CN112183490A - Face snapshot picture filing method and device - Google Patents
Face snapshot picture filing method and device Download PDFInfo
- Publication number
- CN112183490A CN112183490A CN202011217645.2A CN202011217645A CN112183490A CN 112183490 A CN112183490 A CN 112183490A CN 202011217645 A CN202011217645 A CN 202011217645A CN 112183490 A CN112183490 A CN 112183490A
- Authority
- CN
- China
- Prior art keywords
- face
- picture
- temporary
- archive
- filing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000004364 calculation method Methods 0.000 claims description 29
- 238000001514 detection method Methods 0.000 claims description 13
- 238000001914 filtration Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 description 14
- 238000004590 computer program Methods 0.000 description 10
- 238000012545 processing Methods 0.000 description 8
- 230000001815 facial effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000004140 cleaning Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000013441 quality evaluation Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000007667 floating Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention provides a face snapshot picture filing method and a face snapshot picture filing device, wherein the method comprises the following steps: acquiring a snapshot batch of face snapshot picture sets; according to the picture similarity, filing all face snap-shot pictures in the batch of face snap-shot picture sets to obtain at least one temporary file; calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library aiming at each temporary archive; if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file; and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing file is smaller than the filing similarity threshold, newly building the filing file in the filing file library, and classifying the temporary file into the newly built filing file. The archiving efficiency can be improved.
Description
Technical Field
The invention relates to the field of image processing, in particular to a face snapshot picture archiving method and device.
Background
With the rapid development of the face recognition technology, the application scenes of the face recognition technology are greatly widened, and particularly the field of security protection. In both domestic and foreign countries, public security departments in the security field construct security business systems with face recognition as the core by laying a large number of video monitoring terminals, and play a vital role in ensuring the security of the departments. However, with the continuous operation of the security service system, the captured pictures are accumulated continuously, billions or even billions of face captured pictures are accumulated, and the face captured pictures are stored in the database and become sleep data, so that the application value of the data is low.
In order to improve the application value of the accumulated face snapshot pictures, the repeated snapshot pictures of the same person are filed, the hierarchical relationship of 'archive-face snapshot pictures' is formed by filing, and data retrieval and application expansion are carried out based on the archive, so that the application value of the picture data can be improved.
However, in the current face snapshot picture filing method, the face characteristics of each captured face snapshot picture need to be compared with the face characteristics of each filed file in the existing face snapshot picture archive library respectively to determine the filed file to which the captured face snapshot picture belongs, so that the face snapshot picture is filed, a large amount of computing resources need to be consumed, the time required for filing is long, and the filing efficiency is low.
Disclosure of Invention
In view of the above, the present invention provides a method and an apparatus for archiving a face snapshot to improve the archiving efficiency.
In a first aspect, an embodiment of the present invention provides a face snapshot picture archiving method, including:
acquiring a snapshot batch of face snapshot picture sets;
filing each face snapshot picture in the batch of face snapshot picture sets according to the picture similarity to obtain at least one temporary file;
calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library aiming at each temporary archive;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
In a second aspect, an embodiment of the present invention further provides a face snapshot image archiving device, including:
the image set acquisition module is used for acquiring a snapshot batch of face snapshot image sets;
the temporary file generation module is used for archiving each face snapshot picture in the batch of face snapshot picture set according to the picture similarity to obtain at least one temporary file;
an archive module to, for each temporary archive:
calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
In a third aspect, an embodiment of the present application provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the steps of the above method when executing the computer program.
In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, performs the steps of the method described above.
The face snapshot picture archiving method and device provided by the embodiment of the invention have the advantages that a snapshot batch of face snapshot picture sets are obtained; filing each face snapshot picture in the batch of face snapshot picture sets according to the picture similarity to obtain at least one temporary file; for each temporary file: calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library; if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file; and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive. Therefore, after the snapshot batch of face snapshot picture sets are obtained, the batch of face snapshot picture sets are filed to form a plurality of temporary files, face feature similarity calculation is respectively carried out on the temporary reference pictures and the filing reference pictures of each filing file from each temporary file to determine the filing files to which the temporary files belong, and when the temporary files are filed, the face features of the temporary reference pictures are respectively compared with the filing reference pictures of the filing files, so that the data processing amount during filing is effectively reduced, the calculation resources are reduced, and the filing efficiency is improved.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 shows a schematic flow diagram of a face snapshot picture archiving method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a face snapshot picture archiving device according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a computer device 300 according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
In the prior art, when a face snapshot picture is filed, the face characteristic comparison between the face snapshot picture and each filed file in an existing face snapshot picture file library is required, so as to determine the filed file to which the face snapshot picture belongs. In the embodiment of the invention, the quality of the captured face snapshot picture is considered to be uneven, and the face snapshot picture with poor quality has no application value even if the face snapshot picture is archived, so that after a first number of images within a certain period of time of snapshot are obtained, the face snapshot pictures which do not meet the preset quality requirement in the first number of images are filtered, the data processing amount and the required computing resource for subsequent archiving can be effectively reduced, the filtered face snapshot pictures are subjected to preliminary archiving to form a plurality of temporary files, each temporary file is subjected to face feature comparison (similarity calculation) according to the temporary reference picture and the face snapshot picture library of each archive file in the existing archive files, therefore, the filing files to which the temporary files belong are determined, one face snapshot picture is selected from the plurality of face snapshot pictures of the temporary files, and the face characteristics of the face snapshot picture are compared with the filing reference pictures of the filing files respectively, so that the data processing amount during filing can be further reduced, the calculation resources are effectively reduced, the time required for filing is greatly shortened, a large number of face snapshot pictures can be efficiently and accurately filed, and the cost for filing, constructing and operating the face snapshot pictures is reduced.
The embodiment of the invention provides a face snapshot picture filing method and device, which are described below through an embodiment.
Fig. 1 shows a schematic flow diagram of a face snapshot picture archiving method provided by an embodiment of the present invention. As shown in fig. 1, the method includes:
in the embodiment of the present invention, as an optional embodiment, a human face system product and a monitoring device for monitoring a safe city detect and capture a human face, obtain a human face captured picture, and transmit the human face captured picture to a background server, where the background server processes the received human face captured picture according to a preset time period, for example, every four hours, every day, every week, etc., and the human face captured pictures received within the preset time period form a batch of human face captured picture sets.
