CN215264825U - Binocular face recognition device based on multiband light - Google Patents
Binocular face recognition device based on multiband light Download PDFInfo
- Publication number
- CN215264825U CN215264825U CN202023166105.4U CN202023166105U CN215264825U CN 215264825 U CN215264825 U CN 215264825U CN 202023166105 U CN202023166105 U CN 202023166105U CN 215264825 U CN215264825 U CN 215264825U
- Authority
- CN
- China
- Prior art keywords
- light
- infrared
- band
- face recognition
- device based
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Studio Devices (AREA)
Abstract
The utility model provides a two mesh face recognition device based on multiband light, including the circuit board that has identification control module and installing on the circuit board and connect in control module's left camera, right camera, the infrared light filling lamp of high band and the infrared light filling lamp of low band, left camera on install the dual band light filter that supplies visible light and low band infrared light to pass through, right camera on install the infrared light filter that supplies low band infrared light and high band infrared light to pass through. The scheme solves the common problems of the conventional scheme of dazzling visible light supplementary lighting in a dark environment by adopting a multiband infrared supplementary lighting mode.
Description
Technical Field
The utility model belongs to the technical field of face detection, especially, relate to a two mesh face recognition device based on multiband light.
Background
With the wide application of face recognition in various industries, various defects of face recognition, such as various technologies, costs, processes and the like, are exposed one by one in the application process, for example, the problems of low recognition rate, high cost, low yield and the like when light is weak are solved.
In order to solve the above technical problems, people have long searched for, for example, a chinese patent discloses a face recognition method based on visible light and near infrared [ application No.: 201910833604.7], the method simultaneously acquires visible light images and near infrared images, simultaneously carries out face detection and recognition on the visible light images and the near infrared images, and finally improves the face recognition effect by combining the advantages of visible light and near infrared in the recognition process from the perspective of binocular face recognition.
Although the scheme combines visible light and near infrared, one of the two cameras in the scheme can only collect a visible light imaging picture and the other can only collect an infrared light imaging picture, and the camera is only suitable for face recognition under the condition of sufficient illumination and cannot solve the problem of non-dazzling light supplement in a dark environment.
SUMMERY OF THE UTILITY MODEL
The utility model aims at the above-mentioned problem, provide a two mesh face recognition device based on multiband light.
In order to achieve the above purpose, the utility model adopts the following technical proposal:
a binocular face recognition device based on multiband light comprises a circuit board with a recognition control module, a left camera, a right camera, a high-waveband infrared light supplement lamp and a low-waveband infrared light supplement lamp, wherein the circuit board is installed on the circuit board and connected to the recognition control module, a dual-waveband light filter for visible light and low-waveband infrared light to pass through is installed on the left camera, and an infrared light filter for low-waveband infrared light and high-waveband infrared light to pass through is installed on the right camera.
In the binocular face recognition device based on multiband light, the left camera, the right camera, the high-waveband infrared light supplement lamp and the low-waveband infrared light supplement lamp are arranged on the same plane and the same axis of the circuit board.
In the binocular face recognition device based on multiband light, the high-waveband infrared light supplement lamp and the low-waveband infrared light supplement lamp are both located between the left camera and the right camera.
In the binocular face recognition device based on multiband light, the low-band infrared light supplement lamp is located on one side close to the left camera, and the high-band infrared light supplement lamp is located on one side close to the right camera.
In the binocular face recognition device based on multiband light, the left camera and the right camera are both infrared enhanced CMOS.
In the binocular face recognition device based on multiband light, the high-band infrared fill-in lamp is used for providing an infrared fill-in lamp with a 940nm band, and the low-band infrared fill-in lamp is used for providing an infrared fill-in lamp with a 850nm band;
the infrared filter is used for passing infrared light with wave bands of at least 850nm and 940 nm; the dual-band optical filter is used for passing visible light and infrared light with at least 850nm band but not passing infrared light with 940nm band;
the left camera is used for collecting visible light imaging and infrared imaging of 850nm wave band;
the right camera is used for collecting infrared imaging of 850nm wave band and infrared imaging of 940 nm.
In the binocular face recognition device based on multi-band light, the recognition control module comprises a control module and an ambient light detection module connected with the control module.
In the binocular face recognition device based on multiband light, the recognition control module is connected with the display module and the storage module.
In the binocular face recognition device based on multiband light, the control module is further connected with a face detection module, a depth calculation module and a face recognition module.
In the binocular face recognition device based on multiband light, the recognition control module is a highly integrated SOC chip;
or, the identification control module comprises a plurality of hardware functional modules.
