An identity recognition method, device and system based on millimeter wave radar technology

Abstract

The present invention is an identity recognition method, device and system based on millimeter wave radar technology, and the method includes the following steps: 1) Millimeter wave radar capture action: identify and capture action through millimeter wave radar. 2) Feature extraction: identify physical objects through millimeter wave radar and extract features through the motion trajectory of the physical object. 3) Classification recognition: classify the features. 4) Model training: train the classified data based on a dual-modal fusion algorithm; 5) Data storage: save the trained data in a carrier. 6) Actual output application. The present invention can integrate gestures and individual biometrics, accurately identify user identities and realize convenient interaction, and has the characteristics of high accuracy, complete privacy security and strong adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of motion recognition and biometric identification, and in particular to an identity recognition method, device and system based on millimeter wave radar technology. Background Art

[0002] At present, identity recognition technologies mainly include fingerprint recognition, face recognition, iris recognition, and voiceprint recognition, etc. These technologies have shortcomings in certain scenarios, for example: fingerprint recognition is easily affected by surface contamination or damage; face recognition is easily affected by lighting, angle changes or occlusion; iris recognition requires user cooperation and the equipment cost is relatively high; voiceprint recognition is sensitive to environmental noise.

[0003] In recent years, millimeter waves have been widely used to acquire characteristic signals. Compared with Wi-Fi, millimeter waves have higher frequencies and shorter wavelengths, can achieve higher spatial resolution and fine imaging, are suitable for the detection of complex surfaces and tiny targets, and have significant advantages in signal acquisition and feature recognition. At the same time, millimeter waves have stronger directionality and penetration performance, and perform well in environmental perception, target recognition and other fields.

[0004] Most of the existing millimeter wave radar-based recognition technologies are limited to motion detection or simple recognition, and have not yet deeply integrated gestures and individual identity features, resulting in insufficient recognition accuracy and system applicability. To this end, a method, device and system for identity recognition by extracting bone features from the sound of knuckle knocking is designed to overcome the above problems. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide an identity recognition method, device and system based on millimeter wave radar technology. The invention can integrate gestures and individual biometrics to accurately identify the user's identity and achieve convenient interaction, and has the characteristics of high precision, complete privacy security and strong adaptability.

[0006] The present invention is implemented by the following technical solution: an identity recognition method based on millimeter wave radar technology, the method comprising the following steps: 1) Millimeter-wave radar captures motion: Use millimeter-wave radar to identify and capture motion; 2) Feature extraction: Identify the object through millimeter-wave radar and extract features through the movement trajectory of the object; 3) Classification and identification: classify the feature; 4) Model training: training the classified data based on the bimodal fusion algorithm; 5) Data storage: After training, the data is stored in a carrier; 6) Actual output application.

[0007] Preferably, in step 1), the user's gestures and gesture movements are identified and captured, the gesture movements include static gestures and dynamic hand holding, and feature point cloud data are generated.

[0008] Preferably: the point cloud data includes the motion trajectory and speed or acceleration changes of dynamic gesture features, the hand size and micro-motion features of static identity features, wherein the hand size includes the fingernail size and the size from the upper fingers to the middle joints, and the size from the middle joints to the lower fingers.

[0009] Preferably, the acquired features are based on a bimodal fusion algorithm, using a convolutional neural network (CNN) and a long short-term memory network (LSTM) to jointly learn dynamic gesture features and static identity features.

[0010] An identity recognition device based on millimeter wave radar technology, the device comprising: Millimeter wave radar module: used to capture the user's specific parts and the movements of these parts in real time; and generate point cloud data including hand movement trajectory, speed, acceleration and micro-motion characteristics; Feature extraction module: generates point cloud data including the motion trajectory, speed, acceleration and micro-motion features of the part; Classification and recognition module: extracts multi-dimensional features from point cloud data, including dynamic gesture features: motion trajectory and speed change, static identity features: hand size and micro-motion features; Training module: Based on the bimodal fusion algorithm, the convolutional neural network and long short-term memory network are used to jointly learn dynamic gesture features and static identity features, and output user identity and gesture category; Data storage module: used to store user feature data and recognition models, supporting offline updates and dynamic expansion; Output interface module: It realizes various forms of output, such as identity authentication results and gesture control instructions, so that it can be seamlessly connected with other smart devices or systems.