For example, acquiring a snapshot batch of face snapshot photo collections may include:
acquiring a first number of images acquired by a multi-path image acquisition device within a preset time period;
and carrying out face detection on the first number of images to obtain a second number of face images, and forming the batch of face snapshot image sets.
Wherein each image of the first number of images may include information of at least one of: the location of the image capturing device (e.g., longitude, latitude), the Identification (ID) of the image capturing device, the time of capture, etc.
In one implementation, quality judgment can be performed before and/or after face detection is performed, so that data cleaning is performed based on the result of the quality judgment to ensure the quality of pictures in the batch of face snapshot picture sets in S101.
In one embodiment, a first number of images may be face detected, resulting in an N1 number of face images, and then a second number of face images may be obtained by selecting face images from the N1 number of face images in which the face image quality is above a first quality threshold.
It is to be understood that the images acquired by the image acquisition device do not necessarily have a portrait, that is, the partial images in the first number of images are only the environmental images, and therefore the partial environmental images in which no portrait exists can be deleted based on the face detection. Or, optionally, in the image acquired by the image acquisition device, a plurality of faces may exist in a certain image, and then a plurality of face images corresponding to the plurality of faces one to one may be obtained based on the image through face detection. It can be seen that N1 does not have a necessarily magnitude relationship with the first quantity.
When the facial image quality judgment is performed on N1 facial images, one or more of the attributes (such as resolution) of the image acquisition device, the environment (such as brightness) in which the image acquisition device is located, the relationship (such as distance, direction, angle, and the like) between the acquired face and the image acquisition device, the shielding condition of the face, and the like may be considered to comprehensively obtain the facial image quality.
Optionally, after obtaining the first number of images acquired by the multi-path image acquisition device, because the face captured by different devices snap-shot images, or the face captured by the same device at different time periods, the image quality is uneven, and for the face captured by the poor quality, for example, the face captured by the captured face is a side face, even if the face captured image is filed, because the face feature information that can be provided by the filed face captured image is very small, the face captured image has no application value, therefore, optionally, the quality judgment can be performed first to perform data cleaning, and a face captured image set is formed according to the images acquired after the data cleaning. Thus, performing a quality determination on the N1 number of images results in a second number of face images, including:
and according to a preset image quality grading strategy, calculating the snapshot quality grading of each face snapshot image in the N1 number of images, and filtering the face snapshot images with the snapshot quality grading smaller than a preset snapshot quality grading threshold value to obtain a second number of face images.
In the embodiment of the invention, the batch of face snapshot pictures collectively comprise a second number of face images. The picture quality scoring policy may be set according to actual needs, and as an optional embodiment, the picture quality scoring policy includes: the face angle scoring sub-strategy, the pitch angle scoring sub-strategy, the rotation angle scoring sub-strategy and the deflection angle scoring sub-strategy are photographed, so that the snapshot quality scoring of each face snapshot picture in the N1 number of images is calculated, and the method comprises the following steps:
according to the shot face angle scoring strategy and the face angle of the face snapshot picture to be scored, obtaining the face angle score of the face snapshot picture to be scored;
obtaining a pitch angle score of the face snapshot picture to be scored according to a pitch angle scoring strategy and the pitch angle of the face snapshot picture to be scored;
obtaining a rotation angle score of the face snapshot picture to be scored according to a rotation angle scoring strategy and a rotation angle of the face snapshot picture to be scored;
according to the deflection angle evaluation molecular strategy and the deflection angle of the face snapshot picture to be scored, obtaining the deflection angle score of the face snapshot picture to be scored;
and carrying out weighted summation on the face angle score, the pitch angle score, the rotation angle score and the deflection angle score to obtain the snapshot quality score of the face snapshot picture to be scored.
In the embodiment of the present invention, as another optional embodiment, a face occlusion evaluation policy and the like may also be set, which is not limited in the embodiment of the present invention.
In the embodiment of the present invention, in order to make the snapshot quality scores have a uniform scoring scale and avoid the snapshot quality score deviation caused by inconsistent scoring scales, as an optional embodiment, the face angle scoring sub-strategy, the pitch angle scoring sub-strategy, the rotation angle scoring sub-strategy, and the deflection angle scoring sub-strategy have the same scoring levels and scoring scores, for example, five scoring levels, which are respectively a first scoring level to a fifth scoring level, where the scoring score corresponding to the first scoring level is 100, the scoring score corresponding to the second scoring level is 80, the scoring score corresponding to the third scoring level is 60, the scoring score corresponding to the fourth scoring level is 30, and the scoring score corresponding to the fifth scoring level is 0.
In the embodiment of the present invention, as an optional embodiment, each sub-policy has the same weighting coefficient, as another optional embodiment, the weighting coefficients of each sub-policy may also be different, and the weighting coefficients of each sub-policy may be set according to actual statistics and analysis. The sum of the weighting coefficients of the sub-strategies equals 1.
In the embodiment of the invention, the face snapshot pictures with lower snapshot quality scores can be effectively avoided by filtering the first number of images, and the situation that the accuracy of the generated temporary file is lower due to the fact that the similarity is close when the similarity of the subsequent pictures is calculated is avoided, so that the generation precision of the temporary file is improved.
In another embodiment, the image quality of each image in the first number of images may be obtained and the images in which the image quality is below the second quality threshold may be culled, resulting in remaining images with image quality equal to or above the second quality threshold, assuming a number N2. After that, face detection is performed on the N2 number of images, resulting in a second number of face images.
The image quality may be related to factors such as resolution, brightness, and blur degree of the image, and one or more of the factors may be combined to obtain an image quality score.
In another embodiment, a first number of images acquired by the multi-path image acquisition device may be acquired, and a second number of facial images may be acquired by retaining N2 number of images in which the image quality is equal to or higher than the second quality threshold, and then obtaining N3 number of images through facial detection, and then retaining facial images in which the facial image quality is higher than the first quality threshold.
