An identity recognition method and system

Abstract

The application provides an identity recognition method and system, which comprises the following steps: collecting real-time authentication video information of a user's identity in an identity security recognition platform; performing video face feature, video color feature and video dynamic feature recognition on the collected real-time authentication video information; performing evaluation on the recognized video face recognition feature, video environment color recognition feature and video dynamic feature in a pre-constructed feature evaluation model; calculating according to the obtained face feature recognition result, video environment color feature recognition result and video dynamic feature recognition result; and activating the authentication permission of the identity security recognition platform based on the obtained identity verification index. In the embodiment of the application, in the identity recognition process, in addition to recognizing the face feature, diversified information such as video environment color and video dynamic feature which affect the face video are also recognized, and the recognized features are comprehensively recognized to achieve the purpose of effectively improving the accuracy of identity recognition.

Description

Technical Field

[0001] This invention relates to the field of video identity recognition technology, and in particular to an identity recognition method and system. Background Technology

[0002] Facial recognition, as one of the main identification technologies for identity verification, is widely used in various fields such as security, transportation, and payment, and has become inseparable from human social life.

[0003] In existing technologies, facial recognition is performed by collecting video information containing faces to further verify identities. However, in different scenarios, due to variations in lighting, angles, and obstructions, the collected video information will differ. Consequently, during identity verification, these differences in video information can lead to recognition errors or even failure to recognize individuals. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide an identity recognition method and system to solve the problem of failure or incorrect recognition when using existing technologies for identity recognition.

[0005] To address the above problems, embodiments of the present invention provide the following technical solutions:

[0006] The first aspect of this invention discloses an identity recognition method, the method comprising:

[0007] Collect real-time authentication video information of the user to be identified, and input the real-time authentication video information into a pre-built identity security recognition platform;

[0008] In the identity security recognition platform, the real-time authentication video information is subjected to video face feature recognition, video color feature recognition and video dynamic feature recognition respectively, and the corresponding video face recognition features, video environment color recognition features and video dynamic features are output.

[0009] The video face recognition features, video environment color recognition features, and video dynamic features are evaluated using a pre-built feature evaluation model to obtain face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results. The feature evaluation model is constructed based on sample evaluation information.

[0010] The identity security verification index is calculated based on the facial feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results.

[0011] The authentication permissions corresponding to the identity security recognition platform are activated based on the aforementioned identity verification indicators.

[0012] Preferably, the step of performing video facial feature recognition on the real-time authentication video information in the identity security recognition platform to obtain video facial recognition features includes:

[0013] In the video face recognition module of the identity security recognition platform, video face feature recognition is performed on the real-time authentication video information to obtain video face recognition features;

[0014] The video facial feature recognition includes:

[0015] Extract the facial video information from the real-time authentication video information;

[0016] Based on the facial video information, the eye region is divided, and iris detection is performed in the eye region to obtain the iris features of the eye.

[0017] Based on the facial video information, the facial features of the user to be identified are analyzed, the facial features of the user to be identified are located, and the facial feature proportion features are calculated based on the located facial features of the user to be identified.

[0018] By combining the iris features and the facial proportion features, video face recognition features are obtained.

[0019] Preferably, the step of performing video color feature recognition on the real-time authentication video information in the identity security recognition platform to obtain video environment color recognition features includes:

[0020] In the video color recognition module of the identity security recognition platform, video color feature recognition is performed on the real-time authentication video information to obtain video environment color recognition features;

[0021] The video color feature recognition includes:

[0022] Extract the environmental information of the user whose identity is to be identified from the real-time authentication video information;

[0023] Analyze the environmental information to obtain the first environmental color level information;

[0024] The environmental color gradation information is compared based on a preset identification environment comparison library to obtain video environment color recognition features.

[0025] Preferably, the step of comparing the first environment color level information based on a preset identification environment comparison library to obtain video environment color recognition features includes:

[0026] The environmental information is compared with the environmental information in the preset identifier environment comparison database;

[0027] If consistent information is found, it is determined that the environmental information of the user whose identity is to be identified is in the preset identification environment comparison database;

[0028] Facial recognition is performed on the user whose identity is to be identified to obtain the recognition angle of the user's face that can be identified;

[0029] Brightness comparison is performed in the preset identification environment comparison library to obtain second environmental color level information that matches the recognition angle;

[0030] The second environmental color level information is compared with the first environmental color level information, and the resulting information variance is used as a video environment color recognition feature.

[0031] Preferably, the step of performing video dynamic feature recognition on the real-time authentication video information in the identity security recognition platform to obtain video dynamic features includes:

[0032] The video dynamic recognition module of the identity security recognition platform performs video dynamic feature recognition on the real-time authentication video information to obtain video dynamic features;

[0033] The video dynamic feature recognition includes:

[0034] Obtain video action information from the real-time authentication video information;

[0035] Based on the video motion information, analyze the dynamic changes in the body of the user whose identity is to be identified to obtain the user's dynamic body information;

[0036] Extract the dynamic video frames from the video motion information to obtain continuous video frame information;

[0037] By combining the user's limb dynamic information and the video frame continuity information, video dynamic features are obtained.

