CN112149447A - Personnel identification method and device and electronic equipment - Google Patents
Personnel identification method and device and electronic equipment Download PDFInfo
- Publication number
- CN112149447A CN112149447A CN201910559901.7A CN201910559901A CN112149447A CN 112149447 A CN112149447 A CN 112149447A CN 201910559901 A CN201910559901 A CN 201910559901A CN 112149447 A CN112149447 A CN 112149447A
- Authority
- CN
- China
- Prior art keywords
- person
- identified
- image
- recognized
- persons
- 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 43
- 238000004590 computer program Methods 0.000 claims description 10
- 230000002596 correlated effect Effects 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 abstract description 20
- 239000002699 waste material Substances 0.000 abstract description 4
- 230000036544 posture Effects 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000007787 solid Substances 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
- G06V40/70—Multimodal biometrics, e.g. combining information from different biometric modalities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a person identification method, a person identification device and electronic equipment. The method comprises the following steps: acquiring a person image to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification; for each recognized person in a plurality of recognized persons, determining a confidence score of the recognized person based on the similarity between the human body features in the recognized person image of the recognized person and the human body features in the image of the person to be recognized, wherein the information content of the human face features included in the human body features is lower than a preset information content threshold value; and taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified. Monitoring data in the monitoring system can be fully utilized, personnel images with unclear face images can be identified through human body characteristics based on the personnel images, and waste of monitoring resources is effectively reduced.
Description
Technical Field
The present invention relates to the field of image recognition technologies, and in particular, to a method and an apparatus for identifying a person, and an electronic device.
Background
In some application scenarios, personnel identification may need to be performed on personnel appearing in the monitoring picture to determine personnel identification of the personnel, and the personnel identification may refer to an identity card number, a employee card number, and a pedestrian number according to different application scenarios, where the personnel identification is used to uniquely identify the personnel, that is, the personnel identification of different personnel is different. Face recognition may be performed on the monitoring picture to determine the person identifier of the person. However, the method needs to shoot a clear face image to acquire enough face feature information from the face image so as to perform accurate face recognition.
Due to special reasons, such as insufficient light of a monitored scene, improper shooting angle, low resolution of shooting equipment and the like, a camera does not shoot a clear enough face image or even a face image, and personnel identification of personnel in the image cannot be determined through face recognition. In the related art, the camera may be controlled to continue to follow the person until a sufficiently clear face image of the person is obtained, and then the person identifier of the person is determined through face recognition. However, before the face image with sufficient clarity is shot, the shot image does not include the clear face image, so that personnel identification cannot be performed, and monitoring resources are wasted.
Disclosure of Invention
The embodiment of the invention aims to provide a personnel identification method, which is used for realizing personnel identification based on human body characteristics and also can identify personnel in personnel images without clear face images, so that monitoring data is fully utilized, and the waste of monitoring resources is reduced. The specific technical scheme is as follows:
in a first aspect of an embodiment of the present invention, a method for identifying a person is provided, where the method includes:
acquiring a person image to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification;
for each identified person in the plurality of identified persons, determining a confidence score of the identified person based on the similarity between the human body features in the image of the identified person and the human body features in the image of the person to be identified;
and taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
With reference to the first aspect, in a first embodiment, an image of a person to be identified and a plurality of images of the identified person captured in a plurality of time periods are acquired;
dividing the plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the plurality of recognized person images belong based on a person recognition result obtained through face recognition in advance, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the plurality of recognized person images belong.
With reference to the first possible embodiment of the first aspect, in a second possible embodiment, the determining, for each identified person in the plurality of identified persons, a confidence score of the identified person based on a similarity between a human feature in an image of the identified person and a human feature in an image of the person to be identified includes:
for each identified person image in the plurality of identified person images, determining similarity between the human body features in the identified person image and the human body features in the to-be-identified person image;
and weighting and superposing the similarity of each recognized person image meeting preset conditions of the recognized person aiming at each recognized person in the plurality of recognized persons to obtain a weighted superposition result as a confidence score of the recognized person, wherein the preset conditions are that the similarity of the recognized person image is greater than a preset similarity threshold value.
With reference to the second possible embodiment of the first aspect, in a third possible embodiment, the weighting and superimposing, for each identified person in the plurality of identified persons, the similarity of each image of the identified person to obtain a weighted and superimposed result, which is used as the confidence score of the identified person, includes:
and for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient in negative correlation with the shooting time of the identified person image, wherein the shooting time is used for representing the time elapsed from the shooting of the identified person image, and the weighted superposition result is used as the confidence score of the identified person.
With reference to the third possible embodiment of the first aspect, in a fourth possible embodiment, the acquiring an image of a person to be identified and a plurality of images of identified persons captured in a plurality of time periods includes:
acquiring an image of a person to be identified and a plurality of images of identified persons shot in nearly N days;
for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient negatively correlated with the shooting time of the identified person image to obtain a weighted superposition result, wherein the weighted superposition result is used as the confidence score of the identified person, and the weighted superposition result comprises the following steps:
for each identified person of the plurality of identified persons, calculating a confidence score for the identified person according to:
wherein s isiAn image of the identified person taken i days ago, and the image of the identified personThe similarity of the person image is identified, and Score is the confidence Score of the identified person.