Therefore, the quality of each face image of the second number of face images contained in the batch of face snapshot picture sets is ensured by considering not only the image quality but also the quality of the face collected in the images. The human face detection method is characterized in that human face detection is carried out on images with the number of N2, the image number base number for carrying out human face detection can be reduced, and the human face detection algorithm is generally an algorithm based on deep learning, so that the time length for carrying out human face detection can be shortened and the processing efficiency is improved by reducing the number to N2. The image quality of each face image in the second number of face images is higher than the first quality threshold, so that the quality of all face images after filing can be ensured. The accuracy of subsequent operations such as retrieval or face recognition based on the archived files can be conveniently ensured.
In the embodiment of the present invention, after filtering the first number of images according to the image quality scoring policy to obtain the second number of face images, the method may further include:
and carrying out mutually exclusive picture attribute marking on the face snapshot pictures in the second number of face images.
In this embodiment of the present invention, as an optional embodiment, the mutually exclusive picture attributes include, but are not limited to: the male and female, children and adults can set labels for the face snapshot pictures through the picture attribute marks, and follow-up retrieval and application are facilitated.
102, archiving each face snapshot picture in the batch of face snapshot picture sets according to picture similarity to obtain at least one temporary file;
in this embodiment, as an optional embodiment, archiving, according to the picture similarity, each face snapshot picture in the batch of face snapshot picture sets to obtain at least one temporary file includes:
extracting a first face snapshot picture from the batch of face snapshot picture set to obtain a first temporary file, wherein a temporary reference picture in the first temporary file is the first face snapshot picture;
extracting face snapshot pictures except the first face snapshot picture one by one from the batch of face snapshot picture set, and calculating the similarity between the extracted face snapshot pictures and temporary reference pictures in a first temporary file;
if the similarity is larger than or equal to the similarity threshold, adding the newly extracted face snapshot picture into a first temporary file;
and updating the temporary reference picture in the first temporary archive.
Specifically, the face snapshot pictures may be extracted one by one from the batch of face snapshot picture sets, and the temporary archives are gradually updated, and one of the temporary archives is marked as a temporary reference picture. Alternatively, the order of extraction one by one may be in time, e.g. extracting the earliest captured picture first.
And when any face snapshot picture is added into the temporary file, a comparison score is marked, and the comparison score represents the similarity between the face snapshot picture and the temporary reference picture when the face snapshot picture is added.
Initially, there may be only one picture in the temporary archive, and then this picture is the temporary reference picture, denoted as LJ. Then, one face snapshot is extracted from the batch of face snapshot picture sets, the similarity (assumed as score1) between the extracted face snapshot and the current LJ in the temporary archive is calculated, if the similarity is greater than or equal to a similarity threshold, the newly extracted face snapshot is added into the temporary archive, and the newly added face snapshot is marked with a comparison score 1. At this time, there are two pictures in the temporary file. Further, the picture quality of the two pictures may be calculated, and the one with the higher picture quality may be taken as LJ.
Similarly, every time a face snapshot is extracted, the similarity between the newly extracted face snapshot and the current LJ in the temporary archive may be calculated, and if the similarity is greater than or equal to the similarity threshold, the newly extracted face snapshot may be added to the temporary archive, so that the number of pictures in the temporary archive is increased by 1. Optionally, after the joining, it may be further determined whether to update LJ.
The updating may be performed based on the number of currently existing pictures in the temporary archive (updating the temporary reference picture in the first temporary archive):
and A, if the number is less than or equal to a preset number first threshold value, taking the maximum quality in all the pictures of the number as LJ.
Specifically, the preset number first threshold may be 10. And if the number of the face snapshot pictures in the first temporary file is less than or equal to a preset number first threshold, selecting the face snapshot picture with the highest quality from all the face snapshot pictures in the first temporary file as the temporary reference picture.
B, if the number is larger than a preset number first threshold value:
B1. if the quality of the current LJ is less than the preset quality score threshold, then the quality in the most recent number second threshold (all number of pictures if the number is less than the number second threshold) is taken as LJ.
Specifically, if the quality of the current temporary reference picture is less than a preset quality score threshold, the current temporary reference picture is replaced by the face snapshot picture with the maximum quality in the face snapshot pictures with the latest number of second thresholds. The number first threshold is 50 and the number second threshold is 100.
B2. If the number of the pictures added after the current LJ is determined to be larger than or equal to the second threshold value, or if the time length between the current time and the time when the current LJ is determined is larger than a set time length (for example, one month), the face snapshot pictures with the quality larger than the quality score threshold value and the highest score in the latest preset comprehensive comparison threshold value are taken as the LJ.
Specifically, after the current temporary reference picture is replaced, if the number of the newly added face snapshot pictures is greater than or equal to a second threshold, or if the time length between the current time and the time when the current temporary reference picture is replaced is greater than a set time length, the face snapshot picture with the quality greater than the quality score threshold value and the highest calculated score in the face snapshot pictures with the latest preset comprehensive comparison threshold value is replaced by the current temporary reference picture. The overall alignment threshold was 200.
Calculate score (current time-snapshot time)/3600/24/365 x compare score. The unit of the current time and the snapshot time is seconds(s).
In the embodiment of the invention, when the number of face snapshot pictures in the temporary file is large, a part of face snapshot pictures are selected to be subjected to score calculation according to the snapshot time and the comparison scores, so that the face snapshot pictures with large time span and high comparison scores can be effectively selected as temporary reference pictures, when the face snapshot pictures are compared with the filing reference pictures, the same person can generate large face change in a time range with large time span to cause filing errors, and the calculation resources required by calculation can be effectively reduced.
It can be understood that if the similarity is smaller than the similarity threshold, the similarity with LJ in other temporary files is calculated again, and the temporary file to be added is determined. If none, then a temporary file is created.
Alternatively, assuming that there are already M1 temporary archives, the similarity of the extracted face snapshot picture to the M1 LJ of the M1 temporary archives can be calculated directly and synchronously to determine more quickly which temporary archive to add to.