[0038] Preferably, the process of pre-constructing a feature evaluation model based on the sample evaluation information includes:

[0039] Acquire sample face recognition features, sample video environment color recognition features, and sample video dynamic features;

[0040] A three-dimensional evaluation model is constructed by using the face recognition features of the sample as the X-axis, the environmental color recognition features of the sample video as the Y-axis, and the dynamic features of the sample video as the Z-axis.

[0041] The sample face recognition features, sample video environment color recognition features, and sample video dynamic features are used as inputs to the three-dimensional evaluation model. Feature proportion analysis is performed on the three-dimensional evaluation model until the sample face feature recognition results, the sample video environment color feature recognition results, and the sample video dynamic feature recognition results are output.

[0042] A second aspect of this invention discloses an identity recognition system, the system comprising:

[0043] A camera device is used to collect real-time authentication video information of the user to be identified and input the real-time authentication video information into a pre-built identity security identification platform;

[0044] The identity security recognition platform is used to perform video face feature recognition, video color feature recognition and video dynamic feature recognition on the input real-time authentication video information, and output the corresponding video face recognition features, video environment color recognition features and video dynamic features.

[0045] A feature evaluation model is used to evaluate the input video face recognition features, video environment color recognition features, and video dynamic features to obtain face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results. The feature evaluation model is constructed based on sample evaluation information.

[0046] An identity security verification index acquisition device is used to receive and calculate the identity security verification index based on the facial feature recognition result, video environment color feature recognition result, and video dynamic feature recognition result.

[0047] An activation device is used to activate the authentication permissions corresponding to the identity security recognition platform based on the identity verification indicators.

[0048] Preferably, the identity security identification platform that performs video facial feature recognition on the real-time authentication video information to obtain video facial recognition features includes:

[0049] The video face recognition module is used to extract face video information from the real-time authentication video information; divide the eye region based on the face video information, and perform iris detection in the eye region to obtain eye iris features; analyze the facial features of the user to be identified based on the face video information, locate the facial features of the user to be identified, and calculate the facial feature proportion features based on the located facial features of the user to be identified; combine the eye iris features and the facial feature proportion features to obtain video face recognition features.

[0050] Preferably, the identity security identification platform that performs video color feature recognition on the real-time authentication video information to obtain video environment color recognition features includes:

[0051] The video color recognition module is used to extract the environmental information of the user whose identity is to be identified in the real-time authentication video information; analyze the environmental information to obtain the first environmental color level information; and compare the environmental color level information with a preset identification environment comparison library to obtain video environmental color recognition features.

[0052] Preferably, the identity security identification platform that performs video dynamic feature recognition on the real-time authentication video information to obtain video dynamic feature recognition includes:

[0053] The video motion recognition module is used to acquire video action information from the real-time authentication video information;

[0054] Based on the video action information, analyze the limb dynamic change information of the user to be identified to obtain user limb dynamic information; extract dynamic video frames from the video action information to obtain video frame continuity information; combine the user limb dynamic information and the video frame continuity information to obtain video dynamic features.

[0055] Based on the above embodiments of the present invention, an identity recognition method and system are provided. The method collects real-time authentication video information of the user to be identified and inputs the real-time authentication video information into a pre-constructed identity security recognition platform. In the identity security recognition platform, the real-time authentication video information is subjected to video face feature recognition, video color feature recognition, and video dynamic feature recognition, outputting corresponding video face recognition features, video environment color recognition features, and video dynamic features. A pre-constructed feature evaluation model is used to evaluate the video face recognition features, video environment color recognition features, and video dynamic features, obtaining face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results. The feature evaluation model is constructed based on sample evaluation information. An identity security verification index is calculated based on the face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results. The authentication permissions corresponding to the identity security recognition platform are activated based on the identity verification index. In this embodiment of the present invention, in addition to recognizing face features during the identity recognition process, diverse information such as video environment color and video dynamic features that affect the face video are also recognized, and the recognized features are comprehensively used for identity recognition, thereby effectively improving the accuracy of identity recognition. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0057] Figure 1 This is a schematic diagram of an identity recognition system architecture disclosed in an embodiment of the present invention;

[0058] Figure 2 This is a flowchart illustrating an identity recognition method disclosed in an embodiment of the present invention;

[0059] Figure 3 This is a flowchart illustrating a method for constructing a feature evaluation model according to an embodiment of the present invention;

[0060] Figure 4 This is a schematic flowchart of a video face feature recognition method disclosed in an embodiment of the present invention;

[0061] Figure 5 This is a flowchart illustrating a video color feature recognition method disclosed in an embodiment of the present invention;

[0062] Figure 6 This is a flowchart illustrating a method for recognizing dynamic features in video, as disclosed in an embodiment of the present invention. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] In this application, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0065] like Figure 1The diagram shown is a schematic representation of an identity recognition system architecture disclosed in an embodiment of the present invention. The identity recognition system includes a smart camera 1 and a high-precision video recognition and processing system 2.

[0066] The smart camera 1 is connected to the high-precision video recognition and processing system 2.

