CN113449567B - Face temperature detection method and device, electronic equipment and storage medium - Google Patents
Face temperature detection method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113449567B CN113449567B CN202010229946.0A CN202010229946A CN113449567B CN 113449567 B CN113449567 B CN 113449567B CN 202010229946 A CN202010229946 A CN 202010229946A CN 113449567 B CN113449567 B CN 113449567B
- Authority
- CN
- China
- Prior art keywords
- face
- trained
- detection model
- rgb
- infrared heat
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 208
- 230000001815 facial effect Effects 0.000 claims abstract description 58
- 238000000034 method Methods 0.000 claims description 49
- 238000000605 extraction Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 11
- 230000036760 body temperature Effects 0.000 abstract description 16
- 206010005908 Body temperature conditions Diseases 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 13
- 238000001931 thermography Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J5/0022—Radiation pyrometry, e.g. infrared or optical thermometry for sensing the radiation of moving bodies
- G01J5/0025—Living bodies
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J2005/0077—Imaging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (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)
- Spectroscopy & Molecular Physics (AREA)
- Image Analysis (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention provides a face temperature detection method, a face temperature detection device, electronic equipment and a storage medium, wherein the face temperature detection method comprises the following steps: acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map; inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized; the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples. Therefore, the identity recognition and the body temperature test of the target personnel can be completed directly through collecting the facial infrared heat map in the facial image of the target personnel, and the body temperature condition of the target personnel can be monitored and managed conveniently. And further reduces the labor cost and improves the working efficiency of temperature detection.
Description
Technical Field
The present invention relates to the field of artificial intelligence technologies, and in particular, to a face temperature detection method and apparatus, an electronic device, and a storage medium.
Background
At present, if the body temperature of the corresponding target person needs to be monitored and managed, when the body temperature of the target person needs to be measured in scenes such as hospitals, communities, schools and the like, intelligent identification of the body temperature cannot be achieved, and only relevant staff can be assigned to measure the temperature of each target person and inquire relevant identity information and register the temperature, so that the mode is low in working efficiency and high in labor cost.
Disclosure of Invention
The embodiment of the invention provides a face temperature detection method, which can complete identification and body temperature detection of target personnel by using an infrared heat map and can improve the working efficiency.
In a first aspect, an embodiment of the present invention provides a face temperature detection method, including the following steps:
acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map;
inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples.
Optionally, the face sample set further includes a first face RGB pattern book, and the training method of the pre-trained face temperature detection model includes:
acquiring the face sample set;
acquiring a pre-trained face detection model;
and inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of the first face RGB image sample in the training process so that the pre-trained face detection model learns the prediction of the face temperature information according to the face infrared thermal pattern, thereby obtaining the pre-trained face temperature detection model.
Optionally, the inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of the first face RGB pattern book in the training process includes:
performing branch reduction processing on the pre-trained face detection model so as to enable the pre-trained face detection model to adapt to single-channel input of the face infrared heat map sample;
and inputting the face sample set into the training model of the face detection after the branch is subtracted to train, and reducing the input weight of the first face RGB image sample to zero in the training process.
Optionally, the acquiring the face sample set includes:
acquiring a first face RGB pattern book set, wherein the first face RGB pattern book comprises a first face identity tag;
acquiring a human face infrared heat pattern book set, wherein the human face infrared heat pattern book comprises a human face temperature label;
and forming the face sample set according to the first face RGB image sample set and the face infrared heat map sample set.
Optionally, the acquiring the first face RGB pattern book set includes:
acquiring a plurality of first face RGB images;
inputting the plurality of first face RGB images into a face RGB primary feature extraction model to extract face RGB image primary features so as to obtain a plurality of first face RGB image primary features;
and obtaining the first face RGB pattern book set based on the plurality of first face RGB pattern primary features.
Optionally, the acquiring the set of infrared thermal patterns of the face includes:
acquiring a plurality of human face infrared heat maps;
inputting the plurality of facial infrared heat maps into a facial infrared heat map primary feature extraction model to extract facial infrared heat map primary features so as to obtain a plurality of facial infrared heat map primary features;
and obtaining the facial infrared thermal pattern book set based on the primary characteristics of the plurality of facial infrared thermal patterns.
Optionally, the acquiring a pre-trained face detection model includes:
acquiring a second face RGB pattern book set, wherein the second face RGB pattern book comprises a second face identity tag;
and inputting the second human face RGB pattern set into a pre-training model for training, so that the pre-training model learns the prediction of human face identity information according to the second human face RGB pattern set to obtain the pre-trained human face detection model.
In a second aspect, an embodiment of the present invention further provides a face temperature detection apparatus, where the apparatus includes:
the acquisition module is used for acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map;
the recognition module is used for inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples.
In a third aspect, an embodiment of the present invention provides an electronic device, including: the facial temperature detection method comprises the steps of a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the steps in the facial temperature detection method are realized when the processor executes the computer program.
In a fourth aspect, an embodiment of the present invention provides a computer readable storage medium, where a computer program is stored, where the computer program when executed by a processor implements the steps in the face temperature detection method provided in the embodiment of the present invention.