With reference to the first aspect, in a fifth possible embodiment, the determining, for each identified person in the plurality of identified persons, a confidence score of the identified person based on a similarity between a human feature in an image of the identified person and a human feature in an image of the person to be identified includes:
for each identified person in the plurality of identified persons, performing human body modeling on an identified person image of the identified person to obtain an identified human body model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
With reference to the first aspect, in a sixth possible implementation manner, the to-be-identified person image is a plurality of person images obtained by shooting the same person at a plurality of different times;
the determining the confidence score of the identified person based on the similarity between the human body features in the identified person image of the identified person and the human body features in the image of the person to be identified includes:
and respectively determining the similarity between the human body features in the identified person image of the identified person and the human body features in each image of the person to be identified, and taking the maximum value or the average value of the similarities as the confidence score of the identified person.
With reference to the first aspect, in a seventh possible application scenario, the acquiring an image of a person to be identified and an image of an identified person of each of a plurality of identified persons includes:
acquiring a person image shot by first shooting equipment as a person image to be identified, wherein the first shooting equipment is shooting equipment with the image resolution lower than a first preset resolution threshold;
the method comprises the steps of obtaining an image of an identified person, shot by a second shooting device, of each of a plurality of identified persons, wherein the second shooting device is a shooting device with an image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than the first preset resolution threshold.
In a second aspect of embodiments of the present invention, there is provided a person identification apparatus, the apparatus including:
the image acquisition module is used for acquiring an image of a person to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification;
the confidence degree scoring module is used for determining the confidence degree score of each identified person in the plurality of identified persons based on the similarity between the human body features in the identified person image of the identified person and the human body features in the to-be-identified person image;
and the person identification module is used for taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
With reference to the second aspect, in a first possible implementation manner, the image obtaining module is specifically configured to obtain an image of a person to be identified and a plurality of images of the identified person captured in a plurality of time periods;
dividing the plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the plurality of recognized person images belong based on a person recognition result obtained through face recognition in advance, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the plurality of recognized person images belong.
With reference to the first possible implementation manner of the second aspect, in a second possible implementation manner, the confidence score module is specifically configured to determine, for each of the multiple recognized person images, a similarity between a human feature in the recognized person image and a human feature in the to-be-recognized person image;
for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting a preset condition to obtain a weighted superposition result as a confidence score of the identified person, wherein the preset condition is that the similarity of the identified person image is greater than a preset similarity threshold;
and determining the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
With reference to the second possible implementation manner of the second aspect, in a third possible implementation manner, the confidence score module is specifically configured to, for each identified person in the multiple identified persons, perform weighted superposition on the similarity of each identified person image of the identified person, where the similarity satisfies a preset condition, according to a weighting coefficient that is negatively correlated with a shooting time of the identified person image, to obtain a weighted superposition result, which is used as the confidence score of the identified person, where the shooting time is used to indicate an elapsed time since the identified person image was shot.
With reference to the third possible implementation manner of the second aspect, in a fourth possible implementation manner, the image obtaining module is specifically configured to obtain an image of a person to be identified and a plurality of images of identified persons captured in nearly N days;
the confidence score module is specifically configured to calculate, for each identified person of the plurality of identified persons, a confidence score for the identified person according to the following formula:
wherein s isiAnd the Score is the confidence Score of the identified person for the similarity between the identified person image of the identified person shot from the previous i days and the image of the person to be identified.
With reference to the second aspect, in a fifth possible implementation manner, the confidence score module is specifically configured to, for each identified person in a plurality of identified persons, perform human body modeling on an identified person image of the identified person to obtain an identified human body model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
With reference to the second aspect, in a sixth possible implementation manner, the to-be-identified person image is a plurality of person images obtained by shooting the same person at a plurality of different times;
the confidence score module is specifically configured to determine similarity between the human body features in the image of the identified person and the human body features in each image of the person to be identified, and use a maximum value or an average value of the similarity as the confidence score of the identified person.
With reference to the second aspect, in a seventh possible application scenario, the image obtaining module is specifically configured to obtain a person image captured by a first capturing device as a person image to be identified, where the first capturing device is a capturing device with an image resolution lower than a first preset resolution threshold;
the method comprises the steps of obtaining an image of an identified person, shot by a second shooting device, of each of a plurality of identified persons, wherein the second shooting device is a shooting device with an image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than the first preset resolution threshold.
In a third aspect of embodiments of the present invention, there is provided an electronic device, including:
a memory for storing a computer program;
a processor for implementing the person identification method according to any one of the first aspect described above when executing a program stored in the memory.
In a fourth aspect of embodiments of the present invention, there is provided a computer-readable storage medium having stored therein a computer program which, when executed by a processor, implements the person identification method of any one of the above-described first aspects.
According to the person identification method, the person identification device and the electronic equipment provided by the embodiment of the invention, the face identification result matched with the image of the person to be identified can be determined as the person identification result of the image of the person to be identified by associating the face with the human body and utilizing the similarity of the human body characteristics. Monitoring data in the monitoring system can be fully utilized, personnel images with unclear face images can be identified through human body characteristics based on the personnel images, and waste of monitoring resources is effectively reduced. Of course, not all of the advantages described above need to be achieved at the same time in the practice of any one product or method of the invention.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a method for identifying a person according to an embodiment of the present invention;
fig. 2 is another schematic flow chart of a person identification method according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a person identification device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
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. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a schematic flow chart of a method for identifying a person according to an embodiment of the present invention, which may include:
s101, acquiring an image of a person to be identified and an image of an identified person of each of a plurality of identified persons.
The image of the person to be recognized may be a person image from which a person recognition result cannot be obtained through face recognition. The identified person image is a person image for which a person identification result has been obtained by face identification. In an alternative embodiment, a recognized person image database may be constructed in advance, a person image of which a person recognition result is obtained through face recognition is stored in the recognized person image database, and the person image is labeled with the person recognition result of the person image. The step may be reading a plurality of recognized person images from the recognized person image database, and dividing the plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which they belong based on the person recognition results of the plurality of recognized person images, wherein all the recognized person images in each of the recognized person image groups are regarded as recognized person images of the recognized persons to which they belong.