[0011] A smart home identity recognition system, the system installs the above method or system in a smart home, realizes identity recognition by capturing user gestures, and uses the program provided by the smart home to complete light switch and curtain control operations.

[0012] An intelligent security identity recognition and verification system, the system installs the above method or system in a security system, realizes door unlocking by recognizing the gestures and identity features of the user in front of the door, and triggers an alarm when an unauthorized user comes to identify.

[0013] The beneficial effects of the present invention are as follows: Compared with existing identity recognition and gesture recognition methods, the method implemented by the present invention has significant improvements in many aspects: by integrating gestures and individual biometrics, the present method achieves an identity recognition accuracy of 97%. In view of the many advantages of millimeter waves, the present invention is insensitive to ambient lighting, occlusion and noise, and is applicable to a variety of complex scenarios. The present invention uses millimeter wave radar to collect user data, in which the user's biometrics are basically impossible to replicate, which helps to protect user privacy. The present invention has a wide range of application scenarios and can be used in smart homes, identity authentication, medical rehabilitation, human-computer interaction and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a system framework diagram of the present invention. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to more clearly understand the objectives, technical solutions and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0016] In the description of the present invention, it is necessary to understand that the orientation or position relationship indicated by the terms such as "up", "down", "left", "right", "inside", "outside", "horizontal" and "vertical" is based on the orientation or position relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention, rather than indicating or implying that the device or original referred to must have a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0017] The present invention will be described in detail below with reference to the accompanying drawings: Figure 1 As shown, an identity recognition method based on millimeter wave radar technology comprises the following steps: 1) Millimeter-wave radar captures motion: Use millimeter-wave radar to identify and capture motion; 2) Feature extraction: Identify the object through millimeter-wave radar and extract features through the movement trajectory of the object; 3) Classification and identification: classify the feature; 4) Model training: training the classified data based on the bimodal fusion algorithm; 5) Data storage: After training, the data is stored in a carrier; 6) Actual output application.

[0018] In the step 1), the user's gestures and gesture movements are identified and captured, the gesture movements including static gestures and dynamic hand holding, and feature point cloud data are generated.

[0019] The point cloud data includes the motion trajectory and speed or acceleration changes of dynamic gesture features, the hand size and micro-motion features of static identity features, wherein the hand size includes the size of the fingernail and the size from the upper finger to the middle joint, and the size from the middle joint to the lower finger.

[0020] The acquired features are based on a bimodal fusion algorithm, using a convolutional neural network (CNN) and a long short-term memory network (LSTM) to jointly learn dynamic gesture features and static identity features.

[0021] An identity recognition device based on millimeter wave radar technology, the device comprising: Millimeter wave radar module: used to capture the user's specific parts and the movements of these parts in real time; and generate point cloud data including hand movement trajectory, speed, acceleration and micro-motion characteristics; Feature extraction module: generates point cloud data including the motion trajectory, speed, acceleration and micro-motion features of the part; Classification and recognition module: extracts multi-dimensional features from point cloud data, including dynamic gesture features: motion trajectory and speed change, static identity features: hand size and micro-motion features; Training module: Based on the bimodal fusion algorithm, the convolutional neural network and long short-term memory network are used to jointly learn dynamic gesture features and static identity features, and output user identity and gesture category; Data storage module: used to store user feature data and recognition models, supporting offline updates and dynamic expansion; Output interface module: It realizes various forms of output, such as identity authentication results and gesture control instructions, so that it can be seamlessly connected with other smart devices or systems.

[0022] A smart home identity recognition system, the system installs the above method or system in a smart home, realizes identity recognition by capturing the user's gestures, and uses the program built into the smart home to complete the light switch and curtain control operations. For example, when the user sends a light-on command to the radar through a specific gesture, the system not only verifies the user's identity, but also executes the command.

[0023] An intelligent security identity recognition and verification system, the system installs the above method or system in a security system, realizes door unlocking by recognizing the gestures and identity features of the user in front of the door, and triggers an alarm when an unauthorized user comes to identify.