In the embodiment of the invention, the face image to be identified is obtained and comprises a face infrared heat map; inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized; the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples. Therefore, the identity recognition and the body temperature test of the target personnel can be completed directly through collecting the facial infrared heat map in the facial image of the target personnel, and the body temperature condition of the target personnel can be monitored and managed conveniently. Furthermore, the problems of high labor cost and low temperature detection working efficiency caused by the fact that in the prior art, intelligent body temperature identification cannot be achieved, related staff can only be assigned to measure temperature and inquire identity information for each target person and register the temperature are solved. And further reduces the labor cost and improves the working efficiency of temperature detection.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a face temperature detection method provided in an embodiment of the present invention;
FIG. 2 is a flowchart of a training method of a face temperature detection model according to an embodiment of the present invention;
FIG. 3 is a flow chart of a method provided by step 201 in the embodiment of FIG. 2;
FIG. 4 is a flow chart of one method provided by step 202 in the embodiment of FIG. 2;
FIG. 5 is a flow chart of a method provided by step 203 of the embodiment of FIG. 2;
FIG. 6 is a schematic diagram of an infrared heat map model detection provided by an embodiment of the present invention;
fig. 7 is a schematic structural diagram of a face temperature detection device according to an embodiment of the present invention;
FIG. 8 is a schematic diagram of another configuration provided by the training module of an embodiment of the present invention;
FIG. 9 is a schematic diagram of one configuration provided by the training sub-module of the embodiment of FIG. 8;
FIG. 10 is a schematic diagram of a configuration provided by the first acquisition sub-module in the embodiment of FIG. 8;
FIG. 11 is a schematic diagram of a configuration provided by the second acquisition sub-module in the embodiment of FIG. 8;
fig. 12 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a flowchart of a face temperature detection method according to an embodiment of the present invention, as shown in fig. 1, including the following steps:
step 101, acquiring a face image to be recognized.
The face image to be identified is a face image of a target person to be identified in the application scene. The application scene can be a hospital, a district, a campus, a unit, etc., or other application scenes needing to detect the body temperature of the target person. More specifically, the system can be used in an access control system of a hospital, a community, a campus, an entrance or an exit of a unit, and the like. In this embodiment, an entrance guard system for an entrance of an application scene in which body temperature needs to be detected will be mainly described as an example. The face image to be identified can be a face image of a target person entering and exiting a cell, a campus, a unit entrance or an exit. The face image to be identified includes a face infrared thermal image, which may be referred to as an infrared thermal image. The target person may be one or more. The number of the face images to be identified can be one or more, and when the number of the target persons is multiple, the number of the corresponding face images to be identified is also multiple. Corresponding human face infrared heat maps are also multiple.
The human face infrared heat map can be acquired by infrared heat map imaging equipment. The facial infrared thermography imaging apparatus may be a thermal infrared imager or the like. The facial infrared heat map can be acquired by disposing an infrared thermal imager of a hospital, at a district, a campus, a unit entrance or an exit and aiming at the face of a target person, and is used for detecting the face temperature of the target person and further detecting the body temperature of the target person. The infrared thermal imager can also be called as an infrared thermal imager, and the infrared thermal imager receives an infrared radiation energy distribution pattern of a detected target by using an infrared detector and an optical imaging objective lens and reflects the infrared radiation energy distribution pattern onto a photosensitive element of the infrared detector, so that an infrared thermal image is obtained, and the thermal image corresponds to a thermal distribution field on the surface of an object.
The face image to be identified may further include a face R (red) G (green) B (blue) image, which is used for identifying face identity information of the target person. The face RGB image may be an image frame of a still image or a dynamic video of a face of a target person acquired by a face image acquisition device (camera or the like) disposed at a cell portal, a campus portal, or a unit portal. Of course, the camera may be a face camera, a portrait camera or a common RGB image camera.
It should be understood that the infrared thermal imaging device is arranged on the same horizontal line as the face camera, for example, the infrared thermal imaging device is arranged up and down on the same horizontal line, left and right on the same horizontal line, and left and right on the same horizontal line. Therefore, the face angles of the target personnel aligned by the infrared thermal imaging instrument and corresponding to the face cameras are basically consistent. And the aim that the infrared thermal imaging instrument is aimed at is further ensured to be the same person as the aim that the face camera is aimed at.
Specifically, when a target person needs to enter and exit the entrance guard, the target person is aligned through the infrared thermal imager, a face infrared heat map corresponding to the target person is acquired, and then a face image to be identified of the target person is obtained. Of course, the target person can be aligned by the face camera, the RGB image of the target person is collected, and the face image to be recognized is formed by the face infrared thermal pattern.
Step 102, inputting the face image to be recognized into a pre-trained face temperature detection model for recognition, and obtaining face identity information and face temperature information corresponding to the face image to be recognized.
The pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, wherein the face sample set comprises face infrared heat map samples.
The pre-trained face temperature detection model is used for detecting the face temperature of a target person, and then detecting the body temperature of the target person. The face temperature detection model may be a face temperature detection neural network, a face temperature detection convolutional network, or the like.
The face identity information is the face identity information of the target person corresponding to the face image to be identified. The face identity information may be information including name, face image, document number, contact, home address, work unit, physical condition, etc. The face identity information may be stored in an identity database. The identity database may be provided by an authoritative entity or institution for storing the identity information of the face of the target person. The identity database may store face identity information for a plurality of persons.
The face temperature information may be face temperature information of a target person corresponding to the face image to be identified, and the corresponding face temperature information may be body temperature information of the target person corresponding to the face image to be identified. The face temperature information may include a temperature size, a time of measuring the temperature, etc., for example, a temperature detected by person a at 12 points is 36.1C deg.c, etc.
The face sample set is a face sample set formed by face samples for training a pre-trained face detection model. The face sample set may include a face infrared thermal pattern book, and may also include a first face RGB map sample. The face sample set may be stored in a local preset face sample set database, or in an external memory, such as a usb disk, etc. When the face sample set is used, the face sample set can be directly called from a local preset face sample set database or from an external memory. The face sample set in the remote face sample set database may also be invoked remotely via a network link. The face sample set comprises a plurality of face samples, specifically, a face infrared thermal pattern book of a plurality of people can be included, and a first face RGB pattern book of a plurality of people can be also included. Of course, the first face RGB pattern book is used together with the face infrared thermal pattern book as input of a pre-trained face detection model to train the pre-trained face detection model to obtain the face RGB pattern book of the face temperature detection model. Of course, the pre-trained face detection model can be trained without the first face RGB pattern book serving as the input of the pre-trained face detection model, and the pre-trained face temperature detection model can be obtained by training the pre-trained face detection model only by using the face infrared heat map serving as the input of the pre-trained face detection model. It should be noted that the pre-trained face detection model may be a pre-trained face detection neural network or a convolutional network.