For example, assuming that 10 recognized person images are read from the recognized person database and are respectively denoted as images 1 to 10, wherein the person recognition results of the images 1 to 3 are person 1, the person recognition results of the images 4 to 8 are person 2, and the person recognition results of the images 9 to 10 are person 3, the 10 images may be divided into three recognized person image groups, which are respectively: the group of recognized persons 1{ image 1, image 2, image 3}, the group of recognized persons 2{ image 4, image 5, image 6, image 7, image 8}, and the group of recognized persons 3{ image 9, image 10 }. It is understood that, in this example, the person 1, the person 2, and the person 3 are recognized persons, and that all the images in the group of recognized person images 1 are recognized person images of the person 1, all the images in the group of recognized person images 2 are recognized person images of the person 2, and all the images in the group of recognized person images 3 are recognized person images of the person 3.
The image of the person to be identified and the image of the identified person can be shot by the same shooting device or can be shot by different shooting devices. For example, in an alternative embodiment, one or more first photographing devices and one or more second photographing devices may be included in the monitoring system. The image resolution of the first shooting device is lower than a first preset resolution threshold, the image resolution of the second shooting device is higher than a preset image resolution threshold, the second preset resolution threshold is not lower than the first preset resolution threshold, and in some application scenes, the second preset resolution threshold may be equal to the first preset resolution threshold, that is, the image resolution of the second shooting device is higher than the image resolution of the first shooting device. The specific values of the first preset resolution threshold and the second preset resolution threshold may be set according to user requirements or experience.
Because the image resolution of the first shooting device is low, the person image resolution in the person image shot by the first shooting device can be considered to be low, and the person image cannot be subjected to face recognition, so that the person image shot by the first shooting device can be used as the person image to be recognized. The image resolution of the second shooting device is high, so that the person image shot by the second shooting device can be considered to include a sufficiently clear face image, and therefore the person image shot by the second shooting device can be subjected to face recognition to serve as a recognized person image. It can be understood that the cost of the shooting device is usually positively correlated with the image resolution of the shooting device, so that the embodiment can be adopted to assist the shot personnel image by using the second shooting device with higher image resolution, such as a personnel snapshot machine, so as to realize personnel identification on the shot personnel image of the first shooting device with lower image resolution, such as a conventional video device. The cost of the monitoring system can be effectively reduced.
S102, aiming at each identified person in a plurality of identified persons, determining the confidence score of the identified person based on the similarity between the human body features in the image of the identified person and the human body features in the image of the person to be identified.
The human body features may include a posture feature, a wearing style feature, and a posture feature of a person, and the human body features may include a part of a face feature, such as a face contour feature, but in this embodiment, an information amount of the face feature included in the human body features is less than a preset information amount threshold, that is, a person recognition result of the person image cannot be obtained through face recognition based on the face feature included in the human body features. If the information amount of the face features included in the human body features is not less than the preset information amount threshold, it can be considered that the person recognition result of the person image can be obtained through face recognition, that is, the technical problem to be solved by the present invention does not exist.
For one identified person, there may be only one identified person image, or there may be multiple identified person images, as shown in the example in S101, similarly, the image of the person to be identified may be one person image, or multiple person images obtained by shooting the same person at multiple different times, and for example, multiple person images obtained by continuously shooting one person may be used.
For convenience of discussion, the following description will first take a case where the image of the person to be recognized is one person image as an example, and in a case where the image of the recognized person is one image, the similarity between the human features in the one image and the human features in the image of the person to be recognized may be calculated as the confidence score of the recognized person. The similarity can be calculated by performing human body modeling on the image to obtain an identified human body model, and performing human body modeling on the image of the person to be identified to obtain the human body model to be identified. And calculating the similarity between the recognized human body model and the human body model to be recognized, wherein the similarity is the similarity between the human body characteristics in the image and the human body characteristics in the image of the person to be recognized.
For the case where the identified person image of the identified person is a plurality of images, the confidence score may be calculated in different ways according to different application scenarios. For example, it is assumed that there are n recognized person images of the recognized person in total, which are respectively denoted as recognized person image 1-recognized person image n. The human body characteristics in the n recognized personnel images can be respectively calculated, and the similarity between the human body characteristics in the n recognized personnel images and the human body characteristics in the personnel images to be recognized is obtained to obtain a similarity set { sim1、sim2…、simnWhere sim is1Representing the similarity, sim, of the human features in the image 1 of the identified person with the human features in the image of the person to be identified2…、simnThe meanings indicated are analogized in turn.
In an alternative embodiment, each similarity in the similarity set may be weighted and superimposed, and the resulting superimposed result is used as the similarity between the human features in the image of the identified person and the human features in the image of the identified person, i.e. the confidence score of the identified person. In other optional embodiments, a maximum value, a minimum value, or a median value in the similarity set may also be selected as the similarity between the human features in the image of the identified person and the human features in the image of the identified person, that is, the confidence score of the identified person, which is not limited in this embodiment.
It is understood that the higher the similarity between the human body feature in the recognized person image of the recognized person and the human body feature in the recognized person image, the higher the possibility that the person recognition result of the person image to be recognized is the recognized person.