[0024] The specific functions of the identity recognition device in the present invention are as follows: 1. Millimeter wave radar module: Used to capture user gestures in real time and generate point cloud data including hand motion trajectory, speed, acceleration and micro-motion features.

[0025] 2. Feature extraction module: Signal processing and deep learning methods are used to extract multidimensional features from point cloud data, including dynamic gesture features (motion trajectory, speed change) and static identity features (hand size, micro-motion features).

[0026] 3. Classification and recognition module: Based on the bimodal fusion algorithm, the dynamic gesture features and static identity features are jointly learned using convolutional neural networks (CNN) and long short-term memory networks (LSTM) to output user identity and gesture category.

[0027] 4. Data storage module: Used to store user feature data and recognition models, supporting offline updates and dynamic expansion.

[0028] 5. Output interface module: It realizes various forms of output, such as identity authentication results and gesture control instructions, so that it can be seamlessly connected with other smart devices or systems.

[0029] The method implemented by the present invention has significant improvements in many aspects, as follows: This method achieves an identity recognition accuracy of 97% by integrating gestures and individual biometrics. In view of the many advantages of millimeter waves, the present invention is insensitive to ambient lighting, occlusion and noise, and is applicable to a variety of complex scenarios. The present invention uses millimeter wave radar to collect user data, in which the user's biometrics are basically impossible to replicate, which helps to protect user privacy. The present invention has a wide range of application scenarios and can be used in smart homes, identity authentication, medical rehabilitation, human-computer interaction and other fields.

[0030] Anyone familiar with the technology may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. An identity recognition method based on millimeter wave radar technology, characterized in that: The method comprises the following steps: 1) Millimeter-wave radar captures motion: Use millimeter-wave radar to identify and capture motion; 2) Feature extraction: Identify the object through millimeter-wave radar and extract features through the movement trajectory of the object; 3) Classification and identification: classify the feature; 4) Model training: training the classified data based on the bimodal fusion algorithm; 5) Data storage: After training, the data is stored in a carrier; 6) Actual output application.

2. The identity recognition method based on millimeter wave radar technology according to claim 1 is characterized in that: In the step 1), the user's gestures and gesture movements are identified and captured, the gesture movements include static gestures and dynamic hand holding, and feature point cloud data are generated.

3. The identity recognition method based on millimeter wave radar technology according to claim 2 is characterized in that: The point cloud data includes the motion trajectory and speed or acceleration changes of dynamic gesture features, the hand size and micro-motion features of static identity features, wherein the hand size includes the size of the fingernail and the size from the upper finger to the middle joint, and the size from the middle joint to the lower finger.

4. The identity recognition method based on millimeter wave radar technology according to claim 3 is characterized in that: The acquired features are based on a bimodal fusion algorithm, using a convolutional neural network (CNN) and a long short-term memory network (LSTM) to jointly learn dynamic gesture features and static identity features.

5. An identity recognition device based on millimeter wave radar technology, characterized in that: The device comprises: Millimeter wave radar module: used to capture the user's specific parts and the movements of these parts in real time; and generate point cloud data including hand movement trajectory, speed, acceleration and micro-motion characteristics; Feature extraction module: generates point cloud data containing the motion trajectory, speed, acceleration and micro-motion features of the part; classification and recognition module: extracts multi-dimensional features from the point cloud data, including dynamic gesture features: motion trajectory and speed changes, static identity features: hand size and micro-motion features; Training module: Based on the bimodal fusion algorithm, the convolutional neural network and long short-term memory network are used to jointly learn dynamic gesture features and static identity features, and output user identity and gesture category; Data storage module: used to store user feature data and recognition models, supporting offline updates and dynamic expansion; Output interface module: It realizes various forms of output, such as identity authentication results and gesture control instructions, so that it can be seamlessly connected with other smart devices or systems.

6. A smart home identity recognition system, characterized in that: The system installs the method or system described in claim 4 or 5 in a smart home, realizes identity recognition by capturing the user's gestures, and uses the program provided in the smart home to complete the light switch and curtain control operations.

7. An intelligent security identity recognition and verification system, characterized in that: The system installs the method or system described in claim 4 or 5 in a security system, and unlocks the door by identifying the gestures and identity features of the user in front of the door. When an unauthorized user is identified, an alarm is triggered.

Images