Remote near-infrared hand back vein image acquisition device and acquisition method thereof

Abstract

The invention relates to a remote near-infrared hand vein image acquisition device and an acquisition method thereof. The system is mainly composed of a high-definition visible light camera, a near-infrared camera, a controllable transparency liquid crystal screen, a near-infrared light source holder and a central control terminal. And the central control terminal is responsible for coordination and data processing of each component. According to the invention, a high-definition visible light camera and a near-infrared camera are combined, and a binocular vision system is utilized to realize accurate positioning and distance estimation of a hand back area. Based on the parameters, the system can intelligently adjust the irradiation mode of the active near-infrared light source so as to realize accurate irradiation of the hand back area. Afterwards, a near-infrared camera collects hand vein images, and the images are input into the focusing network provided by the invention for processing, so that the sharpness of the hand vein images is realized, and subsequent identity recognition and living body authentication are facilitated. Through the steps, the hand back vein image is efficiently and accurately collected and recognized within the long-distance range of 3-10 meters, and the problems of defocus blurring and glare interference in the long-distance vein recognition process are successfully solved. The system not only has the convenience similar to that of face recognition, but also has the natural anti-counterfeiting characteristic of vein recognition, and is suitable for various remote identity verification scenes.

Description

Technical Field

[0001] The invention relates to a hand back vein image acquisition technology, and in particular to a remote near-infrared hand back vein image acquisition device and an acquisition method thereof. Background Art

[0002] Existing hand vein recognition systems mainly rely on close-range acquisition technology, in which cameras with fixed depth of field are used to capture images. Users need to ensure that the back of the hand is placed within the effective depth of field of the camera in order to successfully acquire a clear vein image. Since the depth of field is large at a long distance, while the depth of field of a general near-infrared camera is short, it is difficult to respond immediately even if a zoom camera is used, so it is necessary to perform software focusing and clear processing on the acquired remote defocused hand vein image. During the hand vein recognition process, the distance from the back of the hand to the camera varies greatly, while the depth of field of a general camera is short and is mostly fixed-focus. Even for a zoom camera, it is necessary to accurately know the specific distance from the back of the hand to the camera in order to perform a focusing operation, so as to capture a clear image that meets the requirements of vein recognition. As for lighting conditions, current hand vein recognition systems mostly use uniform lighting technology. Although this lighting method can ensure that the entire acquisition area has sufficient brightness, it does not specifically optimize the lighting for the hand contour area. Optimizing lighting conditions, especially adaptive lighting for the hand contour area, can help improve image quality, thereby improving the accuracy and efficiency of vein recognition. In terms of distance measurement, currently the commonly used methods of hand back vein recognition system are binocular distance measurement and ultrasonic distance measurement. These methods can effectively obtain the distance information from the back of the hand to the camera, and provide necessary data support for the camera's focus and lighting adjustment. By combining these distance measurement technologies, the system can adjust the camera parameters more intelligently and achieve more accurate and efficient hand back vein recognition.

[0003] Hand vein recognition technology has developed feature extraction and matching recognition methods under various postures. Traditional methods mainly include geometric feature representation methods, binary vein line feature representation methods, skeleton feature representation methods, and vein line extraction and segmentation methods. Deep learning methods cover technologies based on deep convolutional neural networks (CNN), recurrent neural networks (RNN) and their variants, and autoencoders. In order to further improve the convenience and accuracy of recognition, we cleverly integrate key components such as high-definition cameras, near-infrared cameras, and near-infrared light sources on the gimbal to ensure that the device can efficiently track the user within 10 meters.

[0004] Traditional image fusion techniques usually rely on pre-trained neural networks, which require large-scale datasets and extensive training. In contrast, the "Deep Image Prior" focusing algorithm developed by Dmitry Ulyanov et al. performs image restoration tasks by exploiting the inherent structure of randomly initialized neural networks without prior training on datasets.