For the case that the to-be-identified person image is a plurality of person images, the similarity between the identified person image and each to-be-identified person image may be calculated respectively, and the calculation of the similarity may refer to the related description of the case that the to-be-identified person image is one image, which is not described herein again, and a plurality of similarities are obtained. The average of these similarity degrees may be an arithmetic average, a weighted average, a geometric average, or the like, or a maximum value, depending on the application scenario, as the confidence score of the identified person.
It is understood that the human features in an image of a person may be affected by the pose of the person when the image of the person was taken. For example, for the same person, the human body characteristics of the image of the person captured when the person walks may be different from the human body characteristics of the image of the person captured when the person sits down. The postures of the personnel can be approximately considered to be randomly distributed in the time domain and are influenced by subjective consciousness of the personnel, so that the personnel images shot by the same personnel at a plurality of different moments are selected as the images of the personnel to be recognized, the influence of the postures of the personnel on the human body characteristics can be effectively reduced, the similarity between the human body characteristics of the calculated images of the personnel to be recognized and the human body characteristics of the images of the recognized personnel can be better reflected, and the confidence coefficient of the images of the personnel to be recognized and the confidence coefficient of the images of the recognized personnel belonging to the same personnel can be better reflected, namely the accuracy of the confidence coefficient score is improved.
And S103, using the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
It is considered that when there are a sufficient number of recognized persons included in the plurality of recognized persons, the person image to be recognized should be a person image of one of the plurality of recognized persons. Therefore, the recognized person with the highest confidence score among the plurality of recognized persons can be regarded as the person to which the image of the person to be recognized belongs, and therefore, the recognized person can be regarded as the person recognition result of the image of the person to be recognized. The identified person may be used as the result of the identification of the person to be identified, and the person identifier of the identified person may be determined as the person identifier corresponding to the person image.
By adopting the embodiment, the human face and the human body can be associated, and the human face recognition result matched with the image of the person to be recognized is determined by utilizing the similarity of the human body characteristics and is used as the person recognition result of the image of the person to be recognized. Monitoring data in the monitoring system can be fully utilized, personnel images with unclear face images can be identified through human body characteristics based on the personnel images, and waste of monitoring resources is effectively reduced.
In order to more clearly describe the person identification method provided by the embodiment of the present invention, a specific application scenario will be taken as an example below. The urban monitoring system is provided with human body snapshot equipment and video equipment, wherein the resolution ratio of the human body snapshot equipment is higher than that of the video equipment, the video equipment is used for shooting a monitoring scene for a long time to obtain a monitoring picture of the monitoring scene, and the personnel snapshot machine is used for snapshotting personnel images of pedestrians. In other optional embodiments, the definition of the face image in the person image is determined to be greater than a preset definition threshold, and if the definition of the face image in the person image is greater than the preset definition threshold, the person recognition result of the person image may be determined through a preset face recognition algorithm.
And the personnel image marked with the personnel identification result is stored in the labeled data set as the identified personnel image. In the tagged data set, the person images are sorted by date of capture. And if the definition of the face image in the personnel image is not greater than a preset definition threshold value, storing the personnel image in a non-label data set.
In this application scenario, an image q in the unlabeled dataset is described below, in this embodiment, the image q may be regarded as an image of a person to be identified, and person identification is performed as an example, which may be referred to in fig. 2, and includes:
s201, carrying out human body modeling on the image q to obtain a human body model to be recognized, and carrying out human body modeling on the recognized image of the data set to be marked for near N days to obtain a plurality of recognized human body models.
The last N days are within the last N days from the current date, and include the current date, namely the current date, the day before the current date, …, and N days before the current date, wherein N is an integer greater than or equal to 1.
S202, respectively calculating the similarity between the human body model to be recognized and the recognized human body models, and keeping the similarity higher than a preset similarity threshold value as a similarity set.
For example, assuming that there are 5 identified human models, the similarity between the 5 identified human models and the human model to be identified is 0.9,0.8, 07, 0.6, 0.5, respectively, and if the preset similarity threshold is 0.75, the obtained similarity set is {0.9,0.8 }. It can be understood that if the similarity between one identified human body model and the human body model to be identified is lower than the preset similarity threshold, the identified human body model and the human body model to be identified belong to different persons, and the identified human body model may not be referred to when determining the person of the human body model to be identified.
In an alternative embodiment, the similarity sets may also be represented in groups according to dates, which is only different from the representation manner of the similarity sets, and the principle is the same, in other alternative embodiments, the similarity sets may also be represented in other manners, which is not limited in this embodiment, for example, the similarity set may be represented as res0＝{p0,1，p0,2…，p0，j}、res1＝{p1,1，p1,2…，p1，j}…、resN＝{pN,1，pN,2…，pN，jTherein res0Representing the similarity of the human body model with the similarity larger than a preset similarity threshold value in the human body model obtained by modeling the image of the identified person shot at the current date, p0,1Similarity of the human body model obtained by modeling the recognized person image of the person 1 photographed at the current date.
S203, for each identified person, calculating a confidence score of the identified person according to the following formula:
where Score is the confidence Score, s, for the identified personiAnd similarity between the identified human body model corresponding to the identified human body image shot by the identified human body from the current date to i days and the human body model to be identified. E.g. s0And the similarity between the identified human body model corresponding to the identified human body image of the identified human body shot from the current date to the previous 0 days, namely the current date, and the human body model to be identified is represented.
Wherein
Can be regarded as the similarity siIt is understood that a larger i indicates a larger number of previous days, and the weighting coefficient is smaller. It will be appreciated that the human features have a certain timeliness, for example, a person may wear different clothing between two days, and for example, a person may change in posture over time, so that if the time period from the time when an identified image is captured is longer, the reference of the identified image is lower, and thus the weight of the identified human model obtained by human modeling of the identified image can be set lower. In other application scenarios, the weighting coefficients may also be set according to other manners, which is not limited in this embodiment.