[0067] The smart camera 1 collects real-time authentication video information of users whose identities are to be verified. Due to differences in lighting angle and brightness, the collected real-time video information varies accordingly, affecting the information recognition results to varying degrees. Therefore, by collecting the real-time authentication video information of the users whose identities are to be verified, it is used as the source of discriminative analysis for subsequent analysis to determine identity recognition features. The specific analysis is completed by the high-precision video recognition processing system 2.

[0068] The high-precision video recognition and processing system 2 includes an identity security recognition platform 21, a feature evaluation module 22, an identity security verification indicator acquisition module 23, and an authentication permission activation module 24.

[0069] The identity security recognition platform 21 is configured in the identity authentication interface. The identity security recognition platform includes a video face recognition layer, a video color recognition layer, and a video dynamic recognition layer. The identity security recognition platform is used to identify and verify the user's identity.

[0070] The identity security recognition platform can extract the identity identification feature information of the user whose identity is to be confirmed, thereby confirming the user's identity. The establishment of the identity security recognition platform provides the basic conditions for subsequent video identification.

[0071] The video face recognition layer is a network layer that identifies the user's facial features in the video. The video color recognition layer is a network layer that analyzes and identifies the ambient colors in the captured video. The video motion recognition layer is a network layer that extracts the dynamic features of the user's body movements to be verified.

[0072] After the identity security recognition platform 21 identifies the video face recognition features, video environment color recognition features, and video dynamic features, the video face recognition features, video environment color recognition features, and video dynamic features are input into the feature evaluation module 22 for feature evaluation.

[0073] The feature evaluation module 22 is a three-dimensional evaluation model, which is a model for evaluating the proportion of the video face recognition features, the video environment color recognition features, and the video dynamic features.

[0074] The three-dimensional evaluation model uses the preset video face recognition features as the X-axis, the preset color recognition features as the Y-axis, and the preset video dynamic features as the Z-axis. By converting the features into coordinates, the three evaluation feature dimensions are represented by three coordinate axes. Based on the three-dimensional coordinate system, the coordinate proportion of each feature is evaluated to obtain the corresponding face feature recognition results, color feature recognition results, and video dynamic feature recognition results.

[0075] Based on the analysis of the corresponding feature proportions of the three-dimensional evaluation model, the corresponding area ranges of the video face recognition features, the video environment color recognition features, and the video dynamic features can be clearly displayed. By comparing the corresponding area ranges of each feature, the corresponding feature proportions are determined and output as the face feature recognition results, the color feature recognition results, and the video dynamic feature recognition results to the identity security verification indicator acquisition module 23.

[0076] The identity security verification index acquisition module 23 determines the proportion of the influence of the face recognition feature, color recognition feature and video dynamic feature on the identity verification result based on the coordinate ratio of the input face recognition feature, color recognition feature and video dynamic feature. On this basis, it calculates the specific verification feature information of the user whose identity is to be confirmed and obtains the identity security verification index.

[0077] Based on the above identification results, the identity security verification indicator module 23 identifies the corresponding identity information according to the different degrees of influence of each identification feature, thereby further improving the accuracy of identity information identification.

[0078] The authentication permission activation module 24 performs similarity screening in the big data system based on the identity security verification indicators obtained by the identity security verification indicator acquisition module 23, thereby determining the user's identity in the real-time dynamic video. Further, it activates the corresponding authentication permissions of the identity security recognition platform.

[0079] In this embodiment of the invention, by analyzing and judging the real-time dynamic video for multiple dimensions of features, the final user identity information is determined, which effectively improves the recognition accuracy in the video recognition process.

[0080] Based on the identity recognition system architecture disclosed in the above embodiments of the present invention, such as Figure 2 The diagram shown is a flowchart of an identity recognition method disclosed in an embodiment of the present invention. The method includes the following steps:

[0081] S201: Collect real-time authentication video information of the user to be identified, and input the real-time authentication video information into a pre-built identity security identification platform.

[0082] In S201, the camera device may specifically be: Figure 1 The smart camera 1 shown in the image.

[0083] In the specific implementation of S201, real-time authentication video information of the user to be identified is collected through camera equipment or camera device, and the real-time authentication video information is input into a pre-built identity security recognition platform.

[0084] S202: In the identity security recognition platform, the real-time authentication video information is subjected to video face feature recognition, video color feature recognition and video dynamic feature recognition respectively, and the corresponding video face recognition features, video environment color recognition features and video dynamic features are output.

[0085] In S202, an identity security recognition platform is pre-configured on the identity authentication interface. This platform is used to identify and verify the user's identity.

[0086] The identity security recognition platform includes, but is not limited to, a video face recognition layer, a video color recognition layer, and a video motion recognition layer.

[0087] The video face recognition layer, video color recognition layer, and video motion recognition layer can perform targeted recognition and analysis on the collected video information to determine the user's identity information.

[0088] Specifically, the video face recognition layer is used to identify the user's facial features in real-time authentication video information and output video face recognition features.

[0089] The video color recognition layer is used to analyze and identify the ambient color in real-time authentication video information and output video ambient color recognition features.

[0090] The video dynamic recognition layer is used to extract the dynamic features of the body movements of the user whose identity is to be identified in the real-time authentication video information and output the video dynamic features.

[0091] In one embodiment of the present invention, the identity security identification platform may specifically be: Figure 1 The identity security identification platform 21 shown in the figure.

