A method and system for detecting facial information and body temperature information based on a causal model

Abstract

The present invention relates to the technical field of detection of facial information and body temperature information, and specifically to a method and system for detecting facial information and body temperature information based on a causal model. The facial information and body temperature information detection method includes establishing a facial detection causal model and a human body temperature detection causal model, judging the input facial information according to the facial detection causal model, and collecting the human body temperature when the result meets the facial features; judging the collected human body temperature according to the human body temperature detection causal model, and obtaining the normal or abnormal situation of the human body temperature. The beneficial effect of the present invention is that, compared with the prior art, the present invention combines facial recognition with body temperature detection, and combines the algorithm of the causal model to improve the accuracy of facial recognition and the accuracy of temperature detection, thereby realizing rapid detection of one person, one face and one temperature, while having high accuracy and reducing the influence of external influencing factors.

Description

Technical Field

[0001] The present invention relates to the technical field of detection of facial information and body temperature information, and in particular to a method and system for detecting facial information and body temperature information using a causal model. Background Art

[0002] With the continuous progress of society and the continuous improvement of information technology and machine vision technology, face recognition technology and temperature detection technology have also made great progress.

[0003] As a biological feature of a person, face has the characteristics of high acceptance and reliability when used for feature recognition. However, mercury thermometers or other contact temperature testing tools have the risk of cross infection during use.

[0004] Therefore, high-precision non-contact temperature measurement has become a new trend. In some personnel screening situations that are sensitive to human body temperature, there is an urgent need for a method that combines facial recognition and non-contact temperature detection.

[0005] However, due to various influencing factors, its detection accuracy is not high, which is not conducive to its large-scale production and use, and reduces user experience. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a method and system for detecting facial information and body temperature information based on a causal model in response to the above-mentioned defects of the prior art, so as to solve the problem that due to various influencing factors, the detection accuracy is not high, which is not conducive to its large-scale production and use, and reduces the user experience.

[0007] The technical solution adopted by the present invention to solve the technical problem is: to provide a method for detecting face information and body temperature information of a causal model, comprising the steps of:

[0008] Establishing a face detection causal model, the face detection causal model includes face information input through a camera, a face database, and a front-door standard based on removing a first confounding factor, wherein the first confounding factor is an influencing factor that affects the camera's collection of facial feature integrity;

[0009] Establishing a causal model for human body temperature detection, the causal model for human body temperature detection including a human body temperature input through a temperature sensor, a normal temperature range of a human body, and a backdoor standard based on a known second confounding factor, wherein the second confounding factor is an ambient temperature that affects measurement of the temperature sensor and / or affects human body temperature;

[0010] The input face information is judged according to the face detection causal model, and the body temperature is collected when the result matches the face features;

[0011] The collected human body temperature is judged according to the human body temperature detection causal model to obtain the normal or abnormal condition of the human body temperature.

[0012] Among them, the preferred solution is: the formula of the front door standard is P(Y1|do(x1))=∑ M1 P(M1|x1)∑ x1’ P(Y1|x1',M1)P(x1'), where M1 is the face database, x1 is the face information input by the camera, and Y1 is the face judgment result; by removing the front-door standard of the first confounding factor, the cumulative comparison accuracy based on the face database M1 is obtained.

[0013] Among them, the preferred solution is: the formula of the backdoor standard is P(Y2|x2)=∑ M2 (M2|x2)∑ u2 P(Y2|x2,u2)P(u2)P(x2'), where u2 is the second confounding factor, M2 is the normal temperature range of the human body, x2 is the human body temperature input by the temperature sensor, and Y2 is the human body temperature judgment result; the human body temperature is obtained by combining the backdoor standard with the known second confounding factor.

[0014] Among them, the preferred solution is to further include the steps of:

[0015] The input facial information is judged according to the face detection causal model. When the result does not match the facial features, the facial information is acquired and input again through the camera.

[0016] Among them, a preferred solution is that the temperature sensor is a non-contact temperature sensor, and the direction of the non-contact temperature sensor is consistent with or close to the direction of the camera.

[0017] Among them, the preferred solution is to set up a display screen, and the facial information and body temperature information detection method also includes the steps of:

[0018] The collected human body temperature is judged according to the human body temperature detection causal model. When the human body temperature exceeds the normal temperature range of the human body, it is an abnormal situation, and the temperature information and the face image information are displayed in the first color on the display screen;

[0019] When the human body temperature is within the normal temperature range of the human body, it is a normal situation, and the temperature information and the face image information are displayed in the second color on the display screen.