Specifically, after the face image to be recognized is obtained, the obtained face image to be recognized can be used as input data of a face temperature detection model trained in advance, and face identity information of a target person corresponding to the face image to be recognized and face temperature information of the target person corresponding to the face image to be recognized are recognized through the face temperature detection model trained in advance. More specifically, the facial infrared heat map is used as input data of a pre-trained facial temperature detection model, and then the facial infrared heat map is subjected to facial detection through the pre-trained facial temperature detection model, and corresponding facial temperature is given. Thus, when a target person enters and exits a community, a campus, a unit and the like, face identity information of the target person and body temperature corresponding to the target person can be identified. Of course, when the face image to be recognized comprises images of a plurality of persons, the images of the plurality of persons are input into a face temperature detection model trained in advance to detect faces and give corresponding temperatures, so that a plurality of groups of face identity information and face temperature information corresponding to the plurality of persons can be obtained.
In an embodiment of the invention, the identified face identity information and face temperature information can be analyzed, and when a target person with abnormal face temperature is detected, an alarm can be sent out. The method is convenient for reminding workers of related working units to take corresponding management measures, and particularly monitors and manages the temperature abnormality personnel.
In the embodiment of the invention, the face image to be identified is obtained and comprises a face infrared heat map; inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized; the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples. Therefore, the identity recognition and the body temperature test of the target personnel can be completed directly through collecting the facial infrared heat map in the facial image of the target personnel, and the body temperature condition of the target personnel can be monitored and managed conveniently. Furthermore, the problems of high labor cost and low temperature detection working efficiency caused by the fact that in the prior art, intelligent body temperature identification cannot be achieved, related staff can only be assigned to measure temperature and inquire identity information for each target person and register the temperature are solved. And further reduces the labor cost and improves the working efficiency of temperature detection.
Referring to fig. 2, fig. 2 is a flowchart of a training method of a face temperature detection model according to an embodiment of the present invention, where the training method of the face temperature detection model includes the steps of:
step 201, a face sample set is obtained.
Specifically, referring to fig. 3, fig. 3 is a flowchart of a method provided in step 201 in the embodiment of fig. 2, where step 201 includes:
step 301, a first face RGB pattern set is obtained.
The first face RGB pattern set may include a plurality of first face RGB pattern samples, and each first face RGB pattern set includes a first face identity tag. The first face identity tag may be a tag for representing face identity information of a target person corresponding to the first face RGB image sample. The first face identity tag may be a name, a certificate number, a face feature, or the like. The first face RGB pattern set may be stored in a preset face RGB pattern database, or may be acquired in real time by a corresponding face RGB pattern acquisition device. The first face RGB map uniformly and correspondingly comprises a unique first face identity label.
Specifically, a plurality of first face RGB diagrams are acquired. And inputting the plurality of first face RGB images into a face RGB primary feature extraction model to extract the face RGB image primary features so as to obtain a plurality of first face RGB image primary features. And obtaining a first face RGB pattern book set based on the plurality of first face RGB pattern primary features.
More specifically, the first face RGB image primary features of each first face RGB image may be extracted through obtaining first face RGB images corresponding to a plurality of target persons, and the first face RGB image primary features of each first face RGB image may be extracted through a face RGB primary feature extraction model, and stored, and the first face RGB image set may be formed based on the first face RGB image primary features corresponding to the plurality of target persons. When the model training is carried out, the primary characteristics of the first face RGB image can be directly used for training, the whole first face RGB image input model is not required to be trained, and the input data size of the first face RGB image can be reduced. The primary feature extraction model of the face RGB image can promote the dependence of the model on the face RGB image.
Step 302, acquiring a human face infrared heat pattern book.
The above-mentioned facial infrared heat pattern book collection can include a plurality of facial infrared heat pattern books to every facial infrared heat pattern book all includes facial temperature label. The face temperature label may be a label for representing face temperature information of a target person corresponding to the face infrared heat map. The face temperature label may be a degree of temperature. The human face infrared heat pattern book set can be stored in a preset human face infrared heat pattern book database, or can be acquired in real time through human face infrared heat pattern acquisition equipment.
Specifically, a plurality of human face infrared heat maps are obtained. Inputting the plurality of facial infrared heat maps into a facial infrared heat map primary feature extraction model to extract facial infrared heat map primary features so as to obtain a plurality of facial infrared heat map primary features. And obtaining the facial infrared heat pattern book set based on the primary characteristics of the plurality of facial infrared heat patterns.
More specifically, the facial infrared heat patterns corresponding to the target persons can be obtained, the facial infrared heat pattern primary characteristics of each facial infrared heat pattern are extracted through a facial infrared heat pattern primary characteristic extraction model, stored and a facial infrared heat pattern book set is formed based on the facial infrared heat pattern primary characteristics corresponding to the target persons. Therefore, when the model is trained, the primary characteristics of the human face infrared heat map can be directly used for training, the whole human face infrared heat map is not required to be input into the model for training, and the input data quantity of the human face infrared heat map can be reduced. The primary feature extraction model of the human face infrared heat map can promote the dependence of the model on the human face infrared heat map. And the primary feature dimensions of the first face RGB image and the face infrared heat map can be ensured to be consistent.
Step 303, forming a face sample set according to the first face RGB image sample set and the face infrared heat map sample set.
Specifically, after the first face RGB pattern set and the face infrared thermal pattern set are obtained respectively, the two sample sets are combined together to form a face sample set.