[0092] S203: The video face recognition features, video environment color recognition features, and video dynamic features are evaluated using a pre-built feature evaluation model to obtain the face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results.

[0093] In S203, there are many ways to construct the feature evaluation model.

[0094] In one embodiment of the present invention, the feature evaluation model is constructed based on sample evaluation information, and the specific construction process includes:

[0095] First, obtain the facial recognition features of the sample, the environmental color recognition features of the sample video, and the dynamic features of the sample video.

[0096] The sample face recognition features include a preset face recognition feature breadth and corresponding recognition level; the sample video environment color recognition features include a preset environment color gradation range; and the sample video dynamic features include preset limb dynamic information and preset video frame range information.

[0097] Secondly, a three-dimensional evaluation model is constructed by using the face recognition features of the sample as the X-axis, the environmental color recognition features of the sample video as the Y-axis, and the dynamic features of the sample video as the Z-axis.

[0098] Finally, the sample face recognition features, sample video environment color recognition features, and sample video dynamic features are used as inputs to the three-dimensional evaluation model. Feature proportion analysis is performed on the three-dimensional evaluation model until the sample face feature recognition results, the sample video environment color feature recognition results, and the sample video dynamic feature recognition results are output.

[0099] In one embodiment of the present invention, the construction process of the feature evaluation model is as follows: Figure 3 As shown, the main steps include the following:

[0100] S301: Obtain preset face recognition features, preset color recognition features, and preset video dynamic features.

[0101] In S301, the preset face recognition features refer to the preset face recognition feature breadth and corresponding level; the preset color recognition features refer to the preset environmental color gradation range; and the preset video dynamic features refer to the sum of limb dynamic changes and video frame range.

[0102] S302: Using the preset face recognition features as the x-axis, the preset color recognition features as the y-axis, and the preset video dynamic features as the z-axis, a three-dimensional evaluation model is constructed.

[0103] In the specific execution of S302, a three-dimensional Cartesian coordinate system is established based on the preset face recognition features, the preset color recognition features, and the preset video dynamic features. The preset face recognition features are used as the x-axis, the preset color recognition features as the y-axis, and the preset video dynamic features as the z-axis for coordinate transformation. An equivalent model is then established for simulation evaluation to obtain the three-dimensional evaluation model.

[0104] In one embodiment of the present invention, the feature evaluation model may specifically be as follows: Figure 1 Feature evaluation module 22 is shown in the figure.

[0105] In the specific implementation of S203, firstly, the pre-constructed feature evaluation model coordinates the input video face recognition features, video environment color recognition features, and video motion features to obtain video face recognition features, video environment color recognition features, and video motion features expressed in a three-dimensional coordinate system, where each one-dimensional coordinate corresponds to a recognition feature; then, the coordinate proportion of each feature is evaluated in the base three-dimensional coordinate system to obtain the corresponding face feature recognition results, video environment color feature recognition results, and video motion feature recognition results.

[0106] The specific evaluation process is as follows:

[0107] In the feature evaluation model, the whole is used as the base. The proportions of video face recognition features, video environment color recognition features, and video dynamic features are determined separately. The corresponding proportion of each feature is determined and used as the final recognition result to output the face feature recognition result, video environment color feature recognition result, and video dynamic feature recognition result.

[0108] Based on this, by analyzing the proportion of multiple identity recognition features, the degree of influence of each feature on the identity recognition result is determined, and corresponding recognition judgments are made accordingly, so that the recognition result is more accurate.

[0109] S204: Calculate the identity security verification index based on the facial feature recognition result, video environment color feature recognition result, and video dynamic feature recognition result.

[0110] In S204, the identity security verification index refers to a specific index that balances a series of influencing factors to determine the specific identity characteristics of the user whose identity is to be confirmed.

[0111] In the specific execution of S204, firstly, based on the coordinate ratios indicated by the facial feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results, the influence ratio of the facial recognition features, the color recognition features, and the video dynamic features on the identity verification results is determined; then, the identity security verification index is obtained based on this influence ratio.

[0112] S205: Activate the authentication permissions corresponding to the identity security recognition platform based on the identity verification indicators.

[0113] This invention discloses an identity recognition method. An identity security recognition platform is set up on the identity authentication interface. This platform performs user identity verification based on a video face recognition layer, a video color recognition layer, and a video motion recognition layer. Real-time authentication video information of the user to be identified is collected by a camera device and input into the pre-constructed identity security recognition platform. The identity security recognition platform performs video face feature recognition, video color feature recognition, and video motion feature recognition on the real-time authentication video information, outputting corresponding video face recognition features, video environment color recognition features, and video motion features. A pre-constructed feature evaluation model evaluates the video face recognition features, video environment color recognition features, and video motion features from three dimensions to obtain face feature recognition results, video environment color feature recognition results, and video motion feature recognition results. Further, based on the face feature recognition results, video environment color feature recognition results, and video motion feature recognition results, an identity security verification index is calculated. The authentication permissions corresponding to the identity security recognition platform are activated based on the identity verification index.