[0020] The technical solution adopted by the present invention to solve its technical problems is: providing a facial information and body temperature information detection system, the facial information and body temperature information detection system comprising a facial information and body temperature information detection device and a background processing device, the facial information and body temperature information detection device comprising a host body, a camera and a temperature sensor arranged on the host body, a main control unit connected to the camera and the temperature sensor, and a data uploading unit, wherein the temperature sensor is a non-contact temperature sensor, and the direction of the non-contact temperature sensor is consistent with or close to the direction of the camera; the background processing device receives data information collected by the main control unit and transmitted through the data uploading unit, and implements the facial information and body temperature information detection method.

[0021] Among them, the preferred solution is: the facial information and body temperature information detection device also includes a display screen that displays temperature information and facial image information; the camera includes a color camera and a night vision camera.

[0022] The beneficial effect of the present invention is that, compared with the prior art, the present invention combines face recognition with temperature detection and combines the algorithm of the causal model to improve the accuracy of face recognition and temperature detection, thereby realizing rapid detection of one person, one face and one temperature with high accuracy and reducing the impact of external influencing factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0024] Figure 1 It is a flow chart of the method for detecting face information and body temperature information of the causal model of the present invention;

[0025] Figure 2 It is a structural schematic diagram of the face detection causal model of the present invention;

[0026] Figure 3 It is a structural schematic diagram of the causal model of human body temperature detection of the present invention;

[0027] Figure 4 It is a specific flow chart of the method for detecting face information and body temperature information of the present invention;

[0028] Figure 5 It is a structural schematic diagram of the face information and body temperature information detection system of the present invention;

[0029] Figure 6 It is a structural schematic diagram of the face information and body temperature information detection device of the present invention. DETAILED DESCRIPTION

[0030] Now, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0031] like Figures 1 to 3 As shown, the present invention provides a preferred embodiment of a method for detecting facial information and body temperature information based on a causal model.

[0032] A method for detecting facial information and body temperature information based on a causal model, comprising the steps of:

[0033] Step S10, establishing a face detection causal model and a body temperature detection causal model;

[0034] Step S20: actual detection processing step.

[0035] The step S20 comprises the steps of:

[0036] Step S21, judging the input face information according to the face detection causal model, and collecting the body temperature when the result meets the face features;

[0037] Step S22: judging the collected human body temperature according to the human body temperature detection causal model to obtain whether the human body temperature is normal or abnormal.

[0038] Among them, regarding establishing a face detection causal model, the face detection causal model includes face information input through a camera, a face database, and a front-door standard based on removing a first confounding factor, wherein the first confounding factor is an influencing factor that affects the camera's collection of facial feature integrity; regarding establishing a human body temperature detection causal model, the human body temperature detection causal model includes the human body temperature input through a temperature sensor, the normal temperature range of the human body, and a back-door standard based on a known second confounding factor, wherein the second confounding factor is the ambient temperature that affects the measurement of the temperature sensor and / or affects the human body temperature.

[0039] Combining facial recognition with temperature detection, the accuracy of facial recognition and temperature detection can be improved through the algorithm of the causal model, so as to achieve rapid detection of one person, one face and one temperature with high accuracy and reduce the impact of external factors.

[0040] Among them, step S10 is used as a pre-step, and step S20 is used as an actual detection process step.

[0041] In this embodiment, and with reference to Figure 2 , Figure 2 For the face detection causal model, the formula of the front door standard is P(Y1|do(x1))=∑ M1 P(M1|x1)∑ x1’P(Y1|x1',M1)P(x1'), where M1 is the face database, x1 is the face information input by the camera, and Y1 is the face judgment result; by removing the front-door standard of the first confounding factor, the cumulative comparison accuracy based on the face database M1 is obtained. The variable Z1 is the facial feature integrity of the face information x1 input by the camera, and its affected factors include light intensity, facial occlusions, and the quality of the camera itself. It is necessary to obtain the probability of the face judgment result Y1 based on the face information x1 input by the camera, that is, P(Y1|x1,M1). The judgment process needs to use the comparison with the face database M1. If the probability of x1 in M1 is 0, then Y1=0. If the similarity of x1 in M1 is Y1', P(Y1'|x1,M1). Due to the existence of the first confounding factor u1, and u1 cannot be observed, the front-door standard can be used to obtain the formula P(Y1|do(x1))=∑ M1 P(M1|x1)∑ x1’ P(Y1|x1',M1)p(x1'); The front door standard is cited because the first confounding factor u1 blocks the back door path from x1 to Y1, that is, there is no back door path from x1 to u1, and all back door paths from u1 to Y1 are not blocked by x1. Therefore, through the above formula, the first confounding factor u1 can be removed, and only the cumulative comparison accuracy of the face database M1 is concerned.