Thus, the application can establish a temporary archive for later archiving operations. Because the temporary archive is based on the batch face snapshot picture set, the number base number of the processed pictures is small, and the temporary archive is fast and efficient to establish. In addition, in the process of establishing the temporary archives, the temporary archives are updated in a one-by-one extraction mode, and the calculation amount is small. Any temporary file takes one of the images as a temporary reference picture, on one hand, whether the subsequently extracted snapshot picture belongs to the temporary file or not is judged conveniently, the judgment mode of whether the snapshot picture is subjected to the temporary file or not is simple, comparison with all pictures in the temporary file is not needed, the calculated amount is reduced, and the establishment speed of the temporary file is accelerated. On the other hand, the temporary reference picture can be updated based on the number of the existing pictures in the temporary archive, so that the reference value of the temporary reference picture is ensured, and the misjudgment of the recent picture caused by the low quality of the temporary reference picture is prevented. On the other hand, the temporary reference picture can be further used for judging which archive file belongs to in the following step S103, and the comparison between pictures one by one is not needed, so that the archiving efficiency is improved.
103, calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive repository aiming at each temporary archive;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
In the embodiment of the present invention, the temporary archive needs to be merged into the stored archive. As an alternative embodiment, the similarity threshold may be the same as or different from the archive similarity threshold.
In the embodiment of the invention, the determination method of the filing reference picture is the same as the determination method of the temporary reference picture. It is to be understood that after the temporary archive is included in the first archive, the archive reference picture of the first archive may be updated, and the specific updating manner is the same as the above-mentioned manner for updating the temporary reference picture of the temporary archive in S102, for example, the update may be performed with reference to the number of pictures in the first archive and/or the picture quality and/or time of the current archive reference picture, and the details are not described herein again.
In the embodiment of the invention, when the face feature similarity of the temporary reference picture and the filing reference picture of each filing file is calculated, a plurality of face feature similarities which are more than or equal to the filing similarity threshold value and are obtained through calculation may exist. Thus, as an alternative embodiment, the step of including the temporary archive in the first archive comprises:
if the number of the face feature similarities which are larger than or equal to the archiving similarity threshold value is one, the temporary archive is classified into an archiving archive corresponding to the face feature similarity;
and if the number of the face feature similarities which are more than or equal to the filing similarity threshold value and are obtained through calculation is more than one, merging the filing files corresponding to the face feature similarities obtained through calculation, and classifying the temporary files into the merged filing files.
In the embodiment of the present invention, after the temporary archive is classified into the archive file corresponding to the face feature similarity obtained through calculation, the archive reference picture of the classified archive file may be updated according to the picture comprehensive quality evaluation policy.
In the embodiment of the present invention, as the number of face-captured pictures and newly-built archive files in the archive file increases, the requirement for the storage capacity required by the archive file repository increases, and therefore, for an archive file into which no face-captured picture is included for a long time, it indicates that the probability of being able to be captured is extremely low, and corresponding deletion processing can be performed to save the storage space, as an optional embodiment, the method further includes:
acquiring a time stamp of a face snapshot picture which is recently put into each archive in an archive repository;
and calculating the difference value between the current timestamp and the acquired timestamp, and deleting the archive file if the difference value is greater than a preset time span threshold value.
In the embodiment of the invention, the snapshot frequency statistics can be carried out on each face snapshot picture in the archived file, if the snapshot frequency is less than the preset frequency threshold, the latest preset number of face snapshot pictures can be stored, and other face snapshot pictures can be deleted.
In this embodiment of the present invention, as another optional embodiment, the archive file may be further processed according to file activity, and the method further includes:
aiming at each archive file in the archive file library, calculating the file activity of the archive file according to a preset file activity calculation strategy;
and if the calculated file activity is less than the preset activity threshold value, deleting the archive file.
In the embodiment of the present invention, as an optional embodiment, the file liveness is calculated by using the following formula:
xi is the file liveness;
α is a difference between the year of the face snapshot picture captured latest and 2000, for example, if the year of the face snapshot picture captured latest is 2018, α ═ 2018-.
Beta is the month of the face snapshot picture which is snapshot latest, and the value is 1-12;
and gamma is the date of the latest face snapshot picture and is 1-31.
In the embodiment of the invention, the file liveness is determined according to the time of the latest (latest) snapshot picture of the face.
In the embodiment of the invention, the file activity is calculated by utilizing the file activity calculation strategy, so that the activity of the archived files can be distinguished, the archive file state can be rapidly identified, corresponding application is carried out according to the file activity of the archived files, the archiving efficiency of face snapshot pictures is improved, and the occupation of calculation resources and storage resources is favorably reduced.
In the embodiment of the invention, after the face snapshot picture is filed, in order to promote the application of the filed file, the cross collision can be carried out with the population information base so as to determine the identity of the person corresponding to the filed file, so that the filed file can be close to the business, the application of the filed file is further expanded, and the identity authentication of the face snapshot picture is realized. Thus, as an alternative embodiment, the method further comprises:
and aiming at each archive file, carrying out similarity calculation on the archive reference picture in the archive file and the identity picture in a preset population information base, storing the identity picture with the highest similarity into the archive file from the similarity which is obtained by calculation and is greater than or equal to a similarity threshold, and identifying the archive file based on the identity picture with the highest similarity.
In the embodiment of the invention, as an optional embodiment, the population information base is a city temporary population information base.
In the embodiment of the invention, for the archive files for identity identification, identification can be carried out on each face snapshot picture through the database table so as to record the detailed archive information of the face snapshot picture, so that corresponding analysis can be conveniently carried out in subsequent application according to the detailed archive information. Thus, as another alternative embodiment, the method further comprises:
and setting an identity database table for the archive file based on the identity picture identification with the highest similarity, wherein the identity database table comprises an identity master table and an identity slave table, the identity master table is used for marking the personnel information corresponding to the archive file, and the identity slave table is used for marking the archive details.
In the embodiment of the present invention, as an optional embodiment, the person information includes but is not limited to: identity card (idCardNo), liveness (active), latest filing time (timeStamp), latest snapshot time (recentTime), snapshot frequency (frequency), snapshot type (type), etc.; archival details include, but are not limited to: picture identification (id), snapshot identification (captureID), identity card (idCardNo), picture quality score (quality), archive time (onHold _ time), snapshot type (type), snapshot time (capture _ time), zimgnode, snapshot picture summary (picture _ MD5), face picture summary (face _ image _ MD5), face picture uniform resource locator (imageUrl1) and snapshot picture uniform resource locator (imageUrl 2).