The human face infrared heat map has a certain similarity with the human face RGB map in characteristic expression, such as outline and the like.
Step 202, a pre-trained face detection model is obtained.
Specifically, referring to fig. 4, fig. 4 is a flowchart of a method provided in step 202 in the embodiment of fig. 2, where step 202 includes:
step 401, a second face RGB pattern set is obtained, and the second face RGB pattern set includes a second face identity tag.
Step 402, inputting the second RGB image set into a pre-training model for training, so that the pre-training model learns the prediction of the face identity information according to the second RGB image set, to obtain a pre-trained face detection model.
The second face RGB pattern set may include a plurality of second face RGB pattern sets, and each second face RGB pattern set includes a second face identity tag. The second face identity tag may be a tag for representing face identity information of a target person to which the second face RGB pattern corresponds. The second face identity tag may be a name, a certificate number, a face feature, or the like. The second face RGB pattern book set may be stored in a preset face RGB pattern book database, or may be acquired in real time by a corresponding face RGB image acquisition device. The second face RGB map uniformly and correspondingly comprises a unique second face identity label.
Note that the second set of face RGB patterns may be the same as or different from the first set of face RGB patterns. The second set of face RGB patterns may also include face RGB samples of a plurality of persons.
The pre-training model may be a pre-set model that has not been trained yet. The pre-training model may be a pre-training neural network or a convolutional network. The pre-training model is not able to identify or detect any information when the pre-training model has not been trained. The convolution kernel of the pre-training model may be a convolution kernel set to 3 x 3.
Specifically, after the second face RGB pattern set is obtained, the second face RGB pattern set is input into a pre-training model to perform face recognition training, so that the pre-training model can learn the prediction of face identity information according to the second face RGB pattern set, and the pre-training model is trained into a face detection model.
It should be appreciated that when the second set of face RGB patterns is different from the first set of face RGB pattern samples, the pre-trained face detection model is described as being trained only by the second set of face RGB patterns.
When the second face RGB pattern set is the same as the first face RGB pattern sample set, it may also be indicated that the pre-trained face detection model is obtained by training the first face RGB pattern set.
Step 203, inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of the first face RGB image sample in the training process, so that the pre-trained face detection model learns the prediction of the face temperature information according to the face infrared thermal pattern, and the pre-trained face temperature detection model is obtained.
The input weight of the first face RGB image sample is the ratio of the first face RGB image sample in the face sample set, which may be expressed as the first face RGB image sample number/the total face sample number, when the face sample set only includes the first face RGB image sample and the human infrared heat image sample, the input weight of the first face RGB image sample may be expressed as the first face RGB image sample number/(the human infrared heat image sample number+the first face RGB image sample number), for example, the existing face sample set has 100 face samples, where there are 40 first face RGB image samples and 60 human infrared heat image samples, then the 100 face samples in the face sample set are trained as a pre-trained face detection model, and at this time, the input weight of the first face RGB image sample is 40/100, and the input weight of the first face RGB is 2/5. The input weights may also be represented by weight coefficients.
The above-mentioned reducing the input weight of the first face RGB image sample in the training process may be that, when the face sample set is input into the pre-trained face detection model for training, that is, in the iterative process, the specific gravity of the first face RGB image sample in the face sample set is continuously reduced by taking a preset reduction value as a reduction unit.
Specifically, referring to fig. 5, fig. 5 is a flowchart of a method provided in step 203 of the embodiment of fig. 2, step 203 comprising:
and 501, performing branch reduction processing on the pre-trained face detection model so as to enable the pre-trained face detection model to adapt to single-channel input of the face infrared heat map sample.
Wherein the above-mentioned branch-reducing process can be a dimension-reducing process, for example, the pre-trained face detection model includes a convolution kernel of 3 x 3, the 3 x 3 convolution kernel is reduced in dimension to become a 3 x 1 convolution network.
Specifically, the first face RGB diagram in the first face RGB diagram sample is a three-channel, which is an R channel, a G channel, and a B channel, respectively. For this purpose, the pre-trained face detection model may be a 3 x 3 convolution kernel, each layer of convolution kernel convolves the R channel, the G channel and the B channel respectively. In order to better preserve the parameters of the pre-training model and reduce the size of the model, the pre-trained face detection model is subjected to a branch reduction (dimension reduction) process, and 2/3 shared smaller channels are deleted to adapt to the single-channel face infrared heat map, for example, a convolution kernel of 3 x 3 is reduced in dimension and then becomes a convolution kernel of 3 x 1 to adapt to the single channel of the convolution face infrared heat map. In the process of branch reduction, the gradient size corresponding to each convolution kernel is counted according to back propagation, and then the convolution kernel with the smaller gradient is deleted by 2/3. Because the infrared heatmap can provide less information than the RGB map and the model size is correspondingly smaller.
Step 502, inputting the face sample set to a training face detection model after the subtraction, and reducing the input weight of the first face RGB image sample to zero in the training process.
Specifically, training the pre-trained face detection model after performing the branch reduction processing on the pre-trained face detection model, and reducing the input weight of the first face RGB map sample. And until the input weight of the first face RGB image sample is reduced to zero, and then the face temperature detection model is obtained.
The step of reducing the input weight of the first face RGB image sample to zero in the training process may be that when the first face sample set is input into a pre-trained face detection model to perform training, that is, continuously reducing the specific gravity of the first face RGB image sample in the face sample set by taking a preset reduction value as a reduction unit in the iteration process until the input weight of the first face RGB image sample is reduced to zero, so that only an infrared heat image is in the face sample set, and after the pre-trained face detection model is converted into the face temperature detection model, face identity information and face temperature information corresponding to the face infrared heat image can be identified through the face temperature detection model. Specifically, a preset descent value is preset, so that the input weight of the first face RGB image sample is reduced by taking the preset descent value as a unit until the input weight of the first face RGB image sample is reduced to zero, and training of the face temperature detection model can be finished.