The persons included in the plurality of recognized persons depend on the recognized persons corresponding to the similarity stored in the similarity set. For example, assuming that a total of three similarities are stored in the similarity set, where the first similarity corresponds to identified person 1, the second similarity corresponds to identified person 2, and the third similarity corresponds to identified person 2, the identified persons include identified person 1 and identified person 2.
S204, sequencing the multiple recognized persons according to the sequence of the confidence score from high to low, and taking the recognized person at the head as the person recognition result of the image q.
It will be appreciated that the physical characteristics of a person are affected by the dress of the person, and that the physical characteristics of the same person may vary considerably. Thus, in some cases, the human body models obtained by human body modeling two different images of the same person may be very different. However, it is considered that the same person has a similar time for dressing in a certain period of time. For example, one person may dress on monday and tuesday more closely, but within seven days of the week there may be two days to dress closer. Therefore, the embodiment is adopted, and the human body models in the historical time period are integrated, so that the influence of the change of human body characteristics along with time on the similarity is effectively reduced, and the accuracy of the personnel identification result is improved.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a person identification device according to an embodiment of the present invention, which may include:
the image acquisition module 301 is configured to acquire a person image to be identified and respective identified person images of a plurality of identified persons, where the identified person images are person images for which person identification results are obtained through face identification;
the confidence score module 302 is configured to determine, for each recognized person in the multiple recognized persons, a confidence score of the recognized person based on a similarity between a human body feature in a recognized person image of the recognized person and a human body feature in a person image to be recognized, where an information amount of a human face feature included in the human body feature is lower than a preset information amount threshold;
and the person identification module 303 is configured to use the identified person with the highest confidence score among the plurality of identified persons as the person identification result of the image of the person to be identified.
In an optional embodiment, the image obtaining module 301 is specifically configured to obtain an image of a person to be identified and a plurality of images of identified persons captured in a plurality of time periods;
based on a person recognition result obtained by face recognition in advance, dividing a plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the images belong, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the images belong.
In an alternative embodiment, the confidence score module 302 is specifically configured to determine, for each of the multiple recognized person images, a similarity between a human feature in the recognized person image and a human feature in the person image to be recognized;
for each identified person in a plurality of identified persons, performing weighted superposition on the similarity of each image of the identified person, meeting a preset condition, of the identified person to obtain a weighted superposition result, wherein the weighted superposition result is used as a confidence score of the identified person, and the preset condition is that the similarity of the image of the identified person is greater than a preset similarity threshold;
and determining the identified person with the highest confidence score among the plurality of identified persons as the person identification result of the image of the person to be identified.
In an alternative embodiment, the confidence score module 302 is specifically configured to, for each identified person in the multiple identified persons, perform weighted overlap on the similarity of each identified person image of the identified person, where the similarity satisfies a preset condition, according to a weighting coefficient negatively correlated with a shooting time of the identified person image, so as to obtain a weighted overlap result, which is used as the confidence score of the identified person, where the shooting time is used to indicate an elapsed time from shooting of the identified person image.
In an optional embodiment, the image obtaining module 301 is specifically configured to obtain an image of a person to be identified and a plurality of images of identified persons captured in near N days;
the confidence score module is specifically configured to calculate, for each identified person of the plurality of identified persons, a confidence score for the identified person according to the following formula:
wherein s isiTaken i days beforeThe similarity between the image of the identified person and the image of the person to be identified, Score is the confidence Score of the identified person.
In an alternative embodiment, the confidence score module 302 is specifically configured to perform human body modeling on an image of an identified person of a plurality of identified persons to obtain an identified human body model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
In an alternative embodiment, the image of the person to be identified is a plurality of images of persons obtained by shooting the same person at a plurality of different moments;
the confidence score module 302 is specifically configured to determine similarity between human features in the image of the identified person and the human features in each image of the person to be identified, and use a maximum value or an average value of the similarity as the confidence score of the identified person.
In an optional embodiment, the image obtaining module is specifically configured to obtain a person image captured by a first capturing device as a person image to be identified, where the first capturing device is a capturing device whose image resolution is lower than a first preset resolution threshold;
and acquiring the images of the identified persons, which are shot by second shooting equipment by the plurality of identified persons respectively, wherein the second shooting equipment is shooting equipment with the image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than a first preset resolution threshold.
An embodiment of the present invention further provides an electronic device, as shown in fig. 4, including:
a memory 401 for storing a computer program;
the processor 402, when executing the program stored in the memory 401, implements the following steps:
acquiring a person image to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification;
for each recognized person in a plurality of recognized persons, determining a confidence score of the recognized person based on the similarity between the human body features in the recognized person image of the recognized person and the human body features in the image of the person to be recognized, wherein the information content of the human face features included in the human body features is lower than a preset information content threshold value;
and taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
In an alternative embodiment, acquiring an image of a person to be identified and an image of an identified person of each of a plurality of identified persons includes:
acquiring images of a person to be identified and a plurality of identified person images shot in a plurality of time periods;
based on a person recognition result obtained by face recognition in advance, dividing a plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the images belong, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the images belong.
In an alternative embodiment, for each identified person in the plurality of identified persons, determining a confidence score for the identified person based on a similarity between human features in the image of the identified person and human features in the image of the person to be identified includes:
for each identified person image in the plurality of identified person images, determining similarity between the human body features in the identified person image and the human body features in the person image to be identified;
and for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each image of the identified person meeting preset conditions to obtain a weighted superposition result, wherein the weighted superposition result is used as a confidence score of the identified person, and the preset conditions are that the similarity of the image of the identified person is greater than a preset similarity threshold.
