Single-lens multifunctional identification module

Abstract

The utility model provides a single-lens multifunctional identification module, the module comprises an infrared light board, a double-filter IRCUT lens and an algorithm board, the algorithm board is provided with a main control chip, the double-filter IRCUT lens is arranged on the algorithm board through the infrared light board, and by adopting the ISP function of the double-filter IRCUT lens and an MCU, a large wide angle is realized, and the recognition efficiency is improved. And the infrared lamp bead board is designed to be integrated with the algorithm board, so that the whole module is integrated and is convenient to install and use.

Description

Technical Field

[0001] The utility model relates to the technical field of human body recognition, in particular to a single-lens multifunctional recognition module. Background Art

[0002] Currently, smart recognition door locks on the market generally use recognition module structures such as face and palm vein to collect facial image information of the person being identified. The storage unit of the smart recognition door lock pre-stores image information. When the collected image information is compared with the image information pre-stored in the storage unit and is successful, the smart recognition door lock will be unlocked. On this basis, in order to achieve monitoring, the smart recognition door lock will also integrate a cat's eye function.

[0003] To realize face, palm vein recognition and visual cat's eye functions at the same time, multiple cameras are generally used for recognition, which requires a high cost. Utility Model Content

[0004] The utility model aims to provide a single-lens multifunctional recognition module.

[0005] To achieve the above purpose, the technical solution of the utility model is as follows:

[0006] A single-lens multifunctional recognition module, comprising an infrared light board, a dual-filter IRCUT lens, and an algorithm board;

[0007] The algorithm board is provided with a main control chip;

[0008] The dual filter IRCUT lens is set on the algorithm board through the infrared light board;

[0009] The main control chip is electrically connected to the dual filter IRCUT lens;

[0010] The algorithm board is provided with a data sending pin and a data receiving end connected to the host computer, a data receiving pin and a data sending end connected to the host computer;

[0011] The main control chip is provided with a UVC / UAC signal interface and is connected to the cat-eye board;

[0012] The main control chip is provided with a control signal interface and is connected to the lock control board.

[0013] Preferably, the algorithm board is also provided with an infrared lamp, a speaker and a microphone, and the infrared lamp, the speaker and the microphone are electrically connected to the main control chip.

[0014] Preferably, the algorithm board is also provided with a storage chip.

[0015] Preferably, the algorithm board is also provided with a power supply module.

[0016] The utility model realizes a wide-angle, high-definition visual cat's eye function by adopting a dual-filter IRCUT lens and the ISP function of the MCU. In addition, by designing an infrared lamp bead board and integrating it with the algorithm board, the overall module is integrated and easy to install and use.

[0017] The utility model is composed of a main control board, an infrared lamp bead board and a dual-filter IRCUT lens. The overall structure is similar to an independent cat's eye. It uses the characteristics of the infrared / white light dual filters of the lens to see high-definition color cat's eye images in daytime environments and high-definition black and white images at night, thereby realizing the acquisition function. In conjunction with the recognition algorithm built into the main control chip, the storage unit of the intelligent recognition door lock is pre-stored with image information. When the collected image information is compared with the image information pre-stored in the storage unit and is successful, the intelligent recognition door lock will be unlocked, thereby realizing the simultaneous realization of face, palm vein recognition and visual cat's eye functions, greatly reducing manufacturing costs. The infrared light, speaker and microphone can also realize the interaction between users and devices, improving the usage experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the overall structure of the utility model.

[0019] Figure 2 It is an exploded view of the overall structure of the utility model.

[0020] Figure 3 It is a schematic diagram of module connection of the present utility model. DETAILED DESCRIPTION

[0021] The present invention is described in detail below with reference to the accompanying drawings.

[0022] Example 1

[0023] like Figure 1-3 As shown, a single-lens multifunctional recognition module includes an infrared light board 1, a dual-filter IRCUT lens 2 and an algorithm board 3;

[0024] The algorithm board 3 is provided with a main control chip 4;

[0025] The dual filter IRCUT lens 2 is set on the algorithm board 3 through the infrared light board 1;

[0026] The main control chip 4 is electrically connected to the dual filter IRCUT lens 2;

[0027] The algorithm board 3 is provided with a data sending pin and a data receiving terminal connected to the host computer, a data receiving pin and a data sending terminal connected to the host computer;

[0028] The main control chip 4 is provided with a UVC / UAC signal interface and is connected to the cat's eye board 10;

[0029] The main control chip 4 is also electrically connected to the lock control board 11 .

[0030] Preferably, the algorithm board 3 is further provided with an infrared lamp 5 , a speaker 6 and a microphone 7 , and the infrared lamp 5 , the speaker 6 and the microphone 7 are electrically connected to the main control chip 4 .

[0031] Preferably, a storage chip 8 is also provided on the algorithm board 3 .

[0032] Preferably, the algorithm board 3 is further provided with a power supply module 9 .