The preset drop value is set to 1/10, the input weight of the first face RGB pattern book and the input weight of the face infrared heat pattern book when the first face RGB pattern book and the face infrared heat pattern book are input into the pre-trained face detection model for training are 1/2, the input weight of the face infrared heat map at the moment is also 1/2, and the pre-trained face detection model at the moment outputs corresponding face identity information and face temperature information. When the second time of the first face RGB image sample and the face infrared thermal pattern book are input into a pre-trained face detection model for training, the input weight of the first face RGB image sample is reduced to 4/10 based on the first input weight by taking a preset reduction value of 1/10 as a unit, the input weight of the face infrared thermal pattern at the moment is also 6/10, and the pre-trained face detection model outputs corresponding face identity information and face temperature information at the moment.
When the third time of the first face RGB pattern book and the face infrared thermal pattern book are input into a pre-trained face detection model for training, the input weight of a first face RGB pattern sample is reduced to 3/10 based on the second input weight by taking a preset reduction value of 1/10 as a unit, the input weight of the face infrared thermal pattern at the moment is 7/10, and the pre-trained face detection model outputs corresponding face identity information and face temperature information at the moment.
After training for N times (N is an integer greater than 3), the input weight of the last time of the first face RGB pattern and the input weight of the face infrared heat pattern are 0 when the last time of the first face RGB pattern and the face infrared heat pattern are input into a pre-trained face detection model for recognition detection, at this time, the input weight of the face infrared heat pattern samples in the face sample set becomes 1, that is, the face samples in the face sample set are all face infrared heat patterns, and the first face RGB samples do not exist in the face sample set. Under the condition that the first face RGB pattern book does not exist in the face sample set, the face detection model trained in advance can also recognize face identity information and face temperature information corresponding to the corresponding face infrared heat map through the face infrared heat pattern book. When the input weight of the first face RGB pattern book is reduced to zero, the pre-trained face detection model is trained into a face temperature detection model, and the face temperature detection model can identify face identity information and face temperature information corresponding to the face infrared heat map only through the face infrared heat pattern book.
It should be noted that, the smaller the preset drop value is, the slower the input weight of the first face RGB image sample is, the more training times on the pre-trained face detection model will be, and parameters of the pre-trained face detection model can be adjusted more accurately to the model size, so that a better face temperature detection model can be obtained.
Referring to fig. 6, fig. 6 is an infrared heat map model detection schematic diagram provided by the embodiment of the present invention, and a specific engineering flow is that a common RGB image and an infrared heat map are used as input at first, a pre-trained face detection model is trained, and identity information and temperature information corresponding to each face are provided. The weight of the RGB image is gradually reduced in the training process, the weight of the RGB input is reduced to 0 after the training is completed, and the functions of identification and temperature detection of target personnel can be completed only by using an infrared heat map.
In order to improve the performance of the network, the RGB images and the infrared heat maps are not simply overlapped, but the primary characteristics of the two images are extracted by respectively designing two small convolution network modules, and the primary characteristic extraction module can improve the dependence of the network on the RGB images and ensure the consistency of the characteristic dimensions of the two images. In the training process, the contribution of the RGB image to the network is controlled by adjusting the weight coefficient, and the training weight coefficient is reduced to 0.
In an embodiment of the invention, when the first face RGB pattern set is different from the second face RGB pattern set, the face detection model is obtained by training the pre-training model through the second face RGB pattern set, so that the face detection model can identify corresponding face identity information based on the second face RGB pattern. And then, the first face RGB image sample set and the face infrared heat map are used as the input of the pre-trained face detection model together to carry out face recognition and temperature detection training on the pre-trained face detection model, so that the pre-trained face detection model is trained into a face temperature detection model. The input weight of the first face RGB image sample is continuously reduced in the training process of the face temperature detection model, and corresponding face identity information and face temperature information are given out, so that the face temperature detection model can identify the corresponding face identity information and face temperature information only according to the face infrared heat map. And the detection capability of the face temperature detection model is improved.
In another embodiment of the present invention, when the first face RGB pattern set is the same as the second face RGB pattern set, the pre-training model is trained by the first face RGB pattern set to obtain a face detection model, and then the pre-trained face detection model is subjected to face recognition and temperature detection training by using the first face RGB pattern sample set and the face infrared heat map together as inputs of the pre-trained face detection model, so that the pre-trained face detection model is trained into the face temperature detection model. The input weight of the first face RGB image sample is continuously reduced in the training process of the face temperature detection model, and corresponding face identity information and face temperature information are given out, so that the face temperature detection model can identify the corresponding face identity information and face temperature information only according to the face infrared heat map. And the detection capability of the face temperature detection model is improved.
In the embodiment of the invention, the pre-trained face detection model is trained by using the face sample set in advance, so that the pre-trained face detection model learns the identity information of the identified face and the detected face temperature information according to the face sample set, and the identity identification and the temperature detection of the face image to be identified can be carried out by directly using the face temperature detection model through a face infrared heat map. The body temperature condition of management target personnel is convenient to monitor. Furthermore, the problems of high labor cost and low temperature detection working efficiency caused by the fact that in the prior art, intelligent body temperature identification cannot be achieved, related staff can only be assigned to measure temperature and inquire identity information for each target person and register the temperature are solved. And further reduces the labor cost and improves the working efficiency of temperature detection.
Referring to fig. 7, fig. 7 is a schematic structural diagram of a face temperature detection device according to an embodiment of the present invention, and as shown in fig. 7, the face temperature detection device 600 includes:
the acquiring module 601 is configured to acquire a face image to be identified, where the face image to be identified includes a face infrared heat map;
the recognition module 602 is configured to input a face image to be recognized into a pre-trained face temperature detection model for recognition, so as to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, wherein the face sample set comprises face infrared heat map samples.