In an alternative embodiment, for each identified person in the plurality of identified persons, weighted overlapping the similarity of each identified person image of the identified person to obtain a weighted overlapping result as the confidence score of the identified person includes:
and for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient in negative correlation with the shooting time of the identified person image to obtain a weighted superposition result, wherein the weighted superposition result is used as the confidence score of the identified person, and the shooting time is used for representing the time elapsed from the shooting of the identified person image.
In an alternative embodiment, acquiring an image of a person to be identified and a plurality of images of the identified person captured over a plurality of time periods includes:
acquiring an image of a person to be identified and a plurality of images of identified persons shot in nearly N days;
for each identified person in a plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient negatively correlated with the shooting time of the identified person image to obtain a weighted superposition result, wherein the weighted superposition result is used as the confidence score of the identified person, and the weighted superposition result comprises the following steps:
for each identified person of the plurality of identified persons, calculating a confidence score for the identified person according to:
wherein s isiFor the similarity between the image of the identified person shot i days before and the image of the person to be identified, Score is the confidence Score of the identified person.
In an alternative embodiment, for each identified person in the plurality of identified persons, determining a confidence score for the identified person based on a similarity between human features in the image of the identified person and human features in the image of the person to be identified includes:
for each identified person in the plurality of identified persons, performing human body modeling on an identified person image of the identified person to obtain an identified human body model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
In an alternative embodiment, the image of the person to be identified is a plurality of images of persons obtained by shooting the same person at a plurality of different moments;
determining a confidence score of the identified person based on the similarity between the human features in the identified person image of the identified person and the human features in the image of the person to be identified, including:
and respectively determining the similarity between the human body features in the image of the identified person and the human body features in each image of the person to be identified, and taking the maximum value or the average value of the similarity as the confidence score of the identified person.
In an alternative embodiment, acquiring an image of a person to be identified and an image of an identified person of each of a plurality of identified persons includes:
acquiring a person image shot by first shooting equipment as a person image to be identified, wherein the first shooting equipment is shooting equipment with the image resolution lower than a first preset resolution threshold;
and acquiring the images of the identified persons, which are shot by second shooting equipment by the plurality of identified persons respectively, wherein the second shooting equipment is shooting equipment with the image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than a first preset resolution threshold.
The electronic device Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
In yet another embodiment of the present invention, a computer-readable storage medium is further provided, which has instructions stored therein, and when the instructions are executed on a computer, the computer is caused to execute any one of the above-mentioned person identification methods.
In a further embodiment provided by the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform any of the above-described person identification methods.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the electronic device, the computer-readable storage medium, and the computer program product, since they are substantially similar to the method embodiments, the description is relatively simple, and for the relevant points, reference may be made to the partial description of the method embodiments.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (18)
1. A person identification method, characterized in that the method comprises:
acquiring a person image to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification;
for each recognized person in the multiple recognized persons, determining a confidence score of the recognized person based on the similarity between the human body features in the recognized person image of the recognized person and the human body features in the to-be-recognized person image, wherein the information content of the human face features included in the human body features is lower than a preset information content threshold;
and taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
2. The method of claim 1, wherein the obtaining an image of the person to be identified and an image of an identified person for each of a plurality of identified persons comprises:
acquiring images of a person to be identified and a plurality of identified person images shot in a plurality of time periods;
dividing the plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the plurality of recognized person images belong based on a person recognition result obtained through face recognition in advance, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the plurality of recognized person images belong.
3. The method of claim 2, wherein determining, for each identified person in the plurality of identified persons, a confidence score for the identified person based on a similarity between features of the person in the image of the identified person and features of the person in the image of the person to be identified comprises:
for each identified person image in the plurality of identified person images, determining similarity between the human body features in the identified person image and the human body features in the to-be-identified person image;
and weighting and superposing the similarity of each recognized person image meeting preset conditions of the recognized person aiming at each recognized person in the plurality of recognized persons to obtain a weighted superposition result as a confidence score of the recognized person, wherein the preset conditions are that the similarity of the recognized person image is greater than a preset similarity threshold value.
4. The method of claim 3, wherein the weighted overlap-add of the similarity of each identified person image of the identified person for each identified person in the plurality of identified persons to obtain a weighted overlap-add result as the confidence score for the identified person comprises:
and for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient in negative correlation with the shooting time of the identified person image, wherein the shooting time is used for representing the time elapsed from the shooting of the identified person image, and the weighted superposition result is used as the confidence score of the identified person.
5. The method of claim 4, wherein the acquiring the image of the person to be identified and the plurality of images of the identified person captured over the plurality of time periods comprises:
acquiring an image of a person to be identified and a plurality of images of identified persons shot in nearly N days;
for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting preset conditions of the identified person according to a weighting coefficient negatively correlated with the shooting time of the identified person image to obtain a weighted superposition result, wherein the weighted superposition result is used as the confidence score of the identified person, and the weighted superposition result comprises the following steps:
for each identified person of the plurality of identified persons, calculating a confidence score for the identified person according to:
wherein s isiAnd the Score is the confidence Score of the identified person for the similarity between the identified person image of the identified person shot from the previous i days and the image of the person to be identified.