[0114] In this embodiment of the invention, by extracting and judging video face recognition features, video environment color recognition features, and video dynamic features from the real-time video information through multiple recognition network layers in the identity security recognition platform, and by performing information recognition and analysis of multiple feature dimensions, the problem of recognition errors or failure to recognize identities caused by not considering the corresponding recognition result in the prior art when performing identity recognition can be maximized, thereby effectively improving the accuracy of identity recognition.

[0115] Furthermore, feature proportion analysis is performed on the video face recognition features, video color recognition features, and video motion recognition features from three natural dimensions. Based on the degree of influence of each feature in the identity recognition process, targeted evaluation is conducted, which can further improve the verification accuracy.

[0116] Based on the identity recognition method disclosed in the above embodiments of the present invention, the following is executed: Figure 2 The identification of video face recognition features in S202 shown in the diagram specifically involves: performing video face recognition on the real-time authentication video information in the video face recognition module of the identity security identification platform to obtain video face recognition features. The specific process is as follows: Figure 4 As shown, it mainly includes:

[0117] S401: Extract the facial video information from the real-time authentication video information.

[0118] S402: Divide the eye region based on the face video information, and perform iris detection in the eye region to obtain the iris features of the eye.

[0119] In the specific execution of S402, the facial video information is divided into eye regions, and iris detection is performed in these eye regions. The iris refers to the ring-shaped portion located between the black pupil and the white sclera. This portion contains various intricate features such as spots, filaments, coronas, stripes, and crypts. Since the iris is formed during fetal development and remains unchanged throughout life, its unique characteristics are determined, allowing for identity recognition. Iris detection obtains the iris features, including the iris's color, tissue, and structure.

[0120] S403: Analyze the facial features of the user to be identified based on the facial video information, locate the facial features of the user to be identified, and calculate the facial feature proportion features based on the located facial features of the user to be identified.

[0121] In the specific execution of S403, based on the facial video information, a concrete recognition analysis is performed on the proportion, distribution, and facial features of the user whose identity is to be identified, to obtain the corresponding proportional features, namely, the facial feature proportion features.

[0122] S404: Combine the iris features and the facial proportion features to obtain video face recognition features.

[0123] In the specific execution of S404, based on the obtained facial proportion features and iris features, the video face recognition features are obtained. By extracting the face recognition features from the real-time video information, certain information support can be provided for user identity verification, serving as the basis for user identity determination.

[0124] Based on the identity recognition method disclosed in the above embodiments of the present invention, the following is executed: Figure 2 The video color feature recognition shown in S202 is specifically as follows: In the video color recognition module of the identity security recognition platform, video color feature recognition is performed on the real-time authentication video information to obtain video environment color recognition features. The specific process is as follows: Figure 5 As shown, it mainly includes:

[0125] S501: Extract the environmental information of the user whose identity is to be identified from the real-time authentication video information.

[0126] S502: Analyze the environmental information to obtain the first environmental color level information.

[0127] In the specific execution of S502, the environmental information of the user to be identified in real time is further analyzed in detail, the environmental color level is determined, and the first environmental color level information is obtained. The first environmental color level refers to the index standard that describes the brightness of the collected real-time authentication video, and is unrelated to the true color of people and objects in the environment.

[0128] S503: The first environment color level information is compared with the preset identification environment comparison library to obtain the video environment color recognition features.

[0129] In S503, a preset environment comparison library is used to sort and label the environment color levels from light to dark, i.e., the brightest is white and the darkest is black, and these are stored in the video color recognition module.

[0130] The ambient color gradation refers to the change in the inherent color of a person or object caused by the influence of the ambient color reflected from surrounding objects. Based on the influence of ambient color, the facial information of the user to be identified will also be affected, thus impacting the accuracy of information recognition.

[0131] Due to the influence of factors such as lighting, the ambient color is subject to real-time changes. Therefore, by executing S503, comparing the first ambient color level information with a preset identification environment comparison library, and adjusting based on the obtained video ambient color recognition features, the accuracy of user identification can be effectively improved.

[0132] In one embodiment of the present invention, the specific process of comparing the first environment color level information based on a preset identification environment comparison library to obtain video environment color recognition features includes:

[0133] Step 1: Compare the environmental information with the environmental information in the preset identifier environment comparison database. If a match is found, proceed to Step 2.

[0134] Step 2: Determine that the environmental information of the user whose identity is to be identified is in the preset identification environment comparison database.

[0135] Execute steps 1 and 2, compare the real-time environment information of the user to be identified with the preset identifier environment comparison library, and determine whether the real-time environment of the user to be identified is in the preset identifier environment comparison library. If the real-time environment of the user to be identified is in the preset identifier environment comparison library, execute step 3.

[0136] Step 3: Perform facial recognition on the user whose identity is to be identified to obtain the recognition angle of the user's face.

[0137] In the specific execution of step 3, the angle at which the user to be identified is determined, and the identification angle of the user is obtained. Based on the identification angle, the user to be identified can be subjected to corresponding facial recognition.

[0138] Step 4: Perform brightness comparison in the preset identification environment comparison library to obtain the second environment color level information that matches the identification angle.

[0139] In the specific execution of step 4, the second environmental color level information that matches the recognition angle is determined by comparing the brightness in the preset identification environment comparison library.

[0140] Step 5: Compare the second ambient color level information with the first ambient color level information, and use the resulting information variance as a video ambient color recognition feature.