[0005] In order to avoid interference from visible light, traditional palm vein recognition requires users to put their palms into the image acquisition box. In order to improve the accuracy of recognition, a hand fixing bolt is used for contact shooting, which imposes great constraints on users. Non-contact acquisition in an open environment is more popular with users. However, the open environment brings interference of visible light on near-infrared light imaging. The opacity, unevenness and heterogeneity of the skin tissue covering the palm veins cause the near-infrared light irradiating the palm veins to scatter. The visible light in the open environment aggravates the scattering and increases the noise, resulting in unclear palm vein imaging for some people. The non-contact acquisition method increases the intra-class differences of the same sample due to rotation, translation, scaling and lighting changes in multiple shots. Limited by distance and lighting conditions, the quality of the acquired images is unstable. Traditional acquisition methods are easily interfered by ambient light, and reflection and glare problems often appear in the image, thus affecting the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale. Figure 1 is a flow chart of the method of the present invention; Figure 2 This is a diagram of a remote near-infrared hand back vein image acquisition device of the present invention; Figure 3 A diagram of a binocular camera imaging method of the present invention; Figure 4 It is a control module diagram of the present invention. Summary of the invention

[0006] In view of the defects in the prior art, the purpose of the present invention is to provide a remote near-infrared hand back vein image acquisition device and its acquisition method, so as to solve the problems of unstable image quality and susceptibility to ambient light interference in the prior art, and realize efficient and accurate hand back vein recognition.

[0007] In order to achieve the above-mentioned purpose, the present invention provides a remote near-infrared back of hand vein image acquisition device, comprising a near-infrared light source, a controllable transparency liquid crystal screen, a high-definition camera and a near-infrared camera. The near-infrared light source is provided with one and is distributed in front of the human body, forming a structure in which the near-infrared emitted by the near-infrared light source irradiates in the direction of the front. The controllable transparency liquid crystal screen is arranged at the transmitting end of the near-infrared light source assembly, and is used to control the light intensity distribution of the near-infrared light source. The high-definition camera and the near-infrared camera are respectively deployed on the pan-tilt platform, and the controller is electrically connected to the high-definition camera, the near-infrared camera, the near-infrared light source and the controllable transparency liquid crystal screen respectively. Through the coordinated work of the high-definition camera and the near-infrared camera, the binocular vision system can perform accurate distance measurement of the back of the hand area.

[0008] The present invention also provides a remote near-infrared back hand vein image acquisition method, which is applied to the above-mentioned device, and the transmittance of the controllable transparency liquid crystal screen is set as a variable through the controller, and the near-infrared light source is irradiated, and the back hand vein image is collected by the near-infrared camera and transmitted to the controller, and finally a set of vein images is obtained. The back hand area is detected by a high-definition camera, the optical axis is adjusted to align with the center of the back hand, and multiple images with different focal lengths are taken by the near-infrared camera, and full-focus fusion is performed to generate a clear vein image. The present invention introduces an image fusion system using a focusing algorithm to achieve better quality images. The technology uses a convolutional neural network (ConvNet) with randomly initialized weights, and then optimizes specific degraded images. This process uses the network structure as a hand-made prior to achieve effective image denoising, super-resolution and restoration. The fused vein image is processed to obtain a final vein image with uniform grayscale and no bright spots.

[0009] A remote near-infrared hand back vein image acquisition device, characterized in that it includes a near-infrared light source, a controllable transparency liquid crystal screen, a high-definition camera and a near-infrared camera; the near-infrared light source is provided and distributed in front of the human body, thereby forming a structure in which the near-infrared light source emitted by the near-infrared light source irradiates along the direction in front; the controllable transparency liquid crystal screen is provided at the transmitting end of the near-infrared light source component to control the light intensity distribution of the near-infrared light source; the high-definition camera and the near-infrared camera are respectively deployed on a pan-tilt platform, and the controller is electrically connected to the high-definition camera, binocular camera and near-infrared camera, the near-infrared light source component and the controllable liquid crystal screen component respectively.

[0010] Preferably, the infrared and near-infrared light source includes an infrared light source emitter and a first circuit board for driving and controlling the infrared light source emitter, and the emission port of the infrared light source emitter is circular. The light intensity of the infrared light source emitter is controlled by the first circuit board to achieve adjustability of the light intensity, which is conducive to adjusting the grayscale value of the vein image.