The utility model has the advantages that:
1. the common defects of the conventional scheme that visible light is used for supplementing light and dazzling in a dark environment are solved by adopting a multi-band infrared light supplementing mode;
2. the two cameras can acquire infrared light images, the left camera can acquire visible light images and 850nm waveband infrared light images simultaneously, the right camera can acquire 850nm waveband and 940nm waveband infrared light images simultaneously, visible light and near infrared can be combined under the condition of sufficient illumination, and non-dazzling infrared supplementary lighting can be provided for the two cameras simultaneously under the condition of insufficient illumination;
3. the double cameras and the two infrared light supplement lamps are in hardware arrangement design in the same plane and the same optical axis, and complex complexity of a production process can be reduced.
Drawings
Fig. 1 is a schematic view of the structure of the binocular face recognition device based on multiband light according to the present invention;
fig. 2 is a structural block diagram of the binocular face recognition device based on multiband light of the present invention;
fig. 3 is a flow chart of the utility model of the binocular face recognition device based on multiband light.
Reference numerals: identifying the control module 1; an ambient light detection module 11; a face detection module 12; a depth calculation module 13; a face recognition module 14; a control module 15; a circuit board 2; a left camera 3; a dual band filter 31; a right camera 4; an infrared filter 41; a high-band infrared light supplement lamp 5; a low-band infrared light supplement lamp 6; a display module 7; a memory module 8.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
As shown in fig. 1, the present embodiment discloses a binocular face recognition device based on multiband light, which includes a circuit board 2 having a recognition control module 1, and a left camera 3, a right camera 4, a high-band infrared fill-in lamp 5 and a low-band infrared fill-in lamp 6 mounted on the circuit board 2 and connected to the recognition control module 1, where the low-band infrared fill-in lamp 6 is used to provide infrared light of relatively lower band, for example, infrared fill-in light of 850nm band, and certainly, also can be used for infrared fill-in light of other bands, and when visible light is insufficient, the left camera 3 and the right camera 4 are photographed and filled with light at the same time; the high-band infrared fill-in light 5 is used to provide infrared light of a relatively high band, for example, infrared fill-in light of a 940nm band, and certainly, infrared fill-in light of other bands may be used, and when visible light is sufficient, only the right camera 4 is filled in light.
The left camera 3 is internally or above provided with a dual-band optical filter 31 which only allows infrared light with a wavelength below the infrared band provided by the visible light and the high-band infrared fill-in lamp 5 to pass through, specifically, in this embodiment, infrared light with a wavelength below 940nm to pass through (preferably, infrared light with a wavelength below 850nm to pass through), and the right camera 4 is internally or above provided with an infrared optical filter 41 which simultaneously allows infrared light with a low wavelength and infrared light with a high wavelength to pass through, at least infrared light with a wavelength of 850nm and 940nm to pass through, or an optical filter which allows infrared light with a wavelength of 850 and 940nm to pass through, or an optical filter which only allows infrared light with a single wavelength of 850nm and 940nm to pass through.
The left camera 3 is used for collecting left eye images of visible light imaging and 850nm waveband infrared light imaging, and 850nm waveband infrared light enhancement CMOS can be optimized.
The right camera 4 is used for collecting a right-eye image formed by infrared light of 850-940 wave bands, and infrared enhanced CMOS of 850nm and 940nm wave bands can be optimized.
Two cameras homoenergetic of this embodiment can gather the infrared light formation of image, and left camera 3 can gather visible light formation of image and 850nm wave band infrared light formation of image simultaneously, and the infrared light formation of image of 850nm wave band and 940nm wave band can be gathered simultaneously to right camera 4, can combine visible light and near-infrared under the sufficient condition of illumination, can provide the infrared light filling of not dazzling for two cameras simultaneously under the not enough condition of illumination.
Specifically, as shown in fig. 2, the recognition control module 1 is connected to the display module 7 and the storage module 8, and the recognition control module 1 includes a control module 15, and an ambient light detection module 11, a face detection module 12, a depth calculation module 13, and a face recognition module 14 connected to the control module 15, where the modules may be a high-integration SOC chip or a combination module formed by combining a plurality of hardware functional modules.
The display module 7 is preferably a color display screen for displaying the left eye image and the recognition result, including the display prompt of the success and failure of face recognition. Under the condition of sufficient visible light, the color image is directly displayed, and when the visible light is insufficient, the 850nm waveband infrared light supplement lamp is turned on to display the gray image, so that the embarrassment of pure black screen display under a dark environment can be avoided. The color display is performed under the condition that visible light is enough, and the black-and-white gray scale display is performed under the dark environment that the visible light is not enough. Can realize splendid daytime through two cameras, the clear effect evening has can enough guarantee higher formation of image quality, has again with low costs, advantages such as product size is little.
The storage module 8 stores registered face feature template data; the ambient light detection module 11 may detect the ambient visible light brightness by using a light brightness detection circuit, or may detect the ambient visible light brightness of the left eye image, and send the detection result to the control module 15.
Further, the high-band infrared light supplement lamp 5 and the low-band infrared light supplement lamp 6 are both located between the left camera 3 and the right camera 4, the low-band infrared light supplement lamp 6 is located on one side close to the left camera 3, and the high-band infrared light supplement lamp 5 is located on one side close to the right camera 4. Of course, the positions of the high-band infrared fill-in light 5 and the low-band infrared fill-in light 6 and the positions of the left camera 3 and the right camera 4 can be interchanged according to specific situations, and are not limited herein.
Preferably, the left camera 3, the right camera 4, the high-band infrared light supplement lamp 5 and the low-band infrared light supplement lamp 6 are mounted on the same plane and the same axis of the circuit board 2, so that the process complexity can be reduced, and the mounting technology is preferably adopted for mounting.
Further, as shown in fig. 3, a method for using the binocular face recognition device based on multi-band light when being put into use is as follows:
s1, a left camera 3 collects images to collect left eye images and transmits the images to an ambient light detection module 11, and the images are output to a display module 7 to be directly displayed;
s2, the ambient light detection module 11 detects ambient visible light brightness according to the left eye image, if the brightness exceeds a brightness threshold value, the visible light is sufficiently bright, the high-band infrared light supplement lamp 5 is turned on (namely, the 940nm infrared light supplement lamp is turned on) to supplement light for the right camera 4, and otherwise, the low-band infrared light supplement lamp 6 is turned on (namely, the 850nm infrared light supplement lamp is turned on) to supplement light for the left camera 3 and the right camera 4 at the same time; the brightness threshold is set according to specific situations, and is not limited herein;
s31, the left camera 3 and the right camera 4 respectively collect images so as to respectively collect a left eye image and a right eye image, transmit the left eye image and the right eye image to the face detection module 12, and simultaneously transmit the images collected by the left camera 3 to the display module 7 in real time for dynamic display;
s32, the face detection module 12 adopts a depth learning algorithm to respectively perform face detection on the left eye image and the right eye image, and if one of the faces of the left eye image and the right eye image is not detected, the process returns to S31 to perform image acquisition again; if the face is detected in both the left eye image and the right eye image, all the infrared light supplement lamps are turned off, the power consumption is reduced, and meanwhile, the face key point coordinate information of the left eye image and the face key point coordinate information of the right eye image are respectively calibrated and transmitted to the depth calculation module 13;
s33, the depth calculation module 13 carries out three-dimensional depth calculation on the coordinates of the key points of the human faces in the left eye image and the right eye image, the three-dimensional depth calculation is used for judging whether the images are human face photos in an anti-counterfeiting mode, if the images are depth images, the next step of human face recognition is carried out, if the images are not depth images, the display module 7 displays that the human face recognition fails, and the process is ended;
and S34, the face recognition module 14 compares the left eye image and the right eye image with the face feature template data in the storage module 8 respectively to realize face recognition.
In this embodiment, preferably, if there is a pair of images in the left eye image and the right eye image, the result of successful recognition is output and the process is ended, otherwise, the result of failed recognition is output and the process is ended.
Of course, the successful recognition result may also be output when the comparison results of the left eye image and the right eye image with the face feature template data both pass.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications, additions and substitutions for the specific embodiments described herein may be made by those skilled in the art without departing from the spirit of the invention or exceeding the scope of the invention as defined in the accompanying claims.
Although the identification control module 1 is used more herein; an ambient light detection module 11; a face detection module 12; a depth calculation module 13; a face recognition module 14; a control module 15; a circuit board 2; a left camera 3; a dual band filter 31; a right camera 4; an infrared filter 41; a high-band infrared light supplement lamp 5; a low-band infrared light supplement lamp 6; a display module 7; memory module 8, etc., without excluding the possibility of using other terms. These terms are used merely to more conveniently describe and explain the nature of the present invention; they are to be construed in a manner that is inconsistent with the spirit of the invention.
Claims (10)
1. The binocular face recognition device based on multiband light is characterized by comprising a circuit board (2) with a recognition control module (1) and a circuit board (2) installed on the circuit board (2) and connected to a left camera (3), a right camera (4), a high-waveband infrared light supplement lamp (5) and a low-waveband infrared light supplement lamp (6) of the recognition control module (1), wherein a double-waveband light filter (31) only allowing visible light and low-waveband infrared light to pass through is installed on the left camera (3), and an infrared light filter (41) only allowing low-waveband infrared light and high-waveband infrared light to pass through is installed on the right camera (4).