[0141] In the specific execution of step 5, the second environmental color level information corresponding to the recognition angle and the first environmental color level information are compared to determine the information difference between the two, and then the environmental color comparison result, i.e., the information difference, is obtained; the obtained information difference is used as the video environmental color recognition feature.

[0142] Based on this, the difference in lighting at different angles can easily cause information differences between the matching environmental color gradation information corresponding to the first recognition angle, thus affecting the recognition result. Taking the above error into account during the user identification process can effectively reduce the impact of external factors on the recognition accuracy.

[0143] Based on the identity recognition method disclosed in the above embodiments of the present invention, the following is executed: Figure 2 The video dynamic feature recognition in S202 shown in the diagram specifically involves: performing video dynamic feature recognition on the real-time authentication video information in the video dynamic recognition module of the identity security recognition platform to obtain the video dynamic features. The specific process is as follows: Figure 6 As shown, it mainly includes:

[0144] S601: Obtain video action information from the real-time authentication video information.

[0145] During the execution of S601, the video action information of the user whose identity is to be identified is captured to obtain the video action information.

[0146] S602: Analyze the dynamic changes in the body of the user whose identity is to be identified based on the video motion information to obtain the user's dynamic body information.

[0147] During the execution of S602, based on the video action information, the dynamic changes of the body of the user to be identified during the identity verification process are extracted and analyzed to obtain the user's body dynamics, i.e., user body dynamic information.

[0148] S603: Extract the dynamic video frames from the video motion information to obtain continuous video frame information.

[0149] In S603, the frame is the basic unit of video information. When the frame rate is below 15 frames per second, video stuttering will occur. Similarly, as the frame rate increases, the corresponding amount of video data also increases, and the video playback becomes more consistent to a certain extent.

[0150] During the execution of S603, dynamic video frames are extracted from the video action information, i.e., multiple still frames in the dynamic video, to obtain the continuity of video frames, i.e., video frame continuity information.

[0151] S604: Combine the user's limb dynamic information and the video frame continuity information to obtain video dynamic features.

[0152] During the execution of S604, the video dynamic features are output based on the user's limb dynamic information and the video frame continuity information. By extracting the video dynamic features, the breadth of judgment information can be further improved when verifying the identity of the user to be confirmed, enabling diversified information verification and improving verification accuracy.

[0153] Based on the identity recognition system disclosed in the above embodiments of the present invention, the present invention also discloses an identity recognition system, which includes: a camera device, an identity security recognition platform, a feature evaluation model, an identity security verification indicator acquisition device, and an activation device.

[0154] A camera device is used to collect real-time authentication video information of the user to be identified and input the real-time authentication video information into a pre-built identity security identification platform.

[0155] The camera device can be specifically described as Figure 1 The smart camera 1 shown in the image.

[0156] The identity security recognition platform is used to perform video face feature recognition, video color feature recognition, and video dynamic feature recognition on the input real-time authentication video information, and output the corresponding video face recognition features, video environment color recognition features, and video dynamic features.

[0157] This identity security verification platform can be specifically defined as Figure 1 The identity security identification platform 21 shown in the figure.

[0158] A feature evaluation model is used to evaluate the input video face recognition features, video environment color recognition features, and video dynamic features to obtain face feature recognition results, video environment color feature recognition results, and video dynamic feature recognition results. The feature evaluation model is constructed based on sample evaluation information.

[0159] This feature evaluation model can be specifically described as follows: Figure 1 Feature evaluation module 22 is shown in the figure.

[0160] An identity security verification index acquisition device is used to receive and calculate the identity security verification index based on the facial feature recognition result, video environment color feature recognition result, and video dynamic feature recognition result.

[0161] The identity security verification indicator acquisition device can be specifically as follows: Figure 1 The identity security verification indicator acquisition module 23 is shown in the figure.

[0162] An activation device is used to activate the authentication permissions corresponding to the identity security recognition platform based on the identity verification indicators.

[0163] The activation device can be specifically as follows: Figure 1 The authentication permission activation module 24 is shown in the figure.

[0164] In one embodiment of the present invention, a video facial feature recognition platform is obtained by performing video facial recognition feature recognition on the real-time authentication video information, comprising:

[0165] The video face recognition module is used to extract face video information from the real-time authentication video information; divide the eye region based on the face video information, and perform iris detection in the eye region to obtain eye iris features; analyze the facial features of the user to be identified based on the face video information, locate the facial features of the user to be identified, and calculate the facial feature proportion features based on the located facial features of the user to be identified; combine the eye iris features and the facial feature proportion features to obtain video face recognition features.

[0166] In one embodiment of the present invention, a video color feature recognition platform 702 is obtained by performing video environment color recognition feature recognition on the real-time authentication video information, including:

[0167] The video color recognition module is used to extract the environmental information of the user whose identity is to be identified in the real-time authentication video information; analyze the environmental information to obtain the first environmental color level information; and compare the environmental color level information with a preset identification environment comparison library to obtain video environmental color recognition features.

