CN201682580U

CN201682580U - Visual talkback system capable of recognizing human faces

Info

Publication number: CN201682580U
Application number: CN2010201827235U
Authority: CN
Inventors: 乔丙刚
Original assignee: Individual
Current assignee: Individual
Priority date: 2010-05-07
Filing date: 2010-05-07
Publication date: 2010-12-22
Anticipated expiration: 2020-05-07

Abstract

The utility model relates to a visual talkback system capable of recognizing human faces, in particular to a visual talkback system capable of recognizing human faces and with a monitoring function. The visual talkback system comprises a keyboard module, a talkback module, an image pickup module, a driving module, a human face tracking module, an language navigation module and a central processor, and is characterized in that the keyboard module and the talkback module are connected with the central processor through conducting wires together; the human face tracking module and the language navigation module are connected with the central processor through the conducting wires together; and the image pickup module and the driving module are connected with the central processor. The utility model has the advantages of large capacity for face files, high recognition speed, capability of being applied to the outdoor environment and voice navigation function.

Description

A kind of recognition of face visible dialogue system

Technical field

The utility model relates to a kind of recognition of face visible dialogue system, particularly a kind of recognition of face visible dialogue system that is used to monitor.

Technical background

At present the video intercom product installed of sub-district is as follows: low-grade is non-visible dialogue system, and the visiting guest presses the room number on the host keyboard, the owner can with guest's intercommunication, then by the automatically controlled gate, unit of opening; High-grade is the black and white visible dialogue system, can see guest's black and white image; The colorful visual intercom system can be seen coloured image; The owner of band network savvy can also report to the police to room security.

Summary of the invention

The purpose of this utility model be exactly be to overcome above-mentioned deficiency of the prior art and provide that a kind of facial archives capacity is big, recognition speed is fast, the recognition of face visible dialogue system that can be applied in outdoor environment, have the Voice Navigation function.

For achieving the above object, the utility model is to adopt following scheme to be achieved in that a kind of recognition of face visible dialogue system, it is by Keysheet module (1), intercommunication module (2), photographing module (3), driver module (4), face tracking module (6), language navigation module (7), central processing unit (8) is formed, it is characterized in that: Keysheet module (1), intercommunication module (2) is connected with central processing unit (8) by lead jointly, face tracking module (6), language navigation module (7) is connected with central processing unit (8) by lead jointly, photographing module (3), driver module (4) is connected with central processing unit (8).

This cover system that this programme proposes, on the repertoire of original video intercom, the recognition of face of one of bio-identification, face tracking technology are applied in the visible dialogue system, but also have some special utility model functions, for example: (1) facial archives capacity is big, the memory capacity of 8999 people's face archives; (2) recognition speed is fast, is lower than 1 second; (3) can be applied in outdoor environment,, have waterproof, anti-oxidation function because the video intercom main frame all is mounted in position outwardly, gate, unit; (4) Voice Navigation function can guide the guest correctly to use.

The up-to-date embodiment of building science and technology: based on DSP (digital signal processor)+ARM9 processor is the hardware module and the V3.0 software control unit of core, constitutes the solid foundation of face recognition together.When you stop about 1 second in face of the main frame, main frame will automatically snap your face image, and and the archives of inner storage do contrast, when you were eligible, main frame will send: the prompt tone of " your checking is passed through, and come in ", simultaneously, door is opened automatically.

The function of the utility model recognition of face visible dialogue system:

1, people's face is gathered: by the people's face information under the black and white camera collection secondary light source.

2, people's face comparison: the people's face information in the black and white people's face information and date storehouse that collects is compared.

3, the coloured image of colour TV camera collection is sent to resident family's family upstairs, is shown on visual extension set.

4, registration blacklist, and public security part wanted criminal picture.

5, the recognition of face of registration resident family is opened the door: console mode recognition of face video intercom main frame is installed on the building doorway, resident family is carried out the registration of people's face, as " key " that open the door, have only legal resident family also through after the authentication, can successfully open the door.

6, non-registered resident family can't use the Automatic face recognition system to open the door, and to call out resident family, manually verifies bearer identity by resident family only, and whether decision opens the door to messenger.

7, unidentified people is when the recognition of face video intercom main frame, and main frame detects automatically, if when the wanted circular molecule of its image and storage in advance meets, then trigger alarm signal, to administrative staff's mail sending charge per message onsite alarming signal, and take lawless person's photo, solving a case for public security in the future creates conditions.

8, can set the people that each is passed in and out the unit, carry out mug, the picture archiving time is 7 days.

9, modularized design includes: recognition of face module, face tracking module, video drive module, video intercom module, the module of unblanking, warning module.

10, off-line working can independently be used, and also can onlinely use.

11, have general-purpose interface, can increase peripheral apparatus as required, as output equipment, input equipment.

12, tamper function is when equipment goes to pot or during service interruption, alarm function arranged.

13, identification accuracy height, misclassification rate are low: misclassification rate FAR≤0.01% o'clock, discrimination VR 〉=99.9%.

14, identification and comparison speed are fast: recognition speed 1: 1 o'clock＜1 second, and 1: during N＜2 seconds.

15, environmental suitability excellence: be not subjected to the influence of the ethnic colour of skin and sex; Not influenced by light.

16, self-learning function: system can be dynamically user's head portrait in the new database more, thereby the people's face that guarantees the user and the variation of environment.Still can correctly discern fast.

17, network savvy, system has independence, can carry out networking simultaneously and set up, and has reached the infinite expanding to system.

18, this series main frame adopts 4*4 digital-code type button, can be implemented in 8999 and selects the numeral of any appropriate to come indoor unit is carried out geocoding according to demand, both can register 8999 people from resident family face information.

Description of drawings

Fig. 1 is the utility model structural representation

Embodiment

A kind of recognition of face visible dialogue system, it is made up of Keysheet module (1), intercommunication module (2), photographing module (3), driver module (4), face tracking module (6), language navigation module (7), central processing unit (8), it is characterized in that: Keysheet module (1), intercommunication module (2) are connected with central processing unit (8) by lead jointly, face tracking module (6), language navigation module (7) are connected with central processing unit (8) by lead jointly, and photographing module (3), driver module (4) are connected with central processing unit (8).

Before the guest comes the doorway, unit, the outer red sensor circuit of this host computer using human body, sense human body signal after, speech guide system begins to start: " you are good; XX welcomes the sub-district your presence; please rest on the last rice of main frame position, be your identity verification ", " dripping " one, represent face recognition to begin, eligible person " your checking is passed through, and the gate is opened for you now, and come in ".Identification is not by person " you are this unit resident families, would you please call out owner's room number that you will visit ", and " dial, just a moment,please " puts through back prompting " owner and you talk, and please say ".If the owner allows the visitor to enter " door is opened, and come in ".

Claims

1. recognition of face visible dialogue system, it is made up of Keysheet module (1), intercommunication module (2), photographing module (3), driver module (4), face tracking module (6), language navigation module (7), central processing unit (8), it is characterized in that: Keysheet module (1), intercommunication module (2) are connected with central processing unit (8) by lead jointly, face tracking module (6), language navigation module (7) are connected with central processing unit (8) by lead jointly, and photographing module (3), driver module (4) are connected with central processing unit (8).