2. The binocular face recognition device based on multiband light, according to claim 1, wherein the left camera (3), the right camera (4), the high-band infrared supplementary lighting lamp (5) and the low-band infrared supplementary lighting lamp (6) are mounted on the same plane and the same axis of the circuit board (2).
3. The binocular face recognition device based on multiband light, according to claim 2, wherein the high band infrared fill light (5) and the low band infrared fill light (6) are both located between the left camera (3) and the right camera (4).
4. The binocular face recognition device based on multiband light, according to claim 3, wherein the low band infrared fill-in light (6) is located at a side close to the left camera (3), and the high band infrared fill-in light (5) is located at a side close to the right camera (4).
5. The binocular face recognition device based on multiband light, according to claim 1, wherein the left camera (3) and the right camera (4) are both infrared enhanced CMOS.
6. The binocular face recognition device based on multiband light according to claim 1, wherein the high-band infrared fill-in lamp (5) is used for providing 940nm band infrared fill-in light, and the low-band infrared fill-in lamp (6) is used for providing 850nm band infrared fill-in light;
the infrared filter (41) is used for passing infrared light with wave bands of at least 850nm and 940 nm; the dual-band optical filter (31) is used for passing visible light and infrared light with at least 850nm band, and does not pass infrared light with 940nm band;
the left camera (3) is used for collecting visible light imaging and infrared imaging of 850nm wave band;
the right camera (4) is used for collecting infrared imaging of 850nm wave band and infrared imaging of 940nm wave band.
7. Binocular face recognition device based on multiband light according to claim 1, characterized in that the recognition control module (1) comprises a control module (15) and an ambient light detection module (11) connected to the control module (15).
8. Binocular face recognition device based on multiband light according to claim 7, characterized in that the recognition control module (1) is connected with a display module (7) and a storage module (8).
9. Binocular face recognition device based on multiband light according to claim 8, characterized in that the control module (15) is further connected with a face detection module (12), a depth calculation module (13) and a face recognition module (14).
10. The binocular face recognition device based on multiband light of claim 9, wherein the recognition control module (1) is a highly integrated SOC chip;
or, the identification control module (1) comprises a plurality of hardware functional modules.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202023166105.4U CN215264825U (en) | 2020-12-24 | 2020-12-24 | Binocular face recognition device based on multiband light |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202023166105.4U CN215264825U (en) | 2020-12-24 | 2020-12-24 | Binocular face recognition device based on multiband light |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN215264825U true CN215264825U (en) | 2021-12-21 |
Family
ID=79487974
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202023166105.4U Active CN215264825U (en) | 2020-12-24 | 2020-12-24 | Binocular face recognition device based on multiband light |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN215264825U (en) |
-
2020
- 2020-12-24 CN CN202023166105.4U patent/CN215264825U/en active Active
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN201927050U (en) | Face recognition device with infrared human body sensing function | |
| CN201788518U (en) | Identification device with facial image and iris image acquisition functions | |
| CN106385530B (en) | Double-spectrum camera | |
| CN104597622B (en) | Anti-dazzling glasses and method | |
| CN107580163A (en) | A kind of twin-lens black light camera | |
| CN207283655U (en) | A kind of twin-lens black light camera | |
| CN107820066A (en) | A kind of low-luminance color video camera | |
| CN110738952A (en) | HUD display method and system combining human eye recognition, tracking and ray recognition | |
| CN208956216U (en) | Based on the long large car viewing system away from transmission of low-light (level) camera synchronous acquisition | |
| CN103802727A (en) | Low-visibility guiding system | |
| CN215264825U (en) | Binocular face recognition device based on multiband light | |
| CN106934349B (en) | Dual-camera imaging and iris acquisition and recognition integrated equipment | |
| CN112528949B (en) | Binocular face recognition method and system based on multi-band light | |
| CN103945093B (en) | Based on the visible near-infrared integral camera device of ARM platform recognitions of face and method | |
| CN210469476U (en) | Visible light and infrared light bi-pass vehicle-mounted face recognition camera | |
| CN105898122B (en) | A kind of photoelectricity sensory perceptual system | |
| CN112133246A (en) | Control method of LED display screen system and LED display screen system | |
| CN208273078U (en) | A kind of lll night vision imager | |
| CN209170491U (en) | Vision facilities and image procossing mould group | |
| CN110751919B (en) | Transparent display system and method of operating the same | |
| CN204168399U (en) | A kind of colour blindness colour vision servicing unit | |
| CN110211528A (en) | A kind of system that LED display display brightness is adjusted | |
| US20220118909A1 (en) | Driving assistance system and method for driver | |
| CN2508306Y (en) | Vital iris picture collecting device | |
| CN212433787U (en) | Double-reflection light path electronic passport collecting instrument |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |