Drug taking detection method and device, storage medium and equipment

Abstract

The invention discloses a drug taking detection method and device, a storage medium and equipment, and belongs to the field of biological feature recognition. The method comprises the following steps: firstly, acquiring at least one first image sequence containing a face image and at least one second image sequence containing an iris image; then acquiring micro-expression features according to the first image sequence, and acquiring iris texture features according to the second image sequence; and finally, performing drug taking judgment on the detected object according to the micro-expression features and the iris texture features. According to the invention, through fusion analysis of the face micro-expression features and the iris texture features of the detected object, whether the detected object takes drugs or not is comprehensively judged, and the accuracy of drug taking detection is improved. Moreover, non-contact drug taking detection is adopted, cooperation of a detected object is not needed, drug taking detection can be carried out under the condition that the detected object is non-inductive, and the method can adapt to different anti-drug detection environments.

Description

Technical Field

[0001] The present application relates to the field of biometric identification, and in particular to a drug detection method, device, storage medium and equipment. Background Art

[0002] Traditional drug testing usually involves collecting urine, blood, hair, etc. from the subject for biochemical testing. These biochemical drug testing methods usually require the development of special test strips for special confirmation, and have problems such as cumbersome sampling and testing and long testing cycles.

[0003] With the advancement of technology, there is currently an eye mask type drug detection device (one example of which can be found in Chinese patent document CN218004137U). This detection device requires the subject to be detected to fit his nose to the eye mask during detection, so as to judge whether the subject is taking drugs through the response characteristics of the pupil to light stimulation. This eye mask type detection device uses pupil characteristics to detect drug use and is a contact detection device. Although the detection accuracy is high, it requires the subject to be detected to cooperate highly in aligning the eye mask and opening his eyes. Summary of the invention

[0004] In order to address the defects of the prior art, the present application provides a drug detection method, device, storage medium and equipment. The present application does not require the cooperation of the subject being detected, and can perform drug detection without the subject being detected feeling anything, while improving the accuracy of drug detection.

[0005] This application provides the following technical solutions:

[0006] In a first aspect, the present application provides a method for drug abuse detection, the method comprising:

[0007] The face of the detected object is photographed continuously and non-contactly by using a visible light camera and an infrared camera respectively, so as to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image respectively;

[0008] Acquire at least one micro-expression feature of the detected object according to the first image sequence;

[0009] Acquire iris texture features of the detected object according to the second image sequence;

[0010] The detected object is judged to be taking drugs based on the at least one micro-expression feature and the iris texture feature.

[0011] Furthermore, when it is determined that the detected subject is taking drugs, the method further includes:

[0012] Performing flash stimulation on the eyes of the detected object, and performing non-contact continuous photography of the eye region of the detected object by the infrared camera during the entire process from the start of the flash stimulation to the change and recovery of the pupil of the detected object, to obtain a plurality of third image sequences including iris images;

[0013] Acquire iris change characteristics of the detected object according to the third image sequence;

[0014] The drug use of the detected object is determined based on the iris change characteristics.

[0015] Furthermore, acquiring the iris change feature of the detected object according to the third image sequence includes:

[0016] Performing pupil boundary positioning on the third image sequence to obtain a pupil boundary sequence;

[0017] The iris change feature is determined according to the change of the pupil boundary sequence.

[0018] Further, judging drug use of the detected object based on the at least one micro-expression feature and the iris texture feature includes:

[0019] Inputting the at least one micro-expression feature and the iris texture feature into a pre-trained drug detection model, and judging whether the detected subject is taking drugs according to a result output by the drug detection model;

[0020] and / or;

[0021] The step of judging drug use of the detected object according to the iris change characteristics comprises:

[0022] Determine whether the detected object is taking drugs based on changes in the pupil boundary sequence, and / or input the iris change features determined by changes in the pupil boundary sequence into a pre-trained drug-related detection model, and determine whether the detected object is taking drugs based on a result output by the drug-related detection model.

[0023] Furthermore, the method further comprises:

[0024] The health status of the detected object is judged according to the results output by the drug-related detection model.

[0025] Furthermore, the flash stimulation of the eyes of the detected object may also include:

[0026] The two front iris images of the detected object are collected successively by an infrared camera, and it is determined whether the pupil boundaries of the two front iris images collected successively remain stable. If so, the step of flashing the eyes of the detected object is performed, otherwise, this step is repeated.

[0027] Furthermore, the types of micro-expression features include one or more of yawning, tears, dull expression, blurred eyes, dullness, fear, dark skin and sunken eyes.

[0028] Furthermore, the method further comprises:

[0029] Performing facial feature extraction and / or iris feature extraction according to the first image sequence and / or the second image sequence respectively;

[0030] Performing identity recognition on the detected object according to the facial features and / or iris features obtained through the facial feature extraction and / or the iris feature extraction;

[0031] The drug taking history of the detected subject is determined according to the result of the identity recognition of the detected subject.

[0032] In a second aspect, the present application provides a drug detection device, the device comprising:

[0033] A first data acquisition module is used to simultaneously and continuously photograph the face of the detected object using a visible light camera and an infrared camera, respectively, to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image;

[0034] A micro-expression feature extraction module, used for acquiring at least one micro-expression feature of the detected object according to the first image sequence;

[0035] an iris texture feature extraction module, configured to obtain the iris texture feature of the detected object according to the second image sequence;

[0036] The first drug taking judgment module is used to judge drug taking of the detected object according to the at least one micro-expression feature and the iris texture feature.

[0037] Furthermore, when it is determined that the detected object is taking drugs, the device further comprises:

[0038] a second data acquisition module, configured to perform flash stimulation on the eyes of the detected object, and to continuously and non-contactly photograph the eye region of the detected object through the infrared camera during the entire process from the start of the flash stimulation to the change and recovery of the pupil of the detected object, so as to obtain a plurality of third image sequences including iris images;

[0039] an iris change feature extraction module, configured to obtain the iris change feature of the detected object according to the third image sequence;

[0040] The second drug taking judgment module is used to judge drug taking of the detected object according to the iris change characteristics.

[0041] Furthermore, the iris change feature extraction module includes:

[0042] A pupil boundary sequence extraction unit, used for locating the pupil boundary of the third image sequence to obtain a pupil boundary sequence;

[0043] The iris change feature extraction unit is used to determine the iris change feature according to the change of the pupil boundary sequence.

[0044] Furthermore, the first drug-taking determination module is used to: input the at least one micro-expression feature and the iris texture feature into a pre-trained drug-related detection model, and determine whether the detected object is taking drugs based on a result output by the drug-related detection model.

[0045] and / or;

[0046] The second drug taking judgment module is used to: judge whether the detected object is taking drugs according to the changes in the pupil boundary sequence, and / or input the iris change features determined by the changes in the pupil boundary sequence into a pre-trained drug-related detection model, and judge whether the detected object is taking drugs according to the results output by the drug-related detection model.

[0047] Furthermore, the device also includes:

[0048] The health judgment module is used to make a health judgment on the detected object according to the result output by the drug-related detection model.

[0049] Furthermore, the device also includes:

[0050] The stability judgment module is used to successively collect two front iris images of the detected object through an infrared camera, and judge whether the pupil boundaries of the two front iris images collected successively remain stable. If so, execute the second data acquisition module; otherwise, repeat the execution of the stability judgment module.

[0051] Furthermore, the types of micro-expression features include one or more of yawning, tears, dull expression, blurred eyes, dullness, fear, dark skin and sunken eyes.

[0052] Furthermore, the device also includes:

[0053] A face and iris feature extraction module, used to extract face features and / or iris features according to the first image sequence and / or the second image sequence respectively;

[0054] An identity recognition module, used for performing identity recognition on the detected object based on the facial features and / or iris features obtained through the facial feature extraction and / or the iris feature extraction;

[0055] The drug taking history determination module is used to determine the drug taking history of the detected object according to the result of the identity recognition of the detected object.

[0056] In a third aspect, the present application provides a computer-readable storage medium for drug detection, comprising a memory for storing processor-executable instructions, wherein the instructions, when executed by the processor, implement the steps of the drug detection method described in the first aspect.

[0057] In a fourth aspect, the present application provides a device for drug detection, characterized in that it includes at least one processor and a memory storing computer-executable instructions, and when the processor executes the instructions, the steps of the drug detection method described in the first aspect are implemented.

[0058] This application has the following beneficial effects:

[0059] This application comprehensively determines whether the detected object is taking drugs by fusing and analyzing the facial micro-expression features and iris texture features of the detected object, thereby improving the accuracy of drug detection compared to drug detection based on a single feature. In addition, this application is a contactless drug detection, which does not require the cooperation of the detected object, can detect drug use without the detected object being aware of it, and can adapt to different drug detection environments. In addition, the extraction speed of micro-expression features and iris texture features is fast, and can be performed simultaneously with face recognition and iris recognition. When face recognition and iris recognition are performed to determine the identity of the detected object, drug detection can be performed simultaneously and quickly without being perceived by the detected object. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flow chart of the drug detection method of the present application;

[0061] Figure 2 This is a schematic diagram of the drug detection device of the present application. DETAILED DESCRIPTION

[0062] In order to make the technical problems, technical solutions and advantages to be solved by the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with the drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0063] The present application embodiment provides a drug abuse detection method, such as Figure 1 As shown, the method includes:

[0064] S100: Non-contact continuous photographing of the face of the detected object is performed simultaneously by a visible light camera and an infrared camera, respectively, to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image.

[0065] In the present application, the visible light camera may be a color camera, and the visible light camera and the infrared camera may be two independent cameras, or the visible light camera and the infrared camera may be an integrated color-infrared dual-channel camera. The visible light camera is used to capture the entire face of the detected object to obtain a face image; the infrared camera is used to capture the eye area within the face of the detected object to obtain an iris image.

[0066] In one example, the visible light camera has a pixel of more than 2 million and a frame rate of more than 30 frames. The infrared camera can adaptively track the height of a human body in pitch. Its pixel is more than 2 million and the infrared light band is 790nm-880nm. Before taking non-contact continuous photos of the face of the detected object, the focal length and position of the infrared camera lens can be adaptively adjusted according to the height or position of the detected object, so that the captured iris image is of appropriate size and centered. In one example, when adaptively adjusting the lens of the infrared camera, the iris diameter in the iris image in the second image sequence is controlled to be larger than 160 pixels to prevent unclear texture features caused by a small iris.

[0067] Visible light cameras and infrared cameras are non-contact cameras and do not come into contact with the detected object. In some application scenarios, visible light cameras and infrared cameras and other corresponding hardware devices can be deployed at the hotel front desk, security gate entrance center, etc., and from the appearance, it is an identity verification device. When people pass by the above scenes for identity recognition, drug detection is carried out simultaneously without the detected object being aware of it.

[0068] The first image sequence captured may include only one face image or multiple face images, and the multiple face images exist in the form of video frame sequences; similarly, the second image sequence captured may include only one iris image or multiple iris images.

[0069] S200: Acquire at least one micro-expression feature of the detected object according to the first image sequence.

[0070] In the present application, the micro-expression features are expressions that can reflect whether the detected object is taking drugs, and the types include but are not limited to yawning, tears, dull expression, blurred eyes, dullness, fear, dark skin, sunken eyes, etc. When obtaining micro-expression features, it can be implemented based on expression recognition algorithms, such as expression recognition based on support vector machines (SVM), neural networks, decision trees and other algorithms, which will not be repeated in this application.

[0071] S300: Acquire iris texture features of the detected object according to the second image sequence.

[0072] Iris texture features are the details of interlaced spots, filaments, coronal shapes, stripes, crypts, etc. contained in the iris texture area between the inner boundary of the iris (i.e., the pupil boundary) and the outer boundary of the iris. Iris texture features can be obtained by extracting features from the iris texture area. On the one hand, iris texture features can be used for iris recognition to determine the identity of the detected object. On the other hand, there are differences in the iris texture features of drug addicts and normal people. The iris texture features of drug addicts will have corresponding lesions due to drug use. Therefore, it is possible to judge whether the detected object is taking drugs based on the iris texture features.

[0073] The present application does not limit the method for extracting iris texture features. In an example, iris texture features can be obtained through steps such as iris detection, iris effective area segmentation and feature extraction.

[0074] S400: Determine drug use of the detected object based on the at least one micro-expression feature and the iris texture feature.

[0075] The present application does not limit the specific implementation method of judging drug use based on micro-expression features and iris texture features. In one example, the at least one micro-expression feature and the iris texture feature can be input into a pre-trained drug detection model, and the result output by the drug detection model is used to judge whether the detected object is taking drugs. Optionally, when judging drug use, information such as drug use time and drug type can be output at the same time.

[0076] The drug detection model needs to be trained in advance. During the training, a drug detection database needs to be built. The drug detection database includes a series of first image sequences and second image sequences corresponding to drug users and non-drug users. The drug detection model learns based on the drug detection database, updates its parameters, and obtains a drug detection model that can distinguish whether drug users are taking drugs.