[0011] Preferably, the image emitted by the infrared laser emitter is circular. The projection width of the near-infrared beam is 2 mm, the wavelength is 850 nm, and the spectrum width is 5 nm. The infrared light source is completely irradiated in front of the back of the hand, which can reduce overexposure and glare even for metal objects on the wrist, thereby improving the collection success rate.

[0012] Preferably, the controllable transparency LCD screen includes a device for controlling the illumination intensity distribution of the near-infrared beam area on the back of the hand. Since the human body area is large, the controllable transparency LCD screen independently controls the transmittance of the back of the hand and other non-hand parts of the human body, improves the self-adaptation of light intensity, and is beneficial to the distribution of light intensity.

[0013] Preferably, the illumination of the area of ​​interest on the back of the hand is adjusted by an adaptively controllable transparency LCD screen and an off-mirror light source to prevent glare and overexposure caused by ambient light and reflective objects (such as watches and rings).

[0014] Preferably, the controllable transparency liquid crystal screen covers the emission port of the near-infrared light source emitter, so that the emitted near-infrared light source can be controlled, which is conducive to adjusting the gray value of the vein image.

[0015] Preferably, the remote distance measurement is performed by combining a visible light camera and a near infrared camera, which can be understood as an extended form of a binocular camera system. In the system, a high-definition visible light camera and a near infrared camera work together to perform back of hand area detection and image acquisition. Using a binocular camera for remote distance measurement ensures that the optical axis is aligned with the center of the back of the hand.

[0016] Preferably, the pan-tilt platform will use image processing and control algorithms to enable the pan-tilt platform to automatically align and track the back of the hand and collect vein images in a timely manner. First, the back of the hand is detected, and the target is detected in the image through the image processing algorithm. Then the back of the hand is tracked, and when the back of the hand moves in the image, the position of the target is tracked by the target tracking algorithm. Then the control algorithm is used to calculate the rotation angle of the pan-tilt platform according to the position of the back of the hand in the image. Finally, the pan-tilt platform is driven, and the pan-tilt platform is rotated by the driving motor according to the calculated rotation angle, thereby realizing dynamic tracking of the back of the hand.

[0017] Preferably, the grayscale value of one or more segments of the vein image that do not meet the determination requirements is adjusted within a set grayscale value range by changing the transmittance of the controllable liquid crystal screen and the illumination intensity of the infrared light source. Increasing the grayscale value adjustment method of the vein image increases the adjustability of the grayscale value, which is conducive to improving the quality of the vein image.

[0018] Preferably, in the acquired out-of-focus image, the focusing algorithm utilizes a randomly initialized convolutional generative network (ConvNet) and fits it to a single degraded image to perform the restoration task. The network structure inherently captures low-level image statistics, which is critical for solving inverse problems such as denoising, super-resolution, and restoration. This step solves the image blurring problem caused by the limited depth of field of the lens during long-distance acquisition.

[0019] Beneficial effects of the present invention: 1. Improve acquisition accuracy: Through the collaborative work of near-infrared light source, high-definition camera and near-infrared camera, and the application of binocular vision system, the present invention can achieve high-precision hand back vein image acquisition at a distance far from the user. This precise long-range ranging capability significantly improves the acquisition quality of vein images and provides reliable data support for subsequent feature extraction and matching.

[0020] 2. Improve image quality: The present invention adjusts the light intensity distribution of the near-infrared light source through a controllable transparency LCD screen, thereby achieving effective control of the grayscale value of the vein image. This technology can reduce overexposure and glare in the image, improve the contrast of the vein image, make the image clearer and without bright spots, and thus improve the accuracy of the recognition system.

[0021] 3. Automatic tracking and adjustment: The pan / tilt system in the device can automatically align and track the back of the hand through image processing and control algorithms, ensuring that vein images can be stably captured when the user moves. This automatic dynamic tracking function avoids the tedious manual adjustment and improves the convenience of the device.

[0022] 4. Flexible adjustment of lighting conditions: By adjusting the transmittance of the controllable transparency LCD screen and the illumination intensity of the near-infrared light source, the present invention can adapt to the lighting conditions in different environments and optimize the distribution of light intensity. This feature enables the device to maintain stable image quality in various environments and enhances the adaptability of the system.