In the embodiment of the invention, a filing file corresponds to personnel information, and each face snapshot picture in the filing file corresponds to a filing detail.
In the embodiment of the present invention, in the similarity calculation according to the city temporary population information base, since the population mobility is large, after the similarity calculation of the identity authentication is performed, there is no similarity greater than or equal to the similarity threshold in the calculated similarities, and for the part of the files, the part of the files should be floating population files (non-real name files), the method further includes:
if the similarity obtained by calculation based on the archive file does not have the similarity larger than or equal to the similarity threshold, a temporary database table is set for the archive file, wherein the temporary database table comprises a temporary main table, a similarity table and a temporary slave table, the temporary main table is used for marking the flow staff information corresponding to the archive file, the temporary slave table is used for marking the archive details, and the similarity table is used for marking the identity pictures in the staff information base similar to the archive file.
In the embodiment of the present invention, the floating personnel information of the temporary master table includes but is not limited to: the method comprises the following steps of filing identification (id), activity (active), recent filing time (timeStamp), filing reference picture (master), auxiliary filing reference picture (master2), recent snapshot time (recentTime), snapshot frequency (frequency), earliest snapshot time (latestTime) and the like, wherein the filing identification is a randomly generated mark for uniquely identifying the filing archive.
Archival details include, but are not limited to: picture identification (id), snapshot identification (capture _ ID), feature identification (old _ feature ID), picture quality score (quality), filing time (onHold _ time), snapshot time (capture _ time), new feature identification (featureID), feature library (db _ number), zimgnode, snapshot picture summary (picture _ MD5), face picture summary (face _ image _ MD5), face picture uniform resource locator (imageUrl1), snapshot picture uniform resource locator (imageUrl2), snapshot camera identification (camera Id), snapshot address (capture _ location), camera longitude (capture _ lon), and camera latitude (capture _ lature).
Similar archives include, but are not limited to: archive identification (id), capture time (capture _ time), zimgnode, snapshot picture summary (picture _ MD5), face picture summary (face _ image _ MD5), face picture uniform resource locator (imageUrl1), snapshot picture uniform resource locator (imageUrl2), snapshot identification (capture id), feature identification (old _ feature id), snapshot camera identification (camera id), capture address (capture _ location), camera longitude (capture _ location), camera latitude (capture _ lat), similarity (similarity _ score), and archive identification (picture onholeid).
In the embodiment of the invention, the filing file is set in a database table mode, and the personnel information in the filing file is separated from the details of the file; as an optional embodiment, the archived files can be sorted according to mutually exclusive characteristics of the personnel information, such as gender and age, so that frequent data retrieval with large record volume can be avoided, and the access efficiency of the archived files is improved.
The face snapshot picture archiving method provided by the embodiment of the invention can efficiently and accurately archive massive face snapshot pictures, and simultaneously can greatly reduce required resources and reduce construction and operation costs. By utilizing the method of the embodiment of the invention, the following performance indexes can be basically achieved: 8-10 mainstream 2U universal servers (CPU: Intel 2.xGHz 10core 2, memory: 64GB, storage: 4TB) can support 1000 routes of face bayonet monitoring in medium cities with millions of grades of population and daily quasi-real-time filing of face snapshot pictures with the snapshot amount of about 3000 faces per day, and can achieve the face clustering recall rate of more than or equal to 98% and the face clustering accuracy rate of more than or equal to 99%.
In this embodiment of the present invention, as an optional embodiment, the method further includes:
receiving a target face picture, respectively carrying out similarity calculation with each filing reference picture in the filing files, obtaining the filing file with the highest similarity, and outputting the filing file with the highest similarity as a query result of the target face picture.
In the embodiment of the present invention, as an optional embodiment, the images may be sorted according to similarity, and the top n archive files with the similarity rank are output as the query result of the target face image, where n is a natural number. For example, archive files with similarity greater than 90 points are output.
It can be seen from the above embodiments that at least one temporary archive can be established first in the present application, and then it is determined which existing archive the temporary archive belongs to. In addition, the temporary archive is established based on the batch face snapshot picture set, and the picture set is acquired in a preset time period and comprises a limited number of pictures, so that the time consumed for establishing the temporary archive is short, and the archiving speed is further improved.
Fig. 2 shows a schematic structural diagram of a face snapshot picture archiving device according to an embodiment of the present invention. As shown in fig. 2, the apparatus includes:
the image set acquisition module 201 is used for acquiring a snapshot image set of face snapshots in a snapshot batch;
a temporary file generation module 202, configured to archive each face snapshot picture in the batch of face snapshot picture sets according to picture similarity, so as to obtain at least one temporary file;
in this embodiment of the present invention, as an optional embodiment, the temporary archive generating module 202 is specifically configured to:
extracting a first face snapshot picture from the batch of face snapshot picture set to obtain a first temporary file, wherein a temporary reference picture in the first temporary file is the first face snapshot picture;
extracting face snapshot pictures except the first face snapshot picture one by one from the batch of face snapshot picture set, and calculating the similarity between the extracted face snapshot pictures and temporary reference pictures in a first temporary file;
if the similarity is larger than or equal to the similarity threshold, adding the newly extracted face snapshot picture into a first temporary file;
and updating the temporary reference picture in the first temporary archive.
Wherein updating the temporary reference picture in the first temporary archive comprises:
if the number of the face snapshot pictures in the first temporary file is smaller than or equal to a preset number first threshold, selecting the face snapshot picture with the highest quality from all the face snapshot pictures in the first temporary file as the temporary reference picture;
if the number of the face snapshot pictures in the first temporary file is larger than a preset number first threshold value:
and if the quality of the current temporary reference picture is less than a preset quality score threshold, replacing the current temporary reference picture with the face snapshot picture with the maximum quality in the face snapshot pictures with the latest number of second thresholds.
