CN201194137Y

CN201194137Y - Novel hand language recognition device

Info

Publication number: CN201194137Y
Application number: CNU2007200447684U
Authority: CN
Inventors: 刘海宽; 陈传虎
Original assignee: Individual
Current assignee: Individual
Priority date: 2007-10-26
Filing date: 2007-10-26
Publication date: 2009-02-11
Anticipated expiration: 2017-10-26

Abstract

The utility model discloses a novel sign language recognition device mainly comprising a sensor glove part and a data acquisition and processing part. The sensor glove for tracking and measuring the gesture motion trajectory and the temporal order information mainly includes a linear displacement sensor, a contact switch and a proximity switch; the data acquisition and processing part for collecting the output information from the sensor glove and accomplishing the effective recognition of the sign language based on the corresponding rule between the sensor state and the manual alphabet includes an A/D converter AD0809, an AT89C52 single chip, a data memory and a display driving circuit SD0402.

Description

A kind of novel Sign Language Recognition device

Technical field

The utility model relates to a kind of Sign Language Recognition device, particularly adopts the Sign Language Recognition device of linear displacement transducer.

Background technology

Sign language is to express the meaning by the motion of hand, is mainly used in deaf-mute's information interchange.

The effect of Sign Language Recognition device is translated into the Chinese manual alphabet automatically with deaf-mute's sign language exactly, is convenient to deaf-mute and normal person's communication, helps them to participate in every social activities.The Sign Language Recognition device also can be used for sign language training and Intelligent Human-Machine Interface field.

From the realization principle of Sign Language Recognition device, Sign Language Recognition mainly is divided into based on the Sign Language Recognition of vision (image) with based on the identification of data glove.Sign Language Recognition based on vision is to utilize the camera collection to tidy up information, and discerns, and this sign Language Recognition Method needs expensive equipment, the cost height of realization, and be not easy to carry; Sign Language Recognition based on data glove, be to utilize the sensor tracker on the data glove to measure track and the time sequence information of gesture in spatial movement, and this information translated into manual alphabet, utilize the Sign Language Recognition installation cost of this recognition methods design low, and be easy to carry.Therefore obtained using widely, but the sensor on the data glove of this Sign Language Recognition device that uses all adopts the strain resistance type pressure transducer at present, this sensor price is expensive, the life-span short, directly causes Sign Language Recognition installation cost height, the failure rate height.

Summary of the invention

The purpose of this utility model just provides the deficiency that a kind of novel Sign Language Recognition device overcomes present Sign Language Recognition device.The sign Language Recognition Method that the utility model is based on data glove adopts cheap linear displacement transducer to replace the Sign Language Recognition device of strain resistance type pressure transducer, mainly comprise the two large divisions: sensor gloves part and data acquisition and processing section, as Fig. 1.

Wherein the sensor gloves can be used for tracking measurement gesture motion track and time sequence information, the sensor gloves are mainly by linear displacement transducer, contact-making switch, constitute near switch, utilize linear displacement transducer identification finger state, contact-making switch is distinguished contact information between finger, near the direction of motion of switch identification wrist, thereby tracker is measured track and the time sequence information of gesture in spatial movement; The information of data acquisition and the output of processing section pick-up transducers gloves, 30 manual alphabet gestures of centering state sign language are discerned effectively, and can show the result of identification, allow the people who is ignorant of sign language understand the information of sign language, thereby reach the purpose that helps deaf-mute and other people communication.

Description of drawings

Fig. 1 is the structured flowchart of this Sign Language Recognition device.

Fig. 2 is the structural representation block diagram of sensor gloves.

Fig. 3 is the structural drawing of data acquisition and processing section.

Fig. 4 is the course of work synoptic diagram of Sign Language Recognition device.

Embodiment

This device mainly comprises the two large divisions: sensor gloves part and data acquisition and processing section, and as Fig. 1.

Wherein, sensor gloves parts as shown in Figure 2, the sensor gloves are made up of near switch 5 linear displacement transducers, 6 contact-making switches and 1.1-5: linear displacement transducer, 6-11: contact-making switch, 12: near switch.Linear displacement transducer is used to detect the state of finger: 1. straight; 2. less bending; 3. than macrobending; 4. crooked fully.Contact-making switch is used to judge contact information; The direction of motion that is used to monitor wrist near switch.

Data acquisition and processing section are made of A/D converter AD0809, AT89C52 single-chip microcomputer, data-carrier store, display driver circuit SD0402.As shown in Figure 3.

Principle of work and process: the state of 1-5 linear displacement transducers is divided into: straight (00), less bending (01), than macrobending (10), fully crooked (11) }; The state of 6-11 contact-making switches is divided into: { not contacting (0), contact (1) }; Divide near the state of switch the 12nd: { { not contacting (0), contact (1) }.The state of sensor and 30 manual alphabets are set up corresponding 30 rules.With A, B, C is example, represents the state of linear displacement transducer with 2 bits, represents contact-making switch with 1 bit respectively and near the state of switch, quantizes respectively then, and value that each letter is corresponding unique sees Table 1.

Table 1: the state of sensor and manual alphabet rule of correspondence example

When putting on the sensor gloves, the user carries out sign language when operation, each linear displacement transducer and will produce corresponding the variation along with the variation of operator's hand shape near the output of switch.The CPU (central processing unit) that is made of the AT89C52 single-chip microcomputer detects in real time to the output of each sensor, to after the detected data processing with the corresponding comparison of 30 sign languages rule discriminatory analysises, draw the manual alphabet of sign language action expression, and show the output manual alphabet with the form of text, finish the identification of sign language, workflow as shown in Figure 4.

Claims

A kind of Sign Language Recognition device, this device mainly comprises the two large divisions: sensor gloves part and data acquisition and processing section, wherein the sensor gloves are responsible for collecting the variation of operator's hand shape, and this change transitions become electric signal, the signal of data acquisition and the output of processing section pick-up transducers gloves, and the regular corresponding manual alphabet identification of relatively finishing of signal that collects and sign language, it is characterized in that: the sensor gloves are to use linear displacement transducer, contact-making switch, realize near switch, wherein linear displacement transducer detects the state of finger, contact-making switch is used to judge contact information, is used to monitor the direction of motion of wrist near switch.