[0033] In this embodiment, the communication principle between this module and the user device is as follows: The face recognition module uses a UART communication interface and adopts half-duplex asynchronous serial communication according to the UART serial protocol. The default baud rate is 115200bps. The transmitted frame format is 10 bits, with 1 zero-level start bit, 8 data bits (low-order bit first), and 1 stop bit, and no parity bit. It can directly communicate with a single-chip microcomputer or other MCU powered by a 5V to 8.4V power supply. The face recognition module data transmission pin (TXD) is connected to the data receiving pin (RXD) of the host computer, and the face recognition module data receiving pin (RXD) is connected to the data transmission pin (TXD) of the host computer. After the module is powered on, it takes approximately 0.7 seconds to initialize. During this period, the module cannot respond to commands from the host computer. After completing the initialization, the module will immediately send a status notification packet to the host computer, indicating that the module is now able to operate normally and receive commands from the host computer. The main control actively controls the power on and off of the module. When the module is powered on, it begins its startup process. Once it enters standby mode, it notifies the master controller with the message "MSG NOTE READY." The master controller then begins sending command messages, which the module processes and returns. If the module does not respond within 2 seconds, the master controller calls "MSG POWERDOWN." The module processes the message and returns, and the master controller can then power off. To communicate with a host computer operating at RS-232 levels (e.g., a PC), add a level conversion circuit (e.g., a MAX232 circuit) between the module and the host computer. If the MAX232 circuit fails to receive data, verify that there are multiple Master terminals on the RXD line or other circuitry causing interference.

[0034] Taking into account the power consumption of the overall circuit, the face recognition module has only two states: working and non-working, and there is no sleep or standby state; the power input of the face recognition module is turned off normally, and the face recognition module does not work; when the face recognition module needs to be connected to the microcontroller (MCU) function, the power input of the face recognition module is provided, the face recognition module works and completes the corresponding commands. If the face recognition module is no longer used, the power supply of the face recognition module is cut off, and the face recognition module enters the non-working state.

[0035] Furthermore, in this embodiment, the main control chip 4 has a built-in recognition algorithm: the algorithm determines in real time whether there are faces and palms in the acquired image, and adjusts the camera appropriately, thereby achieving better face and palm vein recognition performance, making the switching between face recognition and palm vein recognition more appropriate.

[0036] For the face recognition function, facial data was collected in different environments and after making different postures, and a rich database was built.

[0037] Using deep neural network technology, we designed an optimized network model that can run on the terminal and has better performance.

[0038] This network model was trained on a rich and diverse database, and ultimately produced a face recognition algorithm with stronger recognition capabilities.

[0039] Thanks to its adaptive function, it can quickly and accurately identify people in various environments even when their posture changes.

[0040] For live face recognition, the algorithm simulates various real-world prosthetic attack methods and collects data on these attacks. It then trains the algorithm using the infrared reflection characteristics, image quality, and coarse and fine details of this data to accurately distinguish between live and prosthetic subjects.

[0041] For palm vein recognition, the palm vein image obtained by the camera is preprocessed to eliminate factors that have a negative impact on recognition and enhance the palm vein information as an inherent biometric feature.

[0042] The accuracy of palm vein recognition is improved by combining global and local features reflecting the geometric characteristics of palm veins and utilizing feature quantities of deep neural networks.

[0043] For palm vein liveness recognition, the algorithm conducts an in-depth comparative analysis of the inherent palm vein characteristics of living bodies and various prosthetic palms, and undergoes extensive training based on the results to be able to accurately distinguish between prosthetic palms and living palm veins.

[0044] This module uses a dual-filter IRCUT lens 2 and cooperates with the ISP function of the lock body's MCU to achieve a wide-angle, high-definition visual cat's eye function.

[0045] And considering the high height of the IRCUT lens, a separate infrared light 5-bead board is designed and integrated with the algorithm board 3.

[0046] Building on the above, a facial recognition module uses a single lens to implement face recognition, palm vein recognition, and visual cat's eye functions. The facial recognition module consists of a main control board, a five-bead board with infrared lights, and a dual-filter IRCUT lens. The module's structure is similar to that of a standalone cat's eye, utilizing an infrared / white light dual-filter IRCUT wide-angle lens. It can produce high-definition color cat's eye images even during the day, and high-definition black and white images at night, combined with a unique algorithm.

[0047] The above is a detailed description of the present invention in conjunction with specific embodiments, and the specific implementation methods of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, if a number of equivalent substitutions or obvious modifications are made without departing from the concept of the present invention and the performance or use are the same, they should be considered to fall within the scope of patent protection of the present invention as determined by the submitted claims.

Claims

1. A single-lens multifunctional recognition module, characterized by: The module includes an infrared light board, a dual-filter IRCUT lens and an algorithm board; The algorithm board is provided with a main control chip; The dual filter IRCUT lens is set on the algorithm board through the infrared light board; The main control chip is electrically connected to the dual filter IRCUT lens; The algorithm board is provided with a data sending pin and a data receiving end connected to the host computer, a data receiving pin and a data sending end connected to the host computer; The main control chip is provided with a UVC / UAC signal interface and is connected to the cat-eye board; The main control chip is provided with a control signal interface and is connected to the lock control board.

2. The single-lens multifunctional recognition module according to claim 1, characterized in that: The algorithm board is also provided with an infrared lamp, a speaker and a microphone, and the infrared lamp, the speaker and the microphone are electrically connected to the main control chip.

3. The single-lens multifunctional recognition module according to claim 1, characterized in that: The algorithm board is also provided with a storage chip.

4. The single-lens multifunctional recognition module according to claim 1, characterized in that: The algorithm board is also provided with a power supply module.

Images