[0168] The video color recognition module compares the environmental color level information based on a preset identification environment comparison library to obtain the video environment color recognition features. Specifically, it is used for:

[0169] The environmental information is compared with the environmental information in the preset identification environment comparison library; if there is consistent information, the environmental information of the user to be identified is determined to be in the preset identification environment comparison library; facial recognition is performed on the user to be identified to obtain the recognition angle of the user's face; brightness comparison is performed in the preset identification environment comparison library to obtain the second environmental color level information that matches the recognition angle; the second environmental color level information is compared with the first environmental color level information, and the obtained information asymmetry is used as the video environment color recognition feature.

[0170] In one embodiment of the present invention, a video dynamic feature recognition platform is obtained by performing video dynamic feature recognition on the real-time authentication video information, including:

[0171] The video motion recognition module is used to acquire video action information from the real-time authentication video information;

[0172] Based on the video action information, analyze the limb dynamic change information of the user to be identified to obtain user limb dynamic information; extract dynamic video frames from the video action information to obtain video frame continuity information; combine the user limb dynamic information and the video frame continuity information to obtain video dynamic features.

[0173] The identity recognition system disclosed in this embodiment of the invention further includes:

[0174] The identity security recognition platform settings module is used to configure the identity security recognition platform in the identity authentication interface.

[0175] The identity recognition system disclosed in this embodiment of the invention further includes:

[0176] The first pre-built module is used to construct the feature evaluation model. Specifically:

[0177] Acquire sample face recognition features, sample video environment color recognition features, and sample video dynamic features; construct a three-dimensional evaluation model by using the sample face recognition features as the X-axis, the sample video environment color recognition features as the Y-axis, and the sample video dynamic features as the Z-axis; use the sample face recognition features, sample video environment color recognition features, and sample video dynamic features as inputs to the three-dimensional evaluation model, perform feature proportion analysis on the three-dimensional evaluation model, until the sample face feature recognition result, the sample video environment color feature recognition result, and the sample video dynamic feature recognition result are output.

[0178] The identity recognition system disclosed in this embodiment of the invention further includes:

[0179] The second pre-built module is used to construct the feature evaluation model. Specifically:

[0180] Preset facial recognition features, preset color recognition features, and preset video motion features are obtained. A three-dimensional evaluation model is constructed using the preset facial recognition features as the x-axis, the preset color recognition features as the y-axis, and the preset video motion features as the z-axis.

[0181] In this embodiment of the invention, by extracting and judging video face recognition features, video environment color recognition features, and video dynamic features from the real-time video information through multiple recognition network layers in the identity security recognition platform, and by performing information recognition and analysis of multiple feature dimensions, the problem of recognition errors or failure to recognize identities caused by not considering the corresponding recognition result in the prior art when performing identity recognition can be maximized, thereby effectively improving the accuracy of identity recognition.

[0182] Furthermore, feature proportion analysis is performed on the video face recognition features, video color recognition features, and video motion recognition features from three natural dimensions. Based on the degree of influence of each feature in the identity recognition process, targeted evaluation is conducted, which can further improve the verification accuracy.

[0183] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, for system or system embodiments, since they are fundamentally similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. Components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0184] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0185] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An identity recognition method, characterized by, The method comprises: Collect real-time authentication video information of a user whose identity is to be identified, and input the real-time authentication video information into a pre-constructed identity security identification platform; In the identity security identification platform, video face feature recognition, video color feature recognition and video dynamic feature recognition are performed on the real-time authentication video information respectively, and corresponding video face recognition features, video environment color recognition features and video dynamic features are output; In a pre-constructed feature evaluation model, the proportions of the video face recognition features, the video environment color recognition features and the video dynamic features are evaluated respectively, taking the whole as the base, to determine the corresponding proportion results of each feature as the face feature recognition result, the video environment color feature recognition result and the video dynamic feature recognition result, and the feature evaluation model is constructed based on sample evaluation information; According to the coordinate proportion indicated by the face feature recognition result, the video environment color feature recognition result and the video dynamic feature recognition result, the influence degree proportion value of the face feature recognition result, the video environment color feature recognition result and the video dynamic feature recognition result on the identity verification result is determined, and an identity security verification index is obtained according to the influence degree proportion value; Based on the identity verification index, the authentication permission corresponding to the identity security identification platform is activated; The process of pre-construction of the feature evaluation model based on the sample evaluation information comprises: Obtaining sample face recognition features, sample video environment color recognition features and sample video dynamic features; Taking the sample face recognition features as the X-axis, the sample video environment color recognition features as the Y-axis and the sample video dynamic features as the Z-axis, a three-dimensional evaluation model is constructed; The sample face recognition features, the sample video environment color recognition features and the sample video dynamic features are taken as inputs of the three-dimensional evaluation model, and feature proportion analysis is performed in the three-dimensional evaluation model until sample face feature recognition results, sample video environment color feature recognition results and sample video dynamic feature recognition results are output.

2. The method of claim 1, wherein, The video face feature recognition of the real-time authentication video information in the identity security identification platform comprises: Video face feature recognition of the real-time authentication video information is performed in a video face recognition module of the identity security identification platform to obtain video face recognition features; The video face feature recognition comprises: Extracting face video information from the real-time authentication video information; Based on the face video information, an eye region is divided, and iris detection is performed in the eye region to obtain eye iris features; Based on the face video information, the features of the five organs of the user whose identity is to be identified are analyzed, the five organs of the user whose identity is to be identified are positioned, and five organ proportion features are calculated based on the positioned five organs of the user whose identity is to be identified; The eye iris features and the five organ proportion features are combined to obtain video face recognition features.