An archive module 203 for, for each temporary archive:
calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
In this embodiment of the present invention, as an optional embodiment, the archiving module 203 is specifically configured to:
counting the number of face snapshot pictures in the temporary file;
if the number of the face snapshot pictures is larger than or equal to a preset number threshold value and the face snapshot pictures with the snapshot quality scores larger than a quality score threshold value exist, selecting the face snapshot pictures with the preset comprehensive comparison threshold value and the snapshot quality scores larger than the quality score threshold value according to a time sequence, calculating the picture comprehensive quality scores of the selected face snapshot pictures according to a preset picture comprehensive quality evaluation strategy, selecting the face snapshot pictures with the highest picture comprehensive quality scores, and obtaining a filing reference picture;
if the number of the face snapshot pictures is smaller than a preset number threshold value or no face snapshot pictures with snapshot quality scores larger than a quality score threshold value exist, calculating the picture comprehensive quality scores of all the face snapshot pictures in the temporary file according to a preset picture comprehensive quality evaluation strategy, and selecting the face snapshot picture with the highest picture comprehensive quality score to obtain the filing reference picture.
In this embodiment of the present invention, as another optional embodiment, the archiving module 203 is further configured to:
if the number of the face feature similarities which are larger than or equal to the archiving similarity threshold value is one, the temporary archive is classified into an archiving archive corresponding to the face feature similarity;
and if the number of the face feature similarities which are more than or equal to the filing similarity threshold value and are obtained through calculation is more than one, merging the filing files corresponding to the face feature similarities obtained through calculation, and classifying the temporary files into the merged filing files.
In this embodiment of the present invention, as an optional embodiment, the apparatus further includes:
and the picture filtering module (not shown in the figure) is used for calculating the snapshot quality score of each face snapshot picture in the first number of pictures according to a preset picture quality scoring strategy, and filtering the face snapshot pictures with the snapshot quality scores smaller than a preset snapshot quality scoring threshold value to obtain a second number of face pictures.
In this embodiment of the present invention, as an optional embodiment, the picture quality scoring policy includes: the face angle of shooing assesses sub-strategy, pitch angle sub-strategy, rotation angle sub-strategy, deflection angle sub-strategy, and the picture filtering module is specifically used for:
according to the shot face angle scoring strategy and the face angle of the face snapshot picture to be scored, obtaining the face angle score of the face snapshot picture to be scored;
obtaining a pitch angle score of the face snapshot picture to be scored according to a pitch angle scoring strategy and the pitch angle of the face snapshot picture to be scored;
obtaining a rotation angle score of the face snapshot picture to be scored according to a rotation angle scoring strategy and a rotation angle of the face snapshot picture to be scored;
according to the deflection angle evaluation molecular strategy and the deflection angle of the face snapshot picture to be scored, obtaining the deflection angle score of the face snapshot picture to be scored;
and carrying out weighted summation on the face angle score, the pitch angle score, the rotation angle score and the deflection angle score to obtain the snapshot quality score of the face snapshot picture to be scored.
In this embodiment, as another optional embodiment, the apparatus further includes:
the archive processing module is used for calculating the archive activity of each archive in the archive repository according to a preset archive activity calculation strategy; and if the calculated file activity is less than the preset activity threshold value, deleting the archive file.
In this embodiment, as a further optional embodiment, the apparatus further includes:
and the identity identification module is used for calculating the similarity between the filing reference picture in the filing file and the identity picture in a preset population information base aiming at each filing file, storing the identity picture with the highest similarity into the filing file from the calculated similarity which is more than or equal to the similarity threshold, and identifying the filing file based on the identity picture with the highest similarity.
In this embodiment, as a further optional embodiment, the apparatus further includes:
and the identity database table setting module is used for setting an identity database table for the archive files based on the identity picture identifiers with the highest similarity, wherein the identity database table comprises an identity master table and an identity slave table, the identity master table is used for marking the personnel information corresponding to the archive files, and the identity slave table is used for marking the archive details.
In the embodiment of the invention, a filing file corresponds to personnel information, and each face snapshot picture in the filing file corresponds to a filing detail.
In this embodiment, as a further optional embodiment, the apparatus further includes:
and the temporary database table setting module is used for setting a temporary database table for the archive file if the similarity obtained by calculation based on the archive file does not have the similarity larger than or equal to a similarity threshold, wherein the temporary database table comprises a temporary main table, a similarity table and a temporary slave table, the temporary main table is used for marking the corresponding mobile personnel information of the archive file, the temporary slave table is used for marking the archive details, and the similarity table is used for marking the identity pictures in the personnel information base similar to the archive file.
In this embodiment, as a further optional embodiment, the apparatus further includes:
and the query module is used for receiving the target face picture, respectively carrying out similarity calculation with each filing reference picture in the filing archives to obtain the filing archives with the highest similarity, and outputting the filing archives with the highest similarity as the query result of the target face picture.
As shown in fig. 3, an embodiment of the present application provides a computer device 300, configured to execute the face snapshot image archiving method in fig. 1, where the device includes a memory 301, a processor 302, and a computer program (machine readable instructions) stored in the memory 301 and executable on the processor 302, where the processor 302 implements the steps of the face snapshot image archiving method when executing the computer program.
Specifically, the memory 301 and the processor 302 can be general-purpose memory and processor, and are not limited to specific examples, and when the processor 302 runs a computer program stored in the memory 301, the face snapshot picture archiving method can be executed.
Corresponding to the face snapshot image archiving method in fig. 1, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the face snapshot image archiving method are executed.
Specifically, the storage medium can be a general-purpose storage medium, such as a removable disk, a hard disk, or the like, and when a computer program on the storage medium is executed, the face snapshot image archiving method can be executed.