[0077] This application comprehensively determines whether the detected object is taking drugs by fusing and analyzing the facial micro-expression features and iris texture features of the detected object, thereby improving the accuracy of drug detection compared to drug detection based on a single feature. In addition, this application is a contactless drug detection, which does not require the cooperation of the detected object, can detect drug use without the detected object being aware of it, and can adapt to different drug detection environments. In addition, the extraction speed of micro-expression features and iris texture features is fast, and can be performed simultaneously with face recognition and iris recognition. When face recognition and iris recognition are performed to determine the identity of the detected object, drug detection can be performed simultaneously and quickly without being perceived by the detected object.

[0078] When the detected object is judged to be taking drugs by micro-expression features and iris texture features, a further judgment can be made on whether the detected object is taking drugs. In this case, the method of the present application further includes:

[0079] S500: performing flash stimulation on the eyes of the detected object, and performing non-contact continuous photography of the eye area of ​​the detected object by an infrared camera during the entire process from the start of the flash stimulation to the change and recovery of the pupil of the detected object, to obtain a third image sequence including multiple iris images.

[0080] The source of the flash stimulation for the eyes of the detected object may be a device such as a visible light flash, such as a white light flash. The duration of the flash stimulation may be in the range of 50-200ms. After the flash stimulation, the pupil of the collected object first becomes smaller and then gradually recovers. During this whole process, the eye area in the face of the detected object is continuously photographed in a non-contact manner by the infrared camera to obtain a third image sequence including multiple near-infrared iris images.

[0081] When shooting continuously, the time interval between each two images can be the same or different, and can be set as needed, and this application does not impose any restrictions on this. The duration of shooting can be, for example, 3-5 seconds, which is based on the entire process of pupil change and recovery of the detected object.

[0082] S600: Acquire iris change characteristics of the detected object according to the third image sequence.

[0083] After the flash stimulation, the pupil of the collected object first becomes smaller and then gradually recovers, that is, the inner boundary of the iris first becomes smaller and then gradually recovers. The iris change characteristics can be obtained by the change of the inner boundary of the iris (i.e., the pupil boundary) over time in the third image sequence. Since the pupils of drug addicts and normal people react differently to flash stimulation, their pupil change characteristics are different. Therefore, it is possible to determine whether the detected object is taking drugs through the iris change characteristics.

[0084] In one example, an implementation of this step includes:

[0085] S610: Perform pupil boundary positioning on the third image sequence to obtain a pupil boundary sequence.

[0086] The pupil boundary is the inner boundary of the iris, and various iris detection and boundary positioning methods in the prior art can be used to locate the pupil boundary, which will not be described in detail in this application. When locating the pupil boundary, the pupil boundary can be located as a circle.

[0087] S620: Determine iris change characteristics according to changes in the pupil boundary sequence.

[0088] After the flash stimulation, the pupil of the subject first becomes smaller and then gradually recovers. The iris change characteristics can be determined by the changes in the pupil boundary sequence. When the pupil boundary is located as a circle, the iris change characteristics can be determined based on the changes in the pupil diameter.

[0089] In other examples, the ratio of the pupil diameter to the iris outer boundary diameter belonging to the same iris image may be calculated, and the iris change feature may be determined based on the change in a sequence consisting of the ratios corresponding to all iris images of the third image sequence.

[0090] S700: Determine drug use of the detected subject based on the iris change characteristics.

[0091] In this step, it is possible to directly determine whether the detected object is taking drugs based on the change of the pupil boundary sequence obtained in the aforementioned S620. For example, it is possible to determine whether the detected object is taking drugs based on features such as the change rate of the pupil boundary sequence, or it is possible to draw a curve of the change of the pupil boundary sequence relative to time, and determine whether the detected object is taking drugs based on features such as the slope of the change curve.

[0092] Moreover, similar to the micro-expression features and iris texture features, the iris change features obtained in the aforementioned S620 may also be input into a pre-trained drug detection model to determine whether the detected subject is taking drugs based on the results output by the drug detection model.

[0093] The drug-related detection model used by the iris change feature and the drug-related detection model used by the micro-expression feature and the iris texture feature can be the same drug-related detection model, or they can be drug-related detection models used separately. Regardless of whether they are the same drug-related detection model or drug-related detection models used separately, the drug-related detection model used by the iris change feature needs to be learned in advance using the training data of the iris change feature.

[0094] As an improvement, in the aforementioned S400 or S700, when the drug-related detection model is used to determine drug use, the method of the present application further includes:

[0095] S800: Perform health assessment on the subject according to the output of the drug-related detection model.

[0096] The state of the face and iris is related to the health of the human body. Some common diseases can reflect some targeted features on the face and iris. Therefore, in this application, the results output by the drug detection model can also include the health status of the detected object, so as to make a health judgment on the detected object and determine whether there is a possibility of common diseases.

[0097] In this application, when judging the health of the detected object, drug abuse can be regarded as a special disease, and drug abuse and other diseases can be judged at the same time. Of course, drug abuse can also be not regarded as a special disease, but only the health condition other than drug abuse can be judged.

[0098] Please note that the health status output by the drug-related detection model of this application is only a possible prediction and cannot be used as a disease diagnosis result. It can only be used as an auxiliary reference for the health status of the subject being tested. If you want to obtain an accurate disease diagnosis result, you need to go to a medical institution for diagnosis.

[0099] The present application needs to shoot the third image sequence after the iris and pupil of the detected object are stable. Based on this, the present application also includes before S500:

[0100] The infrared camera is used to successively capture two front iris images of the detected object, and it is determined whether the pupil boundaries of the two successively captured front iris images remain stable. If so, it is proved that the iris and pupil of the detected object are stable, and then S500 is executed. Otherwise, this step is repeated until the pupil boundary remains stable.

[0101] Likewise, when capturing the first image sequence and the second image sequence, a similar method for determining whether the iris and pupil are stable may also be used.

[0102] In the present application, when drug use is detected by micro-expression features and iris texture features, the identity of the detected object can also be identified at the same time. Based on this, the method of the present application also includes:

[0103] S10: Perform facial feature extraction and / or iris feature extraction according to the first image sequence and / or the second image sequence respectively.

[0104] The present application does not limit the method of facial feature extraction and iris feature extraction, and the facial feature extraction algorithm and iris feature extraction algorithm already available in the prior art can be used for implementation, and the present application will not elaborate on this.

[0105] S20: performing identity recognition on the detected object according to the facial features and / or iris features obtained through facial feature extraction and / or iris feature extraction.

[0106] When performing identity recognition, recognition can be performed only through facial features (i.e., only face recognition is performed), recognition can be performed only through iris features (i.e., only iris recognition is performed), or recognition can be performed through both facial features and iris features (i.e., face and iris multimodal recognition).

[0107] S30: Determine the drug use history of the detected subject according to the result of the identification of the detected subject.

[0108] In one example, for a pre-determined drug addict, his / her identity information and drug history are pre-entered into a database, and the drug history may include information such as the type of drug used and the time of drug use. After identity recognition based on facial features and / or iris features, his / her identity information and drug history can be obtained from the stored database. The identity information may be detailed information of the detected object, including name, age, gender, race, ID number, etc. The drug history information can be used to provide guidance for screening or further judgment.

[0109] In another example, the results output by the drug detection model may also include the identity information, drug use time, drug type and drug use history of the detected object. In this case, the identity information of the detected object is the general identity information predicted based on micro-expression features, iris change features and iris texture features, including age, gender, race, etc., and the drug use history is also predicted based on micro-expression features, iris change features and iris texture features.

[0110] To ensure the quality of the first image sequence, the second image sequence, and the third image sequence and ensure that they are clear and usable, the method of the present application further includes:

[0111] Perform image quality detection on the first image sequence / the second image sequence / the third image sequence, and if the image quality detection result does not meet a predetermined condition, reshoot.

[0112] The drug detection model used in this application can be a small sample learning model, which has the advantages of strong interpretability and fast running speed. Of course, the drug detection model used in this application can also be a conventional learning model (i.e., a large sample learning model, such as a recurrent neural network model), and this application does not limit this.

[0113] When training the drug detection model, it is necessary to build a drug detection database. The process of building a drug detection database is described in detail below.

[0114] The data content of the drug detection database includes: a group of dynamic video data, including a first image sequence, a second image sequence and a third image sequence, and the length of the video data varies from 3 to 5 seconds.

[0115] The data content is marked according to the type of drugs used and the time of use.

[0116] The facial image specifications of the first image sequence are as follows: the distance between the two eyes of the facial image is greater than 160 pix, and the color and infrared dual-channel data are sampled simultaneously.

[0117] The iris image specifications of the second image sequence and the third image sequence are: infrared data in the 790-880nm band range with an iris diameter greater than 160 pix.

[0118] At the same time, the data content is marked with basic information of the people, including: gender, age, drug use history, type of use, use time, etc.

[0119] When constructing the data content of the drug detection database, sampling is first carried out through the non-contact iris drug detection analysis method. Specifically: in the targeted drug detection scene, the data of confirmed drug users can be sampled with the collection equipment. The facial micro-expression data and iris dynamic data are sampled at the same time, classified and stored in the drug detection database, and key information is marked (such as age, gender, time of use, type of use, method of use, drug use history, etc.). The types of drugs used include heroin, methamphetamine, K powder, ecstasy, etomidate, methcathinone and other common drugs, and the methods of use include smoking, injection, and sticking to the skin. At the same time, it can also mark whether there are chronic diseases (diabetes, heart disease, etc.) and whether they take medication for a long time and other health information.

[0120] For the data of negative people who do not take drugs, it is necessary to mark their age, gender, whether they have chronic diseases (diabetes, heart disease, etc.), and whether they take long-term medication.

[0121] Then, the sampled data is classified and extracted through the micro-expression analysis method. Specifically: facial micro-expression information is analyzed, classified and analyzed according to different types of drugs, and the characteristics of the same type of expression phenomena are extracted for the design of drug-related detection models. Expression types include but are not limited to blurred eyes, dull expressions, panic, dark skin, sunken eyes, tears, yawns, etc.

[0122] Optionally, the classified data can be further corrected by combining the empirical analysis method. Specifically: the data sorted out by the expression analysis method is analyzed and corrected in combination with the experience of front-line drug inspectors, so that the theory is combined with the practice test, and the questionable data is marked for re-analysis.

[0123] It is also possible to collect data from suspected people who have been treated for drug addiction in drug rehabilitation centers and make secondary corrections to the data. Specifically: Keep a good data file for drug addicts, create files through face and iris recognition, and ensure that the personnel are consistent before and after drug rehabilitation. Then, when they leave the drug rehabilitation center, they will be sampled for drug testing information again to eliminate the influence of other personal factors, make secondary corrections to the data, and improve the robustness of the algorithm.

[0124] To facilitate understanding of the technical solution of the present application, the present application is described in detail below with reference to specific practical application scenarios.

[0125] 1. This application corresponds to the deployment of hardware equipment (color camera and infrared camera and other supporting hardware) at the hotel front desk, gate entrance center, etc. From the appearance, it is an identity verification device.

[0126] 2. When people pass by the above scene, the first image sequence and the second image sequence are collected, and while performing identity recognition, micro-expression analysis and iris texture feature extraction are performed simultaneously, and drug use detection is performed based on the micro-expression and iris texture features. This process can basically be completed simultaneously with identity recognition.

[0127] 3. If the results of micro-expression and iris texture feature detection show that the person has not taken drugs, he or she can pass through normally and smoothly, and the time is controlled within 1 second.

[0128] 4. If the detection results of micro-expression and iris texture features indicate drug use, the flashlight will be turned on, and the third image sequence will be collected to detect drug use based on the iris change characteristics. The whole process will add 2-3 seconds to complete the identity authentication and drug test information screening.

[0129] 5. If the screening result is a suspected drug user, a background warning can be implemented, and the information of the tested person can be synchronized to the local law enforcement unit for further manual or urine test screening.

[0130] The present application also provides a drug detection device, such as Figure 2 As shown, the device comprises:

[0131] The first data acquisition module 100 is used to simultaneously and continuously photograph the face of the detected object using a visible light camera and an infrared camera to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image.

[0132] The micro-expression feature extraction module 200 is used to obtain at least one micro-expression feature of the detected object according to the first image sequence.

[0133] The iris texture feature extraction module 300 is used to obtain the iris texture features of the detected object according to the second image sequence.

[0134] The first drug taking judgment module 400 is used to judge drug taking of the detected object according to at least one micro-expression feature and an iris texture feature.

[0135] This application comprehensively determines whether the detected object is taking drugs by fusing and analyzing the facial micro-expression features and iris texture features of the detected object, thereby improving the accuracy of drug detection compared to drug detection based on a single feature. In addition, this application is a contactless drug detection, which does not require the cooperation of the detected object, can detect drug use without the detected object being aware of it, and can adapt to different drug detection environments. In addition, the extraction speed of micro-expression features and iris texture features is fast, and can be performed simultaneously with face recognition and iris recognition. When face recognition and iris recognition are performed to determine the identity of the detected object, drug detection can be performed simultaneously and quickly without being perceived by the detected object.

