A kind of system that is used to discern people's faceTechnical field
The utility model relates to the biometrics identification technology field, is specifically related to a kind of system that is used to discern people's face.
Background technology
Biometrics identification technology is the complex art that compiles sign anthropology, statistical mathematics, computer utility three big subjects; Because people's face is a 3 d elastic body; The face image of being gathered can have bigger variation because of factors such as age of people, expression, acquisition angles, collection ambient lightings, so the research difficulty of face recognition technology is very big, and anthropology is in the breakthrough conclusion of lifelong unchangeability that does not obtain similar fingerprint texture aspect people's the face feature as yet and uniqueness; Domestic present because of input in this respect limited; So the achievement in research that has obtained is also very limited, really can accomplish the technology that reaches application request, at present also seldom.
The utility model content
The purpose of the utility model has been to provide a kind of system that is used to discern people's face, has solved the problem that does not also have face identification system simple in structure, that function is practical in the prior art.
A kind of system that is used to discern people's face of the utility model; Comprise digital signal processor; The input end of said digital signal processor links to each other with video acquisition module, video module, audio-frequency module; The output terminal of said digital signal processor links to each other with dynamic storage, Ethernet, flash memory, Dual serial ports, LED LCDs, universal input and output port and Wei root interface, in the said video acquisition module infrared narrow band filter is installed.
Preferably, said digital signal processor is a TI DM642 digital signal processor.
Preferably; Said video acquisition module comprises two video acquisition module with two corresponding connections of video inputs mouth; The optical filter that the light that only allows the 850uM wavelength passes through is installed between the camera lens of the camera in one of them video acquisition module and the CMOS chip, 72 LED infrared lamps that can send the 850uM wavelength are housed around the camera lens.
Preferably, said video module is connected with video-out port, and said audio-frequency module is connected with audio output port with audio input port.
Preferably, said Dual serial ports is connected with two 232 interfaces, and one of them 232 interface also is connected with IC-card.
Preferably, said universal input and output port is connected with the key plate.
Beneficial effect: the system that is used to discern people's face of the utility model,, form images down in conjunction with dual camera as digital signal processor with TI DM642; Can accurately discern facial image fast, be used for human face identification work-attendance checking, gate inhibition's equipment, simple in structure; Reduced the volume of system greatly; Function effectiveness is reliable and stable, with the recognition of face device miniaturization, is easy to be embedded in other various systems; Thereby reduced the application threshold of recognition of face, thereby made applying of face recognition technology become possibility.
Description of drawings
Fig. 1 is used to discern the framed structure synoptic diagram of the system of people's face for the utility model a kind of.
Embodiment
Embodiment 1
In conjunction with Fig. 1; A kind of system that is used to discern people's face of the utility model; Comprise TI DM642 digital signal processor; Said TI DM642 digital signal processor is connected with and first video acquisition module, second video acquisition module, video module, audio-frequency module, dynamic storage, Ethernet, flash memory, Dual serial ports, LED LCDs, universal input and output port and Wei root interface.
Said first video acquisition module is connected with the first video inputs mouth; Second video module is connected with the second video inputs mouth; Combining under the dual camera imaging; Embody more accurately recognition of face sharpness, rock or moving process in can be more precise and stable demonstrate image, thereby recognition image fast.Between the camera lens camera lens of the camera in the second video inputs mouth and the CMOS chip infrared narrow band filter is installed; The 850uM near infrared light that the bandwidth of infrared narrow band filter only allows infraluminescence LED pipe group to be sent passes through; 36groups 72 the LED infrared lamps that can send the 850uM wavelength are housed around the camera lens; Stablize the surrounding environment light source, no matter in the sunlight direct projection still under pitch-dark dark surrounds, can be exactly, identification promptly.
Said video module is connected with video-out port, and said audio-frequency module is connected with audio output port with audio input port.
Said Dual serial ports is connected with two 232 interfaces, and one of them 232 interface also is connected with IC-card.
DSP has very high cost performance as a kind of high performance digital signal processor spare.Recognition of face have high computational complexity, and the high-performance of DSP and design feature shows that DSP might become the processing platform of recognition of face as a kind of computing intensive applications.
The utility model is combining under the dual camera imaging, embody more accurately recognition of face sharpness, rock or moving process in can be more precise and stable demonstrate image, thereby recognition image fast.16 touch key-presss such as the system of the utility model is provided with wireless transmission, 72LED lamp, resets, power input, debug port in conjunction with the programming chip, can make keeper's override equipment more flexibly sooner.
Universal input/output interface, network interface, input/output control module links together with said digital signal processor unit, is used to control the mode of said digital signal processor unit being sent by data and transmission speed etc.
The recognition system flow process is: video, digital picture, scanner picture to input carry out the detection of people's face and follow the tracks of; Extract facial characteristics, generate the faceform, with all known person face Model Matching of stock; Carry out the similarity ordering, the identity information of people's face is discerned in the input picture.
The foregoing description is technical conceive and the characteristics for the utility model is described just, its objective is to be the content that lets the one of ordinary skilled in the art can understand the utility model and to implement according to this, can not limit the protection domain of the utility model with this.The variation or the modification of every equivalence of having done according to the essence of the utility model content all should be encompassed in the protection domain of the utility model.