In the embodiments provided in the present application, it should be understood that the disclosed system and method may be implemented in other ways. The above-described system embodiments are merely illustrative, and for example, the division of the units is only one logical functional division, and there may be other divisions in actual implementation, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of systems or units through some communication interfaces, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments provided in the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus once an item is defined in one figure, it need not be further defined and explained in subsequent figures, and moreover, the terms "first", "second", "third", etc. are used merely to distinguish one description from another and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the present disclosure, which should be construed in light of the above teachings. Are intended to be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A face snapshot picture archiving method is characterized by comprising the following steps:
acquiring a snapshot batch of face snapshot picture sets;
filing each face snapshot picture in the batch of face snapshot picture sets according to the picture similarity to obtain at least one temporary file;
for each temporary file:
calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
2. The method of claim 1, wherein obtaining the set of snap-shot batches of face snap-shot pictures comprises:
acquiring a first number of images acquired by a multi-path image acquisition device within a preset time period;
and carrying out face detection on the first number of images to obtain a second number of face images, and forming the batch of face snapshot image sets.
3. The method of claim 2, wherein performing face detection on the first number of images to obtain a second number of face images comprises:
and calculating the snapshot quality score of each face snapshot picture in the first number of pictures according to a preset picture quality scoring strategy, and filtering the face snapshot pictures with the snapshot quality scores smaller than a preset snapshot quality scoring threshold value to obtain a second number of face pictures.
4. The method of claim 3, wherein the picture quality scoring policy comprises: the face angle scoring sub-strategy, the pitch angle scoring sub-strategy, the rotation angle scoring sub-strategy and the deflection angle scoring sub-strategy are photographed, and the snapshot quality score of each face snapshot picture in the first number of images is calculated, and the method comprises the following steps of:
according to the shot face angle scoring strategy and the face angle of the face snapshot picture to be scored, obtaining the face angle score of the face snapshot picture to be scored;
obtaining a pitch angle score of the face snapshot picture to be scored according to a pitch angle scoring strategy and the pitch angle of the face snapshot picture to be scored;
obtaining a rotation angle score of the face snapshot picture to be scored according to a rotation angle scoring strategy and a rotation angle of the face snapshot picture to be scored;
according to the deflection angle evaluation molecular strategy and the deflection angle of the face snapshot picture to be scored, obtaining the deflection angle score of the face snapshot picture to be scored;
and carrying out weighted summation on the face angle score, the pitch angle score, the rotation angle score and the deflection angle score to obtain the snapshot quality score of the face snapshot picture to be scored.
5. The method according to claim 1, wherein the archiving, according to the picture similarity, each face snapshot picture in the batch of face snapshot picture sets to obtain at least one temporary archive comprises:
extracting a first face snapshot picture from the batch of face snapshot picture set to obtain a first temporary file, wherein a temporary reference picture in the first temporary file is the first face snapshot picture;
extracting face snap-shot pictures except the first face snap-shot picture one by one from the batch of face snap-shot picture set, and calculating the similarity between the extracted face snap-shot pictures and temporary reference pictures in a first temporary file;
if the similarity is larger than or equal to the similarity threshold, adding the newly extracted face snapshot picture into a first temporary file;
and updating the temporary reference picture in the first temporary archive.
6. The method of claim 5, wherein the updating the temporary reference picture in the first temporary archive comprises:
if the number of the face snapshot pictures in the first temporary file is smaller than or equal to a preset number first threshold, selecting the face snapshot picture with the highest quality from all the face snapshot pictures in the first temporary file as the temporary reference picture;
if the number of the face snapshot pictures in the first temporary file is larger than a preset number first threshold value:
and if the quality of the current temporary reference picture is less than a preset quality score threshold, replacing the current temporary reference picture with the face snapshot picture with the maximum quality in the face snapshot pictures with the latest number of second thresholds.
7. The method of claim 6, wherein after the replacing the current temporary reference picture, the method further comprises:
and if the number of the newly added face snapshot pictures is greater than or equal to the second threshold, or if the time length between the current time and the time when the current temporary reference picture is replaced is greater than the set time length, replacing the current temporary reference picture with the face snapshot picture with the highest quality and the highest calculated score in the face snapshot pictures with the latest preset comprehensive comparison threshold.
8. The method according to any one of claims 1 to 7, further comprising:
for each archive file, carrying out similarity calculation on an archive reference picture in the archive file and identity pictures in a preset population information base, storing the identity picture with the highest similarity into the archive file from the similarity which is obtained by calculation and is greater than or equal to a similarity threshold, and identifying the archive file based on the identity picture with the highest similarity;
setting an identity database table for an archive file based on an identity picture identifier with the highest similarity, wherein the identity database table comprises an identity master table and an identity slave table, the identity master table is used for marking the personnel information corresponding to the archive file, and the identity slave table is used for marking the archive details;
if the similarity obtained by calculation based on the archive file does not have the similarity larger than or equal to the similarity threshold, a temporary database table is set for the archive file, wherein the temporary database table comprises a temporary main table, a similarity table and a temporary slave table, the temporary main table is used for marking the flow staff information corresponding to the archive file, the temporary slave table is used for marking the archive details, and the similarity table is used for marking the identity pictures in the staff information base similar to the archive file.
9. The utility model provides a picture filing device is taken a candid photograph to people's face which characterized in that includes:
the image set acquisition module is used for acquiring a snapshot batch of face snapshot image sets;
the temporary file generation module is used for archiving each face snapshot picture in the batch of face snapshot picture set according to the picture similarity to obtain at least one temporary file;
an archive module to, for each temporary archive:
calculating the face feature similarity between the temporary reference picture in the temporary archive and the pre-stored archive reference picture of each archive in the archive library;
if the face feature similarity between the temporary reference picture and the filing reference picture in the first filing file is larger than or equal to a preset filing similarity threshold value, the temporary file is classified into the first filing file;
and if the face feature similarity between the temporary reference picture and the filing reference picture in any filing archive is smaller than the filing similarity threshold, newly building a filing archive in the filing archive, and classifying the temporary archive into the newly-built filing archive.