3. The method of claim 1, wherein, The video color feature recognition of the real-time authentication video information in the identity security identification platform comprises: The real-time authentication video information is subjected to video color feature recognition in a video color recognition module of the identity security recognition platform, to obtain video environment color recognition features; The video color feature recognition comprises: extracting environment information in which an identity to-be-recognized user is located in the real-time authentication video information; analyzing the environment information to obtain first environment color scale information; comparing the environment color scale information with a preset identification environment comparison library to obtain video environment color recognition features.

4. The method of claim 3, wherein, The comparison of the environment color scale information with the preset identification environment comparison library to obtain video environment color recognition features comprises: comparing the environment information with environment information in the preset identification environment comparison library; if there is consistent information, it is determined that the environment information in which the identity to-be-recognized user is located is in the preset identification environment comparison library; performing face recognition on the identity to-be-recognized user to obtain a recognition angle of an identifiable identity to-be-recognized user face; performing lightness comparison in the preset identification environment comparison library to obtain second environment color scale information matched with the recognition angle; comparing the second environment color scale information with the first environment color scale information, and taking the obtained information disparity degree as the video environment color recognition features.

5. The method of claim 1, wherein, The real-time authentication video information is subjected to video dynamic feature recognition in the identity security recognition platform, to obtain video dynamic features, which comprises: The real-time authentication video information is subjected to video dynamic feature recognition in a video dynamic recognition module of the identity security recognition platform, to obtain video dynamic features; The video dynamic feature recognition comprises: obtaining video action information in the real-time authentication video information; analyzing body dynamic change information of an identity to-be-recognized user according to the video action information, to obtain user body dynamic information; extracting dynamic video frames in the video action information, to obtain video frame continuous information; collecting the user body dynamic information and the video frame continuous information, to obtain video dynamic features.

6. An identity recognition system characterized by, The system comprises: a camera device configured to collect real-time authentication video information of a to-be-identified user identity, and input the real-time authentication video information into a pre-constructed identity security recognition platform; the identity security recognition platform configured to perform video face feature recognition, video color feature recognition and video dynamic feature recognition on the input real-time authentication video information respectively, and output corresponding video face recognition features, video environment color recognition features and video dynamic features; a feature evaluation model configured to take the whole as a base, respectively evaluate proportions of the input video face recognition features, video environment color recognition features and video dynamic features, and determine corresponding proportion results of each feature as face feature recognition results, video environment color recognition results and video dynamic feature recognition results, the feature evaluation model being constructed based on sample evaluation information; The identity security verification index acquisition device is configured to receive and determine, according to the coordinate proportion indicated by the face feature recognition result, the video environment color feature recognition result and the video dynamic feature recognition result, the influence degree proportion value of the face feature recognition result, the video environment color feature recognition result and the video dynamic feature recognition result on the identity verification result, and obtain the identity security verification index according to the influence degree proportion value. The activation device is configured to activate the authentication permission corresponding to the identity security recognition platform based on the identity verification index. The first pre-construction module is configured to construct the feature evaluation model, specifically configured to: obtain sample face recognition features, sample video environment color recognition features and sample video dynamic features; construct a three-dimensional evaluation model by taking the sample face recognition features as the X-axis, the sample video environment color recognition features as the Y-axis and the sample video dynamic features as the Z-axis; take the sample face recognition features, the sample video environment color recognition features and the sample video dynamic features as the input of the three-dimensional evaluation model, and perform feature proportion analysis on the three-dimensional evaluation model until the sample face feature recognition result, the sample video environment color feature recognition result and the sample video dynamic feature recognition result are output.

7. The system of claim 6, wherein, The identity security recognition platform for video face feature recognition of the real-time authentication video information includes: A video face recognition module is configured to extract face video information from the real-time authentication video information; divide an eye region based on the face video information, and perform iris detection on the eye region to obtain eye iris features; analyze the features of the five organs of an identity-to-be-recognized user based on the face video information, position the five organs of the identity-to-be-recognized user, and calculate the five-organ proportion features based on the positioned five organs of the identity-to-be-recognized user; and combine the eye iris features and the five-organ proportion features to obtain video face recognition features.

8. The system of claim 6, wherein, The identity security recognition platform for video color feature recognition of the real-time authentication video information includes: A video color recognition module is configured to extract environment information in which an identity-to-be-recognized user is located from the real-time authentication video information; analyze the environment information to obtain first environment color level information; compare the environment color level information based on a preset identification environment comparison library to obtain video environment color recognition features.

9. The system of claim 6, wherein, The identity security recognition platform for video dynamic feature recognition of the real-time authentication video information includes: A video dynamic recognition module is configured to obtain video action information from the real-time authentication video information; analyze the body dynamic change information of an identity-to-be-recognized user based on the video action information to obtain user body dynamic information; extract dynamic video frames from the video action information to obtain video frame continuous information; and combine the user body dynamic information and the video frame continuous information to obtain video dynamic features.

Images