Optionally, as shown in fig. 8, the face sample set further includes a first face RGB pattern book, and the training module 603 includes:
a first acquisition submodule 6031 for acquiring a face sample set;
a second acquiring sub-module 6032, configured to acquire a pre-trained face detection model;
the training submodule 6033 is configured to input the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reduce an input weight of the first face RGB image sample in a training process, so that the pre-trained face detection model learns a prediction of face temperature information according to the face infrared thermal pattern, so as to obtain the pre-trained face temperature detection model.
Optionally, as shown in fig. 9, the training sub-module 6033 includes:
the branch-reducing processing unit 60331 is used for performing branch-reducing processing on the pre-trained face detection model so as to enable the pre-trained face detection model to adapt to single-channel input of the face infrared heat map sample;
the first training unit 60332 is configured to input the face sample set to the training model of the training face after the training process, and reduce the input weight of the first face RGB image sample to zero in the training process.
Alternatively, as shown in fig. 10, the first acquisition submodule 6031 includes:
a first obtaining unit 60311, configured to obtain a first RGB human face pattern book, where the first RGB human face pattern book includes a first face identity tag;
a second acquiring unit 60312, configured to acquire a set of facial infrared thermal patterns, where the set of facial infrared thermal patterns includes a facial temperature tag;
a forming unit 60313 is configured to form a face sample set according to the first face RGB image sample set and the face infrared heat map sample set.
Optionally, the first obtaining unit includes:
the first acquisition subunit is used for acquiring a plurality of first face RGB images;
the first extraction subunit is used for inputting the plurality of first face RGB images into a face RGB primary feature extraction model to extract the face RGB image primary features so as to obtain a plurality of first face RGB image primary features;
A first forming subunit configured to form a first face RGB pattern set based on a plurality of first face RGB map primary features.
Optionally, the second obtaining unit includes:
the second acquisition subunit is used for acquiring a plurality of facial infrared heat maps;
the second extraction subunit is used for inputting the plurality of facial infrared heat maps into a facial infrared heat map primary feature extraction model to extract facial infrared heat map primary features so as to obtain a plurality of facial infrared heat map primary features;
the second forming subunit forms a face infrared thermal pattern book set based on the plurality of face infrared thermal pattern primary features.
Optionally, as shown in fig. 11, the second acquisition sub-module 6032 includes:
a third obtaining unit 60321, configured to obtain a second face RGB pattern book, where the second face RGB pattern book includes a second face identity tag;
the second training unit 60322 is configured to input the second RGB image set into the pre-training model for training, so that the pre-training model learns the prediction of the face identity information according to the second RGB image set, to obtain a pre-trained face detection model.
It should be noted that, the face temperature detection apparatus 600 provided in the embodiment of the present invention may be applied to a mobile phone, a monitor, a computer, a server, and other devices that need to perform face temperature detection.
The face temperature detection device 600 provided by the embodiment of the invention can realize each process realized by the face temperature detection method in the method embodiment, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
Referring to fig. 12, fig. 12 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 12, including: memory 702, processor 701, and a computer program stored on memory 702 and executable on processor 701, wherein:
the processor 701 is configured to call a computer program stored in the memory 702, and perform the following steps:
acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map;
inputting the face image to be recognized into a pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, wherein the face sample set comprises face infrared heat map samples.
Optionally, the face sample set further includes a first face RGB chart sample, and the training method performed by the processor 701 for the pre-trained face temperature detection model includes:
Acquiring a face sample set;
acquiring a pre-trained face detection model;
and inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of the first face RGB image sample in the training process so that the pre-trained face detection model learns the prediction of the face temperature information according to the face infrared thermal pattern, thereby obtaining the pre-trained face temperature detection model.
Optionally, the inputting the face sample set into the pre-trained face detection model by the processor 701 is performed to train the pre-trained face detection model, and the reducing the input weight of the first face RGB image sample in the training process includes:
performing branch reduction processing on the pre-trained face detection model so as to enable the pre-trained face detection model to adapt to single-channel input of the face infrared heat map sample;
and inputting the face sample set into the training model of the face after the branch is subtracted, training, and reducing the input weight of the first face RGB image sample to zero in the training process.
Optionally, the acquiring a face sample set performed by the processor 701 includes:
Acquiring a first face RGB pattern book set, wherein the first face RGB pattern book comprises a first face identity tag;
acquiring a human face infrared heat pattern book set, wherein the human face infrared heat pattern book comprises a human face temperature label;
and forming a face sample set according to the first face RGB image sample set and the face infrared heat map sample set.
Optionally, the acquiring the first face RGB pattern set performed by the processor 701 includes:
acquiring a plurality of first face RGB images;
inputting a plurality of first face RGB images into a face RGB primary feature extraction model to extract face RGB image primary features so as to obtain a plurality of first face RGB image primary features;
and obtaining a first face RGB pattern book set based on the plurality of first face RGB pattern primary features.
Optionally, the acquiring the set of infrared thermal patterns of the face executed by the processor 701 includes:
acquiring a plurality of human face infrared heat maps;
inputting the plurality of facial infrared heat maps into a facial infrared heat map primary feature extraction model to extract facial infrared heat map primary features so as to obtain a plurality of facial infrared heat map primary features;
and obtaining the facial infrared heat pattern book set based on the primary characteristics of the plurality of facial infrared heat patterns.
Optionally, the acquiring the pre-trained face detection model performed by the processor 701 includes:
Acquiring a second face RGB pattern book set, wherein the second face RGB pattern book comprises a second face identity tag;
and inputting the second face RGB pattern set into a pre-training model for training, so that the pre-training model learns the prediction of face identity information according to the second face RGB pattern set to obtain a pre-trained face detection model.
