater than 1, Preceding N-1 priority is higher than N priority；

When there are real name archives matched with the facial image to be archived in the real name archives for determining the N priority, The facial image to be archived is included into the matched real name archives.

2. the method as described in claim 1, which is characterized in that further include:

It is not present and the matched real name archives of facial image to be archived in the real name archives for determining the N priority When, the real name archives of N+1 priority are obtained from the offline profile of the facial image database；

Determining in the real name archives of the N+1 priority whether there is and the matched real name shelves of facial image to be archived Case, the real name archives up to finding matched real name archives or each priority, which compare, to be finished.

3. method according to claim 2, which is characterized in that further include:

After determining that the real name archives of each priority compare, it is not present and the matched real name of facial image to be archived When archives, it is determined as the facial image to be archived not file facial image.

4. method as claimed in claim 3, which is characterized in that further include:

Facial image of not filing in sorted each classification is clustered, the multiple first pre- filing archives are generated；

The multiple first pre- filing archives are clustered, determine the second pre- filing archives；

For each second pre- filing archives, according to the spatial information of the described second pre- filing archives, from the facial image database Offline profile in obtain the non-real name archives of M priority；The offline profile of the facial image database is non-including at least one Real name archives, the non-real name archives are the archives without identification information；Described second pre- filing archives are at preceding M-1 Matched non-real name archives are not found in the non-real name archives of priority；M is greater than 1, and preceding M-1 priority is higher than M priority；

Exist in the non-real name archives for determining the M priority and the described second pre- filing matched non-real name archives of archives When, the described second pre- filing archives are included into the matched non-real name archives.

5. method as claimed in claim 4, which is characterized in that further include:

There is no file the matched non-real name shelves of archives in advance with described second in the non-real name archives for determining the M priority When case, the non-real name archives of M+1 priority are obtained from the offline profile of the facial image database；

It determines in the non-real name archives of the M+1 priority and files the matched non-real name of archives in advance with the presence or absence of with described second Archives, the non-real name archives up to finding matched non-real name archives or each priority, which compare, to be finished.

6. method as claimed in claim 5, which is characterized in that further include:

After determining that the non-real name archives of each priority compare, there is no matched non-with the described second pre- filing archives When real name archives, a new non-real name archives are established for the described second pre- filing archives.

7. a kind of filing apparatus characterized by comprising

Module is obtained, for obtaining multiple facial images to be archived in preset period of time and multiple facial images to be archived Spatial information；

Processing module, for being directed to every facial image to be archived, according to the spatial information of the facial image to be archived, from people The real name archives of N priority are obtained in the offline profile in face image library；The offline profile of the facial image database includes at least One real name archives, the archives that the offline profile is filed before being the preset period of time, the real name archives are with identity The archives of identification information；The facial image to be archived does not find matched real name in the real name archives of preceding N-1 priority Archives；N is greater than 1, and preceding N-1 priority is higher than N priority；

Profiling module is matched for existing in the real name archives for determining the N priority with the facial image to be archived Real name archives when, the facial image to be archived is included into the matched real name archives.

8. device as claimed in claim 7, which is characterized in that the processing module is also used to:

9. device as claimed in claim 8, which is characterized in that the processing module is also used to:

10. device as claimed in claim 9, which is characterized in that further include:

Categorization module, for being divided the facial image of not filing according to the spatial information for not filing facial image Class；

It is pre- to generate multiple first for clustering the facial image of not filing in sorted each classification for cluster module File archives；The multiple first pre- filing archives are clustered, determine the second pre- filing archives；

The processing module is also used to be believed for each second pre- filing archives according to the space of the described second pre- filing archives Breath, obtains the non-real name archives of M priority from the offline profile of the facial image database；The facial image database it is offline Archives include at least a non-real name archives, and the non-real name archives are the archives without identification information；Described second Pre- filing archives do not find matched non-real name archives in the non-real name archives of preceding M-1 priority；M is greater than 1, and preceding M-1 is a Priority is higher than M priority；

The profiling module is also used to exist in the non-real name archives for determining the M priority and the described second pre- filing When the matched non-real name archives of archives, the described second pre- filing archives are included into the matched non-real name archives.

11. device as claimed in claim 10, which is characterized in that the processing module is also used to:

12. device as claimed in claim 11, which is characterized in that the processing module is also used to:

13. a kind of archival device, which is characterized in that including at least one processing unit and at least one storage unit, In, the storage unit is stored with computer program, when described program is executed by the processing unit, so that the processing is single First perform claim requires the step of 1~6 any claim the method.

14. a kind of computer-readable medium, which is characterized in that it is stored with the computer program that can be executed by archival device, works as institute Program is stated when running in archival device, so that the step of archival device perform claim requires 1~6 any the method.