[0023] 5. Efficient data processing: The present invention uses the deep learning algorithm and focusing model preset by the full focus fusion technology to process the defocused image and generate a clear focused image. This step solves the image blur problem caused by the depth of field limitation of the lens during long-distance acquisition.

[0024] 6. Contactless identification: Extract vein features from the processed focused image and use advanced recognition algorithms for vein matching and identity recognition. The system provides basic support for remote access control and payment systems, has the convenience of not requiring users to cooperate with specific actions, and maintains the high security of vein recognition.

[0025] Through the above beneficial effects, the present invention solves the problems of unstable image quality and susceptibility to ambient light interference in the prior art, realizes efficient and accurate back of hand vein recognition, and provides reliable technical support for related fields. DETAILED DESCRIPTION

[0026] The following embodiments of the technical solution of the present invention are described in detail. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0027] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0028] like Figure 1As shown, in this embodiment, direction a indicates the top, direction b indicates the bottom, direction c indicates the left, direction d indicates the right, direction e indicates the front, and direction f indicates the back. The remote near-infrared hand back vein image acquisition device includes 1-near-infrared light source, 2-controllable transparency liquid crystal screen, 3-high-definition camera, 4-near-infrared camera, 5-pan platform, the near-infrared light source 1 includes an infrared light source transmitter and a first circuit board for driving and controlling the infrared light source transmitter, the infrared light source transmitter power is 5w, the diameter of the transmitting end of the infrared light source transmitter is 12mm, its transmitting port is circular, the wavelength of the emitted near-infrared light source is 850nm, and the spectrum width is 5nm. The light intensity of the near-infrared light source transmitter is controlled by using controllable transparency liquid crystal to achieve the adjustability of the light intensity, which is conducive to adjusting the grayscale value of the vein image. The upper end surface of the pan-tilt 5 is provided with a high-definition camera 3 for identifying the back of the hand area, and an infrared camera 4 for collecting back of the hand veins is provided next to the high-definition camera 3. The infrared light source transmitter 1 is arranged above the pan-tilt 5 and a controllable liquid crystal screen 2 is provided in front. The infrared light source 1 is arranged directly behind the controllable liquid crystal screen 2, thereby forming a structure in which the infrared laser irradiates along the front and rear directions of the light transmission hole. The light beam emitted by the near-infrared light source emitter is circular, so that the near-infrared light source is completely irradiated in front of the back of the hand, and even for metal objects worn on the hand, the glare can be reduced, and the collection success rate can be improved. The controllable transparency liquid crystal screen is arranged at the transmitting end of the infrared light source assembly 1 to control the infrared light intensity distribution; the infrared light beam forms a near-infrared area after being transmitted through the controllable liquid crystal screen. Since the hand is wearing a metal object, the metal object will produce a glare exposure reaction, so the light beam segment transmitting the wrist part requires a weaker light intensity, while the light beam segment transmitting the back of the hand requires a stronger light intensity. The controllable liquid crystal screen controls the transmittance of the wrist part and the back of the hand, improves the self-adaptation of the light intensity, and is conducive to the distribution of the light intensity. The controllable liquid crystal screen includes a controllable liquid crystal screen and a second circuit board for driving and controlling the controllable liquid crystal screen. The transmittance of the controllable liquid crystal screen is controlled by the first circuit board to achieve the adjustability of the transmittance, which is conducive to adjusting the grayscale value of the vein image. The controllable LCD screen covers the emission port of the infrared light source emitter, so that the emitted infrared light source can be controlled, which is conducive to adjusting the grayscale value of the vein image. The controller is electrically connected to the high-definition camera and the near-infrared camera, the infrared light source and the controllable LCD screen respectively.

[0029] In order to prevent glare during the acquisition process, partial projection is achieved by using controllable liquid crystal. Liquid crystal has voltage-controllable transmission characteristics, and the transmittance can be controlled by adjusting the applied voltage, thereby achieving dorsal venography imaging.

[0030] In order to process the out-of-focus image, the image fusion method of the focusing algorithm is used to randomly initialize the weights of the convolutional neural network, and then the network is fitted to the degraded image by optimizing the network weights. Finally, the network structure is used as a priori to perform image restoration tasks, including denoising, super-resolution and inpainting.