The electronic device 700 may be applied to a mobile phone, a monitor, a computer, a server, or the like, which needs to detect a face temperature.
The electronic device 700 provided in the embodiment of the present invention can implement each process implemented by the face temperature detection method in the above embodiment of the method, and can achieve the same beneficial effects, so that repetition is avoided, and no description is repeated here.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements each process of the face temperature detection method provided by the embodiment of the invention, and can achieve the same technical effects, so that repetition is avoided, and no further description is provided herein.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.
Claims (9)
1. The face temperature detection method is characterized by comprising the following steps of:
acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map;
acquiring a face sample set;
acquiring a pre-trained face detection model;
inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of a first face RGB (red, green and blue) image sample in the training process so that the pre-trained face detection model learns the prediction of the face temperature information according to the face infrared thermal pattern book to obtain the pre-trained face temperature detection model;
inputting the face image to be recognized into the pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples.
2. The method of claim 1, wherein inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model and reducing the input weight of the first face RGB pattern book during training comprises:
performing branch reduction processing on the pre-trained face detection model so as to enable the pre-trained face detection model to adapt to single-channel input of the face infrared heat map sample;
and inputting the face sample set into the training model of the face detection after the branch is subtracted to train, and reducing the input weight of the first face RGB image sample to zero in the training process.
3. The method of claim 1, wherein the obtaining the set of face samples comprises:
acquiring a first face RGB pattern book set, wherein the first face RGB pattern book comprises a first face identity tag;
acquiring a human face infrared heat pattern book set, wherein the human face infrared heat pattern book comprises a human face temperature label;
and forming the face sample set according to the first face RGB image sample set and the face infrared heat map sample set.
4. The method of claim 3, wherein the obtaining the first set of face RGB patterns comprises:
acquiring a plurality of first face RGB images;
inputting the plurality of first face RGB images into a face RGB primary feature extraction model to extract face RGB image primary features so as to obtain a plurality of first face RGB image primary features;
and obtaining the first face RGB pattern book set based on the plurality of first face RGB pattern primary features.
5. The method of claim 3, wherein the acquiring the set of infrared thermal patterns of the face comprises:
acquiring a plurality of human face infrared heat maps;
inputting the plurality of facial infrared heat maps into a facial infrared heat map primary feature extraction model to extract facial infrared heat map primary features so as to obtain a plurality of facial infrared heat map primary features;
and obtaining the facial infrared thermal pattern book set based on the primary characteristics of the plurality of facial infrared thermal patterns.
6. The method of claim 1, wherein the obtaining a pre-trained face detection model comprises:
acquiring a second face RGB pattern book set, wherein the second face RGB pattern book comprises a second face identity tag;
and inputting the second human face RGB pattern set into a pre-training model for training, so that the pre-training model learns the prediction of human face identity information according to the second human face RGB pattern set to obtain the pre-trained human face detection model.
7. A face temperature detection apparatus, the apparatus comprising:
the acquisition module is used for acquiring a face image to be identified, wherein the face image to be identified comprises a face infrared heat map;
the first acquisition submodule is used for acquiring a face sample set;
the second acquisition sub-module is used for acquiring a pre-trained face detection model;
the training sub-module is used for inputting the face sample set into a pre-trained face detection model to train the pre-trained face detection model, and reducing the input weight of a first face RGB image sample in the training process so that the pre-trained face detection model learns the prediction of the face temperature information according to the face infrared thermal pattern book to obtain the pre-trained face temperature detection model;
the recognition module is used for inputting the face image to be recognized into the pre-trained face temperature detection model for recognition to obtain face identity information and face temperature information corresponding to the face image to be recognized;
the pre-trained face temperature detection model is obtained by training the pre-trained face detection model according to a face sample set, and the face sample set comprises face infrared heat map samples.
8. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the face temperature detection method according to any one of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps in the face temperature detection method according to any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010229946.0A CN113449567B (en) | 2020-03-27 | 2020-03-27 | Face temperature detection method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010229946.0A CN113449567B (en) | 2020-03-27 | 2020-03-27 | Face temperature detection method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113449567A CN113449567A (en) | 2021-09-28 |
CN113449567B true CN113449567B (en) | 2024-04-02 |
Family
ID=77807718
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010229946.0A Active CN113449567B (en) | 2020-03-27 | 2020-03-27 | Face temperature detection method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113449567B (en) |
Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2012235415A (en) * | 2011-05-09 | 2012-11-29 | Nec Software Kyushu Ltd | Image processing system, febrile person identifying method, image processing apparatus, and control method and program thereof |