[0042] In this embodiment, and with reference to Figure 3 , Figure 3 is a causal model for human body temperature detection, and the formula of the backdoor standard is P(Y2|x2)=∑ M2 (M2|x2)∑ u2 P(Y2|x2,u2)P(u2)P(x2'), where u2 is the second confounding factor, M2 is the normal temperature range of the human body, x2 is the human body temperature input by the temperature sensor, and Y2 is the result of the human body temperature judgment; the human body temperature is obtained by combining the backdoor standard of the known second confounding factor. The variable Z2 is the environmental parameter of the human body temperature x2 test input by the temperature sensor, and its affected factors include black body reference accuracy and ambient light intensity. If x2 is not within the range of M2, then Y2=0, and its probability within the range is P(Y2|x2,M2). Due to the existence of the second confounding factor u2, and it can be known that the ambient temperature will affect the measured temperature and the normal temperature of the human body at the same time, there is no child node of x2 in u2, and blocking u2 can block all backdoors between x2 and Y2, which meets the backdoor standard. By referring to the front door standard, the formula P(Y2|x2)=∑ M2 (M2|x2)∑ u2 P(Y2|x2,u2)P(u2)P(x2').

[0043] like Figure 4As shown, the present invention provides a preferred embodiment of the method for detecting facial information and body temperature information.

[0044] The facial information and body temperature information detection method further comprises the steps of:

[0045] Step S23: judging the input face information according to the face detection causal model, and when the result does not match the face feature, re-acquiring and inputting the face information through the camera.

[0046] Furthermore, the temperature sensor is a non-contact temperature sensor, and the orientation of the non-contact temperature sensor is consistent with or close to the orientation of the camera; and a display screen is provided. It also includes step S24, judging the collected human body temperature according to the human body temperature detection causal model, when the human body temperature exceeds the normal human body temperature range, it is an abnormal situation, and the temperature information and the face image information are displayed in a first color on the display screen. It also includes step S25, when the human body temperature is within the normal human body temperature range, it is a normal situation, and the temperature information and the face image information are displayed in a second color on the display screen.

[0047] Specifically, regarding step S23, the accuracy of face information judgment is improved through the face detection causal model, so that when the structure is judged not to meet the face features, the next cycle is entered, that is, the face information is re-acquired until the face information meets the requirements or an alarm is issued. At the same time, the accuracy of face recognition is improved by combining related operations, such as answering questions, turning the head, opening the mouth or closing the eyes.

[0048] like Figure 5 and Figure 6 As shown, the present invention provides a preferred embodiment of a facial information and body temperature information detection system.

[0049] A facial information and body temperature information detection system, the facial information and body temperature information detection system comprising a facial information and body temperature information detection device 100 and a background processing device 200, the facial information and body temperature information detection device 100 comprising a host body 160, a camera 120 and a temperature sensor 130 arranged on the host body 160, a main control unit 110 connected to the camera 120 and the temperature sensor 130, and a data upload unit 140, wherein the temperature sensor 130 is a non-contact temperature sensor 130, and the orientation of the non-contact temperature sensor 130 is consistent with or close to the orientation of the camera 120; the background processing device 200 receives data information collected by the main control unit 110 and transmitted through the data upload unit 140, and implements the facial information and body temperature information detection method.

[0050] Specifically, the camera 120, the temperature sensor 130 or the induction sensor is used to obtain whether there is a person at present. If there is, facial information and body temperature information are detected, that is, face recognition and body temperature acquisition. The camera 120 and the temperature sensor 130 are controlled by the main control unit 110 to implement corresponding operations, such as controlling the camera 120 and the temperature sensor 130 to collect data at the same time, obtain data and transmit it to the background processing device 200 through the data upload unit 140 for data processing. Combined with the face detection causal model and the human body temperature detection causal model set in the background processing device 200, the face information and body temperature information detection method is implemented to obtain face information and corresponding temperature information.

[0051] In this embodiment, the facial information and body temperature information detection device 100 further includes a display screen 150 for displaying temperature information and facial image information; the camera 120 includes a color camera 121 and a night vision camera 122 .

[0052] Temperature information and facial image information are displayed on the display screen 150, and a color camera or a night vision camera is used according to the ambient light to improve accuracy.

[0053] The above is only the best embodiment of the present invention and is not intended to limit the scope of the present invention. All equivalent changes or modifications made according to the scope of the patent application of the present invention are covered by the present invention.

Claims

1. A method for detecting facial information and body temperature information based on a causal model, characterized in that: Includes steps: Establishing a face detection causal model, the face detection causal model includes face information input through a camera, a face database, and a front-door standard based on removing a first confounding factor, wherein the first confounding factor is an influencing factor that affects the camera's collection of facial feature integrity, and the influencing factor includes light intensity, facial obstructions, and the quality of the camera itself; Establishing a causal model for human body temperature detection, the causal model for human body temperature detection including a human body temperature input through a temperature sensor, a normal temperature range of a human body, and a backdoor standard based on a known second confounding factor, wherein the second confounding factor is an ambient temperature that affects measurement of the temperature sensor and / or affects human body temperature; The input face information is judged according to the face detection causal model, and the body temperature is collected when the result matches the face features; Judging the collected human body temperature according to the human body temperature detection causal model to obtain the normal or abnormal condition of the human body temperature; The formula of the front door standard is P(Y1|do(x1))=∑ M1 P(M1|x1)∑ x1’ P(Y1|x1',M1)P(x1'), where M1 is the face database, x1 is the face information input by the camera, and Y1 is the face judgment result; by removing the front-door standard of the first confounding factor, the cumulative comparison accuracy based on the face database M1 is obtained; wherein, the first confounding factor u1 cannot be observed, and the first confounding factor u1 blocks the back-door path from x1 to Y1. By using the formula of the above front-door standard, the first confounding factor u1 is removed, and only the cumulative comparison accuracy of the face database M1 is concerned; The formula of the backdoor standard is P(Y2|x2)=∑ M2 (M2|x2)∑ u2 P(Y2|x2,u2)P(u2)P(x2'), where u2 is the second confounding factor, M2 is the normal temperature range of the human body, x2 is the human body temperature input by the temperature sensor, and Y2 is the human body temperature judgment result; the human body temperature is obtained by combining the backdoor standard with the known second confounding factor.

2. The method for detecting facial information and body temperature information according to claim 1, characterized in that: Also includes the steps: The input facial information is judged according to the face detection causal model. When the result does not match the facial features, the facial information is acquired and input again through the camera.

3. The method for detecting facial information and body temperature information according to claim 1, characterized in that: The temperature sensor is a non-contact temperature sensor, and the direction of the non-contact temperature sensor is consistent with or close to the direction of the camera.

4. The method for detecting facial information and body temperature information according to claim 1, characterized in that: A display screen is provided, and the facial information and body temperature information detection method further comprises the steps of: The collected human body temperature is judged according to the human body temperature detection causal model. When the human body temperature exceeds the normal temperature range of the human body, it is an abnormal situation, and the temperature information and the face image information are displayed in the first color on the display screen; When the human body temperature is within the normal temperature range of the human body, it is a normal situation, and the temperature information and the face image information are displayed in the second color on the display screen.

5. A facial information and body temperature information detection system, characterized in that: The facial information and body temperature information detection system includes a facial information and body temperature information detection device and a background processing device, the facial information and body temperature information detection device includes a host body, a camera and a temperature sensor arranged on the host body, a main control unit connected to the camera and the temperature sensor, and a data upload unit, wherein the temperature sensor is a non-contact temperature sensor, and the direction of the non-contact temperature sensor is consistent with or close to the direction of the camera; the background processing device receives data information collected by the main control unit and transmitted through the data upload unit, and implements the facial information and body temperature information detection method as described in claim 1.

6. The facial information and body temperature information detection system according to claim 5, characterized in that: The facial information and body temperature information detection device also includes a display screen that displays temperature information and facial image information; the camera includes a color camera and a night vision camera.

Images