10. A computer device, comprising: a processor, a memory and a bus, the memory storing machine readable instructions executable by the processor, the processor and the memory communicating via the bus when a computer device is running, the machine readable instructions when executed by the processor performing the steps of the face snapshot archiving method as recited in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011217645.2A CN112183490A (en) | 2020-11-04 | 2020-11-04 | Face snapshot picture filing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011217645.2A CN112183490A (en) | 2020-11-04 | 2020-11-04 | Face snapshot picture filing method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112183490A true CN112183490A (en) | 2021-01-05 |
Family
ID=73917068
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011217645.2A Pending CN112183490A (en) | 2020-11-04 | 2020-11-04 | Face snapshot picture filing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112183490A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113220942A (en) * | 2021-05-31 | 2021-08-06 | 杭州海康威视数字技术股份有限公司 | Method, device and equipment for searching associated object |
| CN117333926A (en) * | 2023-11-30 | 2024-01-02 | 深圳须弥云图空间科技有限公司 | Picture aggregation method and device, electronic equipment and readable storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105938552A (en) * | 2016-06-29 | 2016-09-14 | 北京旷视科技有限公司 | Face recognition method capable of realizing base image automatic update and face recognition device |
| CN109800318A (en) * | 2018-12-28 | 2019-05-24 | 上海依图网络科技有限公司 | A kind of archiving method and device |
| CN109800668A (en) * | 2018-12-28 | 2019-05-24 | 上海依图网络科技有限公司 | A kind of archiving method and device |
| CN110276277A (en) * | 2019-06-03 | 2019-09-24 | 罗普特科技集团股份有限公司 | Method and apparatus for detecting facial image |
| CN110287791A (en) * | 2019-05-23 | 2019-09-27 | 罗普特科技集团股份有限公司 | A kind of screening technique and system for face picture |
| CN110378314A (en) * | 2019-05-31 | 2019-10-25 | 北京金山云网络技术有限公司 | A kind of human face region image archiving method, device, electronic equipment and storage medium |
| CN111104825A (en) * | 2018-10-26 | 2020-05-05 | 北京陌陌信息技术有限公司 | Face registry updating method, device, equipment and medium |
| CN111126208A (en) * | 2019-12-11 | 2020-05-08 | 智慧眼科技股份有限公司 | Pedestrian archiving method and device, computer equipment and storage medium |
-
2020
- 2020-11-04 CN CN202011217645.2A patent/CN112183490A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105938552A (en) * | 2016-06-29 | 2016-09-14 | 北京旷视科技有限公司 | Face recognition method capable of realizing base image automatic update and face recognition device |
| CN111104825A (en) * | 2018-10-26 | 2020-05-05 | 北京陌陌信息技术有限公司 | Face registry updating method, device, equipment and medium |
| CN109800318A (en) * | 2018-12-28 | 2019-05-24 | 上海依图网络科技有限公司 | A kind of archiving method and device |
| CN109800668A (en) * | 2018-12-28 | 2019-05-24 | 上海依图网络科技有限公司 | A kind of archiving method and device |
| CN110287791A (en) * | 2019-05-23 | 2019-09-27 | 罗普特科技集团股份有限公司 | A kind of screening technique and system for face picture |
| CN110378314A (en) * | 2019-05-31 | 2019-10-25 | 北京金山云网络技术有限公司 | A kind of human face region image archiving method, device, electronic equipment and storage medium |
| CN110276277A (en) * | 2019-06-03 | 2019-09-24 | 罗普特科技集团股份有限公司 | Method and apparatus for detecting facial image |
| CN111126208A (en) * | 2019-12-11 | 2020-05-08 | 智慧眼科技股份有限公司 | Pedestrian archiving method and device, computer equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| DANIEL SCHOFIELD等: "Chimpanzee face recognition from videos in the wild using deep learning", 《SCIENCE ADVANCES》, vol. 05, no. 09, 4 September 2019 (2019-09-04), pages 1 - 10 * |
| 崔进: "社交关系网人脸照片管理系统的设计与实现", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, vol. 138, no. 2016, 15 December 2016 (2016-12-15), pages 138 - 265 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113220942A (en) * | 2021-05-31 | 2021-08-06 | 杭州海康威视数字技术股份有限公司 | Method, device and equipment for searching associated object |
| CN117333926A (en) * | 2023-11-30 | 2024-01-02 | 深圳须弥云图空间科技有限公司 | Picture aggregation method and device, electronic equipment and readable storage medium |
| CN117333926B (en) * | 2023-11-30 | 2024-03-15 | 深圳须弥云图空间科技有限公司 | Picture aggregation method and device, electronic equipment and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110175549B (en) | Face image processing method, device, equipment and storage medium | |
| US8605956B2 (en) | Automatically mining person models of celebrities for visual search applications | |
| JP2022518469A (en) | Information processing methods and devices, storage media | |
| CN108229674B (en) | Training method and device of neural network for clustering, and clustering method and device | |
| CN111221991B (en) | Method and device for determining personnel identity attribute and electronic equipment | |
| TWI740537B (en) | Information processing method, device and storage medium thereof | |
| EP2742442B1 (en) | A method for detecting a copy of a reference video, corresponding apparatus for extracting a spatio-temporal signature from video data and corresponding computer readable storage medium | |
| Bharati et al. | Beyond pixels: Image provenance analysis leveraging metadata | |
| CN103503000A (en) | Facial recognition | |
| CN103718190A (en) | Incremental image clustering | |
| CN112183490A (en) | Face snapshot picture filing method and device | |
| WO2023273081A1 (en) | Clustering method, clustering apparatus, and non-transitory computer-readable storage medium | |
| CN110865992A (en) | Retrieval library management method, retrieval device and retrieval medium | |
| CN110765215A (en) | Query method and device for personnel common relationship, electronic equipment and storage medium | |
| CN110705297A (en) | Enterprise name-identifying method, system, medium and equipment | |
| WO2023124134A1 (en) | File processing method and apparatus, electronic device, computer storage medium and program | |
| CN111753642A (en) | Method and device for determining key frame | |
| CN103093213A (en) | Video file classification method and terminal | |
| CN116071569A (en) | Image selection method, computer equipment and storage device | |
| KR101138873B1 (en) | Method and apparatus for reducing the number of photo in photo album | |
| CN106934041B (en) | image file management method and device | |
| CN111563479B (en) | Concurrent person weight removing method, partner analyzing method and device and electronic equipment | |
| CN113111689A (en) | Sample mining method, device, equipment and storage medium | |
| KR101100429B1 (en) | Semi automatic enrollment method and apparatus of photo album, and photo album system using them | |
| JP6958952B2 (en) | Master face image registration device, registration method, program, and recording medium |
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 |