6. The method of claim 1, wherein determining, for each identified person in the plurality of identified persons, a confidence score for the identified person based on a similarity between human features in an image of the identified person and human features in an image of the person to be identified comprises:
for each identified person in the plurality of identified persons, performing human body modeling on an identified person image of the identified person to obtain an identified human body model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
7. The method according to claim 1, wherein the image of the person to be identified is a plurality of images of persons obtained by photographing the same person at a plurality of different times;
the determining the confidence score of the identified person based on the similarity between the human body features in the identified person image of the identified person and the human body features in the image of the person to be identified includes:
and respectively determining the similarity between the human body features in the identified person image of the identified person and the human body features in each image of the person to be identified, and taking the maximum value or the average value of the similarities as the confidence score of the identified person.
8. The method of claim 1, wherein the obtaining an image of the person to be identified and an image of an identified person for each of a plurality of identified persons comprises:
acquiring a person image shot by first shooting equipment as a person image to be identified, wherein the first shooting equipment is shooting equipment with the image resolution lower than a first preset resolution threshold;
the method comprises the steps of obtaining an image of an identified person, shot by a second shooting device, of each of a plurality of identified persons, wherein the second shooting device is a shooting device with an image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than the first preset resolution threshold.
9. A person identification device, characterized in that the device comprises:
the image acquisition module is used for acquiring an image of a person to be identified and respective identified person images of a plurality of identified persons, wherein the identified person images are person images of which person identification results are obtained through face identification;
the confidence score module is used for determining the confidence score of each identified person in the plurality of identified persons based on the similarity between the human body features in the image of the identified person and the human body features in the image of the person to be identified, wherein the information content of the human face features included in the human body features is lower than a preset information content threshold value;
and the person identification module is used for taking the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
10. The device according to claim 9, wherein the image acquisition module is specifically configured to acquire an image of a person to be identified and a plurality of images of the identified person captured in a plurality of time periods;
dividing the plurality of recognized person images into a plurality of recognized person image groups according to the recognized persons to which the plurality of recognized person images belong based on a person recognition result obtained through face recognition in advance, wherein all recognized person images in each recognized person image group are used as recognized person images of the recognized persons to which the plurality of recognized person images belong.
11. The apparatus according to claim 10, wherein the confidence score module is specifically configured to determine, for each of the plurality of recognized person images, a similarity between the human features in the recognized person image and the human features in the image of the person to be recognized;
for each identified person in the plurality of identified persons, performing weighted superposition on the similarity of each identified person image meeting a preset condition to obtain a weighted superposition result as a confidence score of the identified person, wherein the preset condition is that the similarity of the identified person image is greater than a preset similarity threshold;
and determining the identified person with the highest confidence score in the plurality of identified persons as the person identification result of the image of the person to be identified.
12. The apparatus according to claim 11, wherein the confidence score module is specifically configured to, for each of the plurality of identified persons, perform weighted overlap on the similarity of each of the images of the identified persons that satisfy the preset condition according to a weighting coefficient that is negatively correlated with a shooting time of the image of the identified person, so as to obtain a weighted overlap result as the confidence score of the identified person, where the shooting time is used to indicate an elapsed time from shooting of the image of the identified person.
13. The device according to claim 12, wherein the image acquisition module is specifically configured to acquire an image of a person to be identified and a plurality of images of identified persons captured within approximately N days;
the confidence score module is specifically configured to calculate, for each identified person of the plurality of identified persons, a confidence score for the identified person according to the following formula:
wherein s isiAnd the Score is the confidence Score of the identified person for the similarity between the identified person image of the identified person shot from the previous i days and the image of the person to be identified.
14. The apparatus according to claim 9, wherein the confidence score module is specifically configured to perform human modeling on an image of an identified person of a plurality of identified persons for each identified person to obtain an identified human model of the identified person;
and calculating the similarity between the recognized human body model of the recognized person and the human body model to be recognized as the confidence score of the recognized person, wherein the human body model to be recognized is obtained by performing human body modeling on the image of the person to be recognized.
15. The device according to claim 9, wherein the image of the person to be identified is a plurality of images of persons obtained by shooting the same person at a plurality of different times;
the confidence score module is specifically configured to determine similarity between the human body features in the image of the identified person and the human body features in each image of the person to be identified, and use a maximum value or an average value of the similarity as the confidence score of the identified person.
16. The apparatus according to claim 9, wherein the image obtaining module is specifically configured to obtain a person image captured by a first capturing device as the person image to be identified, where the first capturing device is a capturing device whose image resolution is lower than a first preset resolution threshold;
the method comprises the steps of obtaining an image of an identified person, shot by a second shooting device, of each of a plurality of identified persons, wherein the second shooting device is a shooting device with an image resolution higher than a second preset resolution threshold, and the second preset resolution threshold is not lower than the first preset resolution threshold.
17. An electronic device, comprising:
a memory for storing a computer program;
a processor for implementing the method steps of any of claims 1 to 8 when executing a program stored in the memory.
18. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910559901.7A CN112149447A (en) | 2019-06-26 | 2019-06-26 | Personnel identification method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910559901.7A CN112149447A (en) | 2019-06-26 | 2019-06-26 | Personnel identification method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112149447A true CN112149447A (en) | 2020-12-29 |
Family
ID=73869721
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910559901.7A Pending CN112149447A (en) | 2019-06-26 | 2019-06-26 | Personnel identification method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112149447A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554693A (en) * | 2021-09-18 | 2021-10-26 | 深圳市安软慧视科技有限公司 | Correlation and judgment method, device and storage medium for edge deployment image |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070092245A1 (en) * | 2005-10-20 | 2007-04-26 | Honeywell International Inc. | Face detection and tracking in a wide field of view |
| JP2012083997A (en) * | 2010-10-13 | 2012-04-26 | Hitachi Information & Communication Engineering Ltd | Authentication system and determination method for authentication reliability |
| CN106412491A (en) * | 2015-07-30 | 2017-02-15 | 中兴通讯股份有限公司 | Video monitoring method, apparatus and system |
| CN107491744A (en) * | 2017-07-31 | 2017-12-19 | 广东欧珀移动通信有限公司 | Human body personal identification method, device, mobile terminal and storage medium |
| CN107527075A (en) * | 2016-06-20 | 2017-12-29 | 杭州海康威视数字技术股份有限公司 | RFID label tag is established with personnel's corresponding relation and trajectory track method and device |
| CN108009465A (en) * | 2016-10-31 | 2018-05-08 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| CN108021846A (en) * | 2016-11-01 | 2018-05-11 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| CN108921008A (en) * | 2018-05-14 | 2018-11-30 | 深圳市商汤科技有限公司 | Portrait identification method, device and electronic equipment |
| US20190026543A1 (en) * | 2017-07-24 | 2019-01-24 | Fuji Xerox Co., Ltd. | Person identification device |
| CN109753920A (en) * | 2018-12-29 | 2019-05-14 | 深圳市商汤科技有限公司 | A kind of pedestrian recognition method and device |
| CN109766755A (en) * | 2018-12-06 | 2019-05-17 | 深圳市天彦通信股份有限公司 | Face identification method and Related product |
-
2019
- 2019-06-26 CN CN201910559901.7A patent/CN112149447A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070092245A1 (en) * | 2005-10-20 | 2007-04-26 | Honeywell International Inc. | Face detection and tracking in a wide field of view |
| JP2012083997A (en) * | 2010-10-13 | 2012-04-26 | Hitachi Information & Communication Engineering Ltd | Authentication system and determination method for authentication reliability |
| CN106412491A (en) * | 2015-07-30 | 2017-02-15 | 中兴通讯股份有限公司 | Video monitoring method, apparatus and system |
| CN107527075A (en) * | 2016-06-20 | 2017-12-29 | 杭州海康威视数字技术股份有限公司 | RFID label tag is established with personnel's corresponding relation and trajectory track method and device |
| CN108009465A (en) * | 2016-10-31 | 2018-05-08 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| CN108021846A (en) * | 2016-11-01 | 2018-05-11 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| US20190026543A1 (en) * | 2017-07-24 | 2019-01-24 | Fuji Xerox Co., Ltd. | Person identification device |
| CN107491744A (en) * | 2017-07-31 | 2017-12-19 | 广东欧珀移动通信有限公司 | Human body personal identification method, device, mobile terminal and storage medium |
| CN108921008A (en) * | 2018-05-14 | 2018-11-30 | 深圳市商汤科技有限公司 | Portrait identification method, device and electronic equipment |
| CN109766755A (en) * | 2018-12-06 | 2019-05-17 | 深圳市天彦通信股份有限公司 | Face identification method and Related product |
| CN109753920A (en) * | 2018-12-29 | 2019-05-14 | 深圳市商汤科技有限公司 | A kind of pedestrian recognition method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554693A (en) * | 2021-09-18 | 2021-10-26 | 深圳市安软慧视科技有限公司 | Correlation and judgment method, device and storage medium for edge deployment image |
| CN113554693B (en) * | 2021-09-18 | 2021-12-31 | 深圳市安软慧视科技有限公司 | Correlation and judgment method, device and storage medium for edge deployment image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10242250B2 (en) | Picture ranking method, and terminal | |
| Lahiri et al. | Biometric animal databases from field photographs: identification of individual zebra in the wild | |
| US9805098B2 (en) | Method for learning a latent interest taxonomy from multimedia metadata | |
| US9147128B1 (en) | Machine learning enhanced facial recognition | |
| CN108629791A (en) | Pedestrian tracting method and device and across camera pedestrian tracting method and device | |
| JP2011118782A (en) | Image processor, image processing method, and program | |
| US11176679B2 (en) | Person segmentations for background replacements | |
| CN110781733B (en) | Image duplicate removal method, storage medium, network equipment and intelligent monitoring system | |
| US20110150278A1 (en) | Information processing apparatus, processing method thereof, and non-transitory storage medium | |
| CN111008935A (en) | Face image enhancement method, device, system and storage medium | |
| CN107741996A (en) | Family's map construction method and device based on recognition of face, computing device | |
| CN105893813A (en) | Biometric Information Registration Apparatus And Biometric Information Registration Method | |
| CN112199530B (en) | Multi-dimensional face library picture automatic updating method, system, equipment and medium | |
| CN109902550A (en) | The recognition methods of pedestrian's attribute and device | |
| WO2023174304A1 (en) | Systems, methods, and storage devices for data clustering | |
| CN109784220A (en) | A kind of method and device of determining passerby track | |
| CN109308704A (en) | Background elimination method, device, computer equipment and storage medium | |
| CN112149447A (en) | Personnel identification method and device and electronic equipment | |
| Wang et al. | Two-stage unsupervised facial image quality measurement | |
| Dong et al. | A detection-regression based framework for fish keypoints detection | |
| CN110019951B (en) | Method and equipment for generating video thumbnail | |
| CN111971686A (en) | Automatic generation of training data sets for object recognition | |
| CN114049608A (en) | Track monitoring method and device, computer equipment and storage medium | |
| CN114898155A (en) | Vehicle damage assessment method, device, equipment and storage medium | |
| CN114092809A (en) | Object identification method and device and electronic equipment |
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 |