[0031] Through the following steps, the hand vein image can be effectively collected to provide data support for subsequent applications such as identity recognition.

[0032] In the first step, when the user enters the collection range, the system starts preparing for collection.

[0033] The second step is to use a high-definition visible light camera to complete the detection of the back of the hand area.

[0034] The third step is to determine whether the optical axis is aligned with the center of the back of the hand. If the optical axis is not aligned, rotate the gimbal to adjust the position and return to the first step to prepare for acquisition again.

[0035] The fourth step is to use a binocular vision system, with a visible light camera and a near-infrared camera to complete the remote distance measurement of the back of the hand. By comparing the image data captured by the two cameras, the depth information of the back of the hand is calculated.

[0036] The fifth step is to irradiate the back of the hand area with near-infrared rays.

[0037] Step 6. Configure the remote connection function, control the PTZ device to track the back of the hand, adjust the horizontal angle, and control the transparency of the LCD screen to perform regional shading and light transmission to prevent glare.

[0038] Step 7: Determine whether the light source completely covers the outline of the back of the hand. If not, adjust the light source and return to step 6; if covered, proceed to step 8.

[0039] Step 8: Use a near-infrared camera to take multiple images with different focal lengths.

[0040] In the ninth step, the multiple near-infrared images with different focal lengths collected in the eighth step are fully focused and fused to generate a clear vein image.

[0041] The tenth step is to process the fused vein image to obtain a final vein image with uniform grayscale and no bright spots.

[0042] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A remote near-infrared hand back vein image acquisition device, characterized in that The invention comprises the following components: a high-definition visible light camera, a near-infrared camera, a controllable transparency liquid crystal screen, a near-infrared light source platform and a central control terminal. The near-infrared light source is provided as one and is located directly in front of the human body to form a structure in which the near-infrared light source irradiates in the direction directly in front. The controllable transparency liquid crystal screen is located at the transmitting end of the near-infrared light source component and is used to control the light intensity distribution of the near-infrared light source. In addition, the high-definition camera and the near-infrared camera are respectively mounted on the platform, and the controller coordinates with the high-definition camera, the near-infrared camera, the near-infrared light source and the controllable transparency liquid crystal screen through electrical connection.

2. The remote near-infrared hand back vein image acquisition device according to claim 1, characterized in that: The visible light camera and the near infrared camera work together to perform accurate distance measurement of the back of the hand area through a binocular vision system.

3. The remote near-infrared hand back vein image acquisition device according to claim 1 or 2, characterized in that: The near-infrared light source includes a near-infrared light source emitter and a first circuit board for driving and controlling the near-infrared light source emitter.

4. The remote near-infrared hand back vein image acquisition device according to claim 3 is characterized in that: The near-infrared light emitted by the near-infrared light source emitter is dispersed in a specific shape to form a uniform light beam. The projection width of the light beam is 2 mm. The wavelength of the near-infrared light source is 850 nm, and the spectrum width is 5 nm.

5. The remote near-infrared hand back vein image acquisition device according to claim 4, characterized in that: The controllable transparency liquid crystal screen covers the emission port of the near-infrared light source emitter.

6. The remote near-infrared hand back vein image acquisition device according to claim 1, characterized in that: The controller sets the transmittance of the controllable liquid crystal screen as a variable.

7. The remote near-infrared hand back vein image acquisition device according to claim 6, characterized in that: By changing the transmittance of the controllable transparency liquid crystal screen and the illumination intensity of the near-infrared light source, the near-infrared camera can collect a sufficiently clear back of hand vein image.

8. A remote near-infrared hand back vein image acquisition method, characterized in that: The remote near-infrared hand back vein image acquisition device is applied to any one of claims 1 to 8, wherein the transmittance of the controllable transparency LCD screen is set as a variable through the controller, and the hand back vein image is acquired by a near-infrared camera and transmitted to the controller, thereby finally obtaining a set of vein images.

9. The method according to claim 8, characterized in that The back of the hand area is detected by a high-definition camera, the optical axis is adjusted to align with the center of the back of the hand, and multiple images with different focal lengths are taken by a near-infrared camera. The focusing network designed in this paper is used to focus and clarify the remote defocused back of the hand vein image.

Images