[0136] As an improvement of the embodiment of the present application, when it is determined that the detected object is taking drugs, the device further includes:

[0137] The second data acquisition module is used to perform flash stimulation on the eyes of the detected object. From the start of the flash stimulation to the change and recovery of the pupil of the detected object, the eye area of ​​the detected object is continuously photographed in a non-contact manner by an infrared camera to obtain a third image sequence containing multiple iris images.

[0138] The iris change feature extraction module is used to obtain the iris change features of the detected object according to the third image sequence.

[0139] The second drug abuse judgment module is used to judge drug abuse of the detected object according to the iris change characteristics.

[0140] In one example, the aforementioned iris change feature extraction module includes:

[0141] The pupil boundary sequence extraction unit is used to locate the pupil boundary of the third image sequence to obtain a pupil boundary sequence.

[0142] The iris change feature extraction unit is used to determine the iris change feature according to the change of the pupil boundary sequence.

[0143] Specifically, the aforementioned first drug abuse judgment module is used to: input at least one micro-expression feature and an iris texture feature into a pre-trained drug-related detection model, and judge whether the detected object is taking drugs according to the result output by the drug-related detection model.

[0144] Correspondingly, the aforementioned second drug abuse judgment module is used to: judge whether the detected object is taking drugs based on the changes in the pupil boundary sequence, and / or, input the iris change features determined by the changes in the pupil boundary sequence into a pre-trained drug-related detection model, and judge whether the detected object is taking drugs based on the results output by the drug-related detection model.

[0145] As another improvement of the embodiment of the present application, the device may further include:

[0146] The health judgment module is used to make health judgments on the detected object based on the results output by the drug-related detection model.

[0147] The stability judgment module is used to successively collect two front iris images of the detected object through an infrared camera, and judge whether the pupil boundaries of the two front iris images collected successively remain stable. If so, execute the second data acquisition module; otherwise, repeat the execution of the stability judgment module.

[0148] In order to realize the identity recognition of the detected object, the device of the present application may also include:

[0149] The face and iris feature extraction module is used to extract face features and / or iris features according to the first image sequence and / or the second image sequence respectively.

[0150] The identity recognition module is used to perform identity recognition on the detected object based on the facial features and / or iris features obtained through facial feature extraction and / or iris feature extraction.

[0151] The drug taking history determination module is used to determine the drug taking history of the detected object according to the result of the identity recognition of the detected object.

[0152] In some examples, the types of micro-expression features may include one or more of yawning, tears, dull expression, blurred eyes, dullness, fear, dark skin, and sunken eyes.

[0153] The device provided in the above embodiment, its implementation principle and the technical effect produced correspond to the embodiment of the above method one by one. For the sake of brief description, the part not mentioned in the embodiment of the device can refer to the corresponding content in the embodiment of the above method. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the modules and units described in the device can refer to the corresponding process in the embodiment of the above method, and will not be repeated here.

[0154] The drug detection method described in the above embodiments provided by the present application can implement the business logic through a computer program and record it on a storage medium, and the storage medium can be read and executed by a computer to achieve the effect of the scheme described in the method embodiment of this specification. Therefore, the embodiment of the present application also provides a computer-readable storage medium for drug detection, including a memory for storing processor-executable instructions, and when the instructions are executed by the processor, the steps of the drug detection method including the above embodiments are implemented.

[0155] The storage medium may include a physical device for storing information, which is usually a medium that digitizes the information and then stores it in an electrical, magnetic or optical manner. The storage medium may include: a device that stores information in an electrical energy manner, such as various memories, such as RAM, ROM, etc.; a device that stores information in a magnetic energy manner, such as a hard disk, a floppy disk, a magnetic tape, a magnetic core memory, a magnetic bubble memory, a USB flash drive; a device that stores information in an optical manner, such as a CD or a DVD. Of course, there are other readable storage media, such as quantum memory, graphene memory, etc.

[0156] The storage medium described above may also include other implementation methods according to the description of the method embodiment. The implementation principle and technical effects produced by this embodiment are the same as those of the aforementioned method embodiment. For details, please refer to the description of the relevant method embodiment, and no further description will be given here.

[0157] The embodiment of the present application also provides a device for drug detection, which may be a separate computer, or may include an actual operating device using one or more of the methods or one or more of the embodiments of the present specification. The drug detection device may include at least one processor and a memory storing computer executable instructions, and when the processor executes the instructions, any one or more of the steps of the drug detection method described above is implemented.

[0158] The device described above may also include other implementation methods according to the description of the method embodiment. The implementation principle and technical effects produced by this embodiment are the same as those of the aforementioned method embodiment. For details, please refer to the description of the relevant method embodiment, and no further description will be given here.

[0159] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. They should all be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A drug abuse detection method, characterized in that: The method comprises: The face of the detected object is continuously photographed by a visible light camera and an infrared camera simultaneously to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image; Acquire at least one micro-expression feature of the detected object according to the first image sequence; Acquire iris texture features of the detected object according to the second image sequence; The detected object is judged to be taking drugs based on the at least one micro-expression feature and the iris texture feature.

2. The drug abuse detection method according to claim 1, characterized in that: When it is determined that the detected subject is taking drugs, the method further comprises: Performing flash stimulation on the eyes of the detected object, and performing non-contact continuous photography of the eye region of the detected object by the infrared camera during the entire process from the start of the flash stimulation to the change and recovery of the pupil of the detected object, to obtain a plurality of third image sequences including iris images; Acquire iris change characteristics of the detected object according to the third image sequence; The drug use of the detected object is determined based on the iris change characteristics.

3. The drug abuse detection method according to claim 2, characterized in that: The step of acquiring the iris change feature of the detected object according to the third image sequence includes: Performing pupil boundary positioning on the third image sequence to obtain a pupil boundary sequence; The iris change feature is determined according to the change of the pupil boundary sequence.

4. The drug abuse detection method according to claim 3, characterized in that: The step of judging drug use of the detected object based on the at least one micro-expression feature and the iris texture feature comprises: Inputting the at least one micro-expression feature and the iris texture feature into a pre-trained drug detection model, and judging whether the detected subject is taking drugs according to a result output by the drug detection model; and / or; The step of judging drug use of the detected object according to the iris change characteristics comprises: Determine whether the detected object is taking drugs based on changes in the pupil boundary sequence, and / or input the iris change features determined by changes in the pupil boundary sequence into a pre-trained drug-related detection model, and determine whether the detected object is taking drugs based on a result output by the drug-related detection model.

5. The drug abuse detection method according to claim 4, characterized in that: The method further comprises: The health status of the detected object is judged according to the results output by the drug-related detection model.

6. The drug abuse detection method according to claim 2, characterized in that: The flash stimulation of the eyes of the detected object also includes: The two front iris images of the detected object are collected successively by an infrared camera, and it is determined whether the pupil boundaries of the two front iris images collected successively remain stable. If so, the step of flashing the eyes of the detected object is performed, otherwise, this step is repeated.

7. The drug abuse detection method according to claim 1, characterized in that: The types of micro-expression features include one or more of yawning, tears, dull expression, blurred eyes, dullness, fear, dark skin and sunken eyes.

8. The drug abuse detection method according to any one of claims 1 to 7, characterized in that: The method further comprises: Performing facial feature extraction and / or iris feature extraction according to the first image sequence and / or the second image sequence respectively; Performing identity recognition on the detected object according to the facial features and / or iris features obtained through the facial feature extraction and / or the iris feature extraction; The drug taking history of the detected subject is determined according to the result of the identity recognition of the detected subject.

9. A drug abuse detection device, characterized in that: The device comprises: A first data acquisition module is used to simultaneously and continuously photograph the face of the detected object using a visible light camera and an infrared camera, respectively, to obtain at least one first image sequence including a face image and at least one second image sequence including an iris image; A micro-expression feature extraction module, used for acquiring at least one micro-expression feature of the detected object according to the first image sequence; an iris texture feature extraction module, configured to obtain the iris texture feature of the detected object according to the second image sequence; The first drug taking judgment module is used to judge drug taking of the detected object according to the at least one micro-expression feature and the iris texture feature.

10. A computer-readable storage medium for drug detection, characterized in that: It comprises a memory for storing processor executable instructions, and when the instructions are executed by the processor, the steps of the drug detection method according to any one of claims 1 to 8 are implemented.

11. A device for drug abuse detection, characterized in that: The method comprises at least one processor and a memory storing computer executable instructions, wherein when the processor executes the instructions, the steps of the drug detection method according to any one of claims 1 to 8 are implemented.

Images