JP2012254222A (en) * | 2011-06-09 | 2012-12-27 | Teikyo Heisei Univ | Device for human dynamic face temperature change measurement by thermography and measuring method, and calibrator used for the same |
CN103493112A (en) * | 2011-02-22 | 2014-01-01 | 菲力尔***公司 | Infrared sensor systems and methods |
CN106599797A (en) * | 2016-11-24 | 2017-04-26 | 北京航空航天大学 | Infrared face identification method based on local parallel nerve network |
CN106845450A (en) * | 2017-02-22 | 2017-06-13 | 武汉科技大学 | Dark surrounds face identification method based near infrared imaging Yu deep learning |
KR101754152B1 (en) * | 2016-01-25 | 2017-07-19 | 주식회사 에스원 | Thermal Patient Monitering System by Using Multiple Band Camera and Method thereof |
CN108875752A (en) * | 2018-03-21 | 2018-11-23 | 北京迈格威科技有限公司 | Image processing method and device, computer readable storage medium |
CN109308436A (en) * | 2017-07-28 | 2019-02-05 | 西南科技大学 | A kind of living body faces recognition methods based on active infrared video |
CN109614910A (en) * | 2018-12-04 | 2019-04-12 | 青岛小鸟看看科技有限公司 | A kind of face identification method and device |
CN109846463A (en) * | 2019-03-04 | 2019-06-07 | 武汉迅检科技有限公司 | Infrared face temp measuring method, system, equipment and storage medium |
CN110032915A (en) * | 2018-01-12 | 2019-07-19 | 杭州海康威视数字技术股份有限公司 | A kind of human face in-vivo detection method, device and electronic equipment |
CN110276248A (en) * | 2019-05-10 | 2019-09-24 | 杭州电子科技大学 | A kind of facial expression recognizing method based on sample weights distribution and deep learning |
CN110543645A (en) * | 2019-09-04 | 2019-12-06 | 网易有道信息技术(北京)有限公司 | Machine learning model training method, medium, device and computing equipment |
CN110797125A (en) * | 2018-08-01 | 2020-02-14 | 深圳云天励飞技术有限公司 | Human health detection system |
-
2020
- 2020-03-27 CN CN202010229946.0A patent/CN113449567B/en active Active
Patent Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103493112A (en) * | 2011-02-22 | 2014-01-01 | 菲力尔***公司 | Infrared sensor systems and methods |
JP2012235415A (en) * | 2011-05-09 | 2012-11-29 | Nec Software Kyushu Ltd | Image processing system, febrile person identifying method, image processing apparatus, and control method and program thereof |
JP2012254222A (en) * | 2011-06-09 | 2012-12-27 | Teikyo Heisei Univ | Device for human dynamic face temperature change measurement by thermography and measuring method, and calibrator used for the same |
KR101754152B1 (en) * | 2016-01-25 | 2017-07-19 | 주식회사 에스원 | Thermal Patient Monitering System by Using Multiple Band Camera and Method thereof |
CN106599797A (en) * | 2016-11-24 | 2017-04-26 | 北京航空航天大学 | Infrared face identification method based on local parallel nerve network |
CN106845450A (en) * | 2017-02-22 | 2017-06-13 | 武汉科技大学 | Dark surrounds face identification method based near infrared imaging Yu deep learning |
CN109308436A (en) * | 2017-07-28 | 2019-02-05 | 西南科技大学 | A kind of living body faces recognition methods based on active infrared video |
CN110032915A (en) * | 2018-01-12 | 2019-07-19 | 杭州海康威视数字技术股份有限公司 | A kind of human face in-vivo detection method, device and electronic equipment |
CN108875752A (en) * | 2018-03-21 | 2018-11-23 | 北京迈格威科技有限公司 | Image processing method and device, computer readable storage medium |
CN110797125A (en) * | 2018-08-01 | 2020-02-14 | 深圳云天励飞技术有限公司 | Human health detection system |
CN109614910A (en) * | 2018-12-04 | 2019-04-12 | 青岛小鸟看看科技有限公司 | A kind of face identification method and device |
CN109846463A (en) * | 2019-03-04 | 2019-06-07 | 武汉迅检科技有限公司 | Infrared face temp measuring method, system, equipment and storage medium |
CN110276248A (en) * | 2019-05-10 | 2019-09-24 | 杭州电子科技大学 | A kind of facial expression recognizing method based on sample weights distribution and deep learning |
CN110543645A (en) * | 2019-09-04 | 2019-12-06 | 网易有道信息技术(北京)有限公司 | Machine learning model training method, medium, device and computing equipment |
Non-Patent Citations (2)
Title |
---|
Thermal Face Recognition Under Temporal Variation Conditions;G. Hermosilla Vigneau等;IEEE Access;第9663-9672页 * |
基于视频的轨道交通车站乘客口罩佩戴检测及测温技术;谢征宇等;中国铁路;第2020年卷(第03期);126-131 * |
Also Published As
Publication number | Publication date |
---|---|
CN113449567A (en) | 2021-09-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2020010785A1 (en) | Classroom teaching cognitive load measuring system | |
Lynnerup et al. | Gait as evidence | |
CN111814661A (en) | Human behavior identification method based on residual error-recurrent neural network | |
CN105930822A (en) | Human face snapshot method and system | |
CN107977656A (en) | A kind of pedestrian recognition methods and system again | |
Andrushia A et al. | Deep learning based thermal crack detection on structural concrete exposed to elevated temperature | |
CN108121961A (en) | Inspection Activity recognition method, apparatus, computer equipment and storage medium | |
CN109410252A (en) | A kind of thermal imaging equipment | |
CN116129353B (en) | Method and system for intelligent monitoring based on image recognition | |
WO2024051067A1 (en) | Infrared image processing method, apparatus, and device, and storage medium | |
CN114098636A (en) | Human health state management method and system and electronic equipment | |
CN116311400A (en) | Palm print image processing method, electronic device and storage medium | |
CN112525355A (en) | Image processing method, device and equipment | |
CN109997199A (en) | Tuberculosis inspection method based on deep learning | |
CN111652314A (en) | Temperature detection method and device, computer equipment and storage medium | |
CN116797977A (en) | Method and device for identifying dynamic target of inspection robot and measuring temperature and storage medium | |
CN114170690A (en) | Method and device for living body identification and construction of living body identification model | |
CN113449567B (en) | Face temperature detection method and device, electronic equipment and storage medium | |
CN110619365B (en) | Method for detecting falling water | |
CN115147705B (en) | Face copying detection method and device, electronic equipment and storage medium | |
CN115273245A (en) | Living body detection method, living body detection device and computer-readable storage medium | |
Cui et al. | Contrast threshold adaptive adjustment algorithm for remote sensing image matching | |
Espinola et al. | Atmospheric turbulence and sensor system effects on biometric algorithm performance | |
Surabhi et al. | Infrared fever body identification using shape and temperature filters | |
Tao et al. | Concealed multiscale feature extraction network for automatic four-bar target detection in infrared imaging |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |