Online early warning system of weariness working and method for early warning thereof
Technical field
The present invention relates to a kind of early warning system, relate in particular to a kind of fatigue tendency and fatigue state that user is occurred in the course of the work and carry out the ONLINE RECOGNITION also online early warning system of weariness working and the method for early warning thereof of alarm.
Background technology
The operator on duty of key post especially at night, the situation of mental fatigue, the decline of energy intensity occurs through regular meeting in long work process.This will cause dizzy, the symptom of dozing off of its appearance, thereby decision error, misperception and operation mistake take place.Decision error mainly is forecasting inaccuracy and misjudge, and misperception comprises negligence, absentminded, improper, the visual deterioration of attention.Operation mistake mainly is that action is inaccurate.The behavior mistake that these fatigues cause is the major reason that causes operating irregularity to take place.According to statistics, society is caused in the motor vehicles vehicle accident of huge life, property loss, surpass 25% owing to driver's mental fatigue causes the ratio that energy is not concentrated even the overtired death by accident that causes drowsiness to cause accounts for the vehicle accident sum.
At present existing multinomial patent relates to by gathering people's bio information, and fatigue state is differentiated.For example, video camera is installed in driver's cabin, is gathered driver's facial information and connect image processing equipment.Owing to well-known reason, the instability of image, user of service's attitude variation etc. can cause the feature extraction difficulty, and especially the accurate tracking of eyes is difficult to realize reliably.In addition, because biological characteristic exists than big-difference between the nonspecific user, cause using merely the method for the facial characteristics judgement mental status not ideal enough.
Also have scheme to gather the bioelectrical signals of user, for example body electricity, brain electricity etc., and the foundation of differentiating as its mental status after treatment.These class methods exist Body contact measurement requirement harshness, thereby cause problems such as harvester output individual variation is big, use inconvenience, and causing can't online use in daily production, life activity.
Along with development of electronic technology, microminiaturized, portable image capture device has been suitable for human body and has dressed.Easy to use and cheap human heart rate and breathing rate sensing device obtain popularization and application at medical domain.Under normal condition, bio signals such as the heart rate of being convenient to gather, breathing rate are less demanding with contacting of health to harvester, the more important thing is, medical research confirms that already there is definite relation in this category information and people's the mental status.
Summary of the invention
Purpose of the present invention is exactly in order to solve used fatigue early warning device at present, can not well satisfy the early warning requirement, can't determine the mental status automatically according to individual variation, play problems such as forewarning function, a kind of specification for structure unification that has is provided, easy to use, can accurately judge the user mental status, the reliable also energy of forewarning function auxiliary improvement user of service's energy intensity, the online early warning system of weariness working and the method for early warning thereof of preventing accident and taking place.
For achieving the above object, the present invention adopts following technical scheme:
The online early warning system of a kind of weariness working, it comprises battery, alarm device, it is arranged on the eyes image harvester that human body head is used to gather the action of people's eye in addition, be arranged on the heart rate and the breathing rate harvester of human lumbar abdominal part, and the training control interface, they all are connected with tired recognition device, and tired recognition device then is connected with alarm device.
Described eyes image harvester comprises working cap, is provided with minisize pick-up head at its shade front end and human eye correspondence position, and is connected with tired recognition device by signal transmssion line.
Be provided with the infrared illumination source of aiming at its eye with user eye correspondence position on the shade of described working cap.
Described heart rate and breathing rate harvester comprise webbing, and its two ends are provided with the union joint and the Connection Block of movable connection, and waist abdomen formula pressure transducer connects firmly on webbing, and is strapped in the human lumbar abdominal part.
Described tired recognition device comprises the digital signal processor that adopts dsp chip, and it is connected with the eyes image harvester with reception eye signal by the picture signal interface, and utilizes the eye messaging software to handle; Be connected with heart rate and breathing rate harvester device by the pressure signal interface, receive the heart rate and the breathing rate signal of human body, and utilize heart rate and breathing rate process software to handle; Form software by training control interface utilization training template and form training template characteristic parameter; Fatigue state by analysis of fatigue identification software identification user; Be connected with alarm device by alarm control interface.
Described alarm device comprises that voice storage and playing circuit, output amplifier, speaker form; Tired recognition device triggers voice storage and playing circuit according to recognition result by alarm control interface, and voice signal is through amplifying output, and the prompting user is on your toes, concentrate one's energy.
The online early warning system method for early warning of a kind of weariness working, its step is
A, system power-up start, and carry out initialization;
B, gather the user eyes image by the eyes image harvester, with the picture frame is unit, utilizes software that the user eyes image is handled, and calculates position, pupil size, the folding degree of eyes, catacleisis time, catacleisis frequency change next step then over to; As not gathered a two field picture, then directly change next step over to;
C, gather the heart rate and the breathing rate pressure data of user, as gathered, then utilize computed in software heart rate and breathing rate respiratory frequency and variation thereof by heart rate and breathing rate harvester; As not gathered, then directly change next step over to;
D, judge whether to be in physical training condition at present, as be in physical training condition, then generate the size that the driver status template parameter comprises eyes, closed frequency, the size of pupil, respiratory frequency by software; As be not in physical training condition, and then judge software by tired tendency, carry out the fatigue state pattern recognition according to driver's eyes image and breathing rate parameter;
E, according to recognition result, if user is in fatigue state then gives the alarm by alarm device; As be not in fatigue state, and then go back among the step b, detect judgement again.
The present invention gathers user eye information, heart rate, breathing rate information in real time in the early warning process.It adopts the wear-type imageing sensor, gathers the eye information of user.The position relative fixed of imageing sensor and face image, the image information of obtaining is vibrated, the influence of ambient light is little, is not subjected to the influence of user posture.These characteristics make the computational processing of eyes image greatly reduce, and the action message of the eyes of extraction is more accurate.Adopt heart rate and the breathing rate and the yawn information of waist abdomen formula sensor acquisition user.Extract the relevant characteristic parameters such as eye pupil variation, catacleisis frequency change, the variation of catacleisis time, changes in heart rate and change of respiratory rate of the reaction user mental status by information processing, the application mode recognition principle, analyze the fatigue tendency that user shows, its fatigue state is mated identification.Equipment is differentiated the result according to fatigue state, the information that gives a warning, thereby auxiliary user concentrate one's energy or take corresponding measure, avoid causing error of performance and accident because of tired reason.
The invention has the beneficial effects as follows: version is unified and standard, and information gathering is accurately convenient, can send early warning more accurately, has reduced tired reason and has caused error of performance and accident.
Description of drawings
Fig. 1 is the online source of early warning block diagram of fatigue driving;
Fig. 2 is an eyes image harvester composition frame chart;
Fig. 3 is heart rate and breathing rate harvester composition frame chart;
Fig. 4 is tired recognition device composition frame chart;
Fig. 5 is the alarm device composition frame chart;
Fig. 6 is the flow chart of the online early warning software of weariness working;
Fig. 7 is an electrical schematic diagram of the present invention;
Fig. 8 is tired recognition device electrical schematic diagram;
Fig. 9 is the electrical schematic diagram of picture signal interface circuit.
Wherein, 1. eyes image harvester, 10. working cap, 11. minisize pick-up head, 12. infrared illumination source, 13. signal transmssion lines, 2. heart rate and breathing rate harvester, 20. webbings, 21. pressure transducers, 22. union joint, 23. Connection Blocks, 3. tired recognition device, 30. digital signal processor, 31. picture signal interfaces, 32. pressure signal interfaces, 33. the eye messaging software, 34. hearts rate and breathing rate process software, 35. training templates form software, 36. analysis of fatigue identification software, 37. alarm control interface, 4. alarm device, storage of 40. voice and playing circuit, 41. output amplifier, 42. speaker, 5. training control interface, 6. battery.
The specific embodiment
The invention will be further described below in conjunction with accompanying drawing and embodiment.
The online source of early warning of weariness working partly is made up of eyes image harvester 1, heart rate and breathing rate harvester 2, tired recognition device 3, alarm device 4, training control interface 5 etc., and the interactively of system's each several part as shown in Figure 1.
The effect of eyes image harvester 1 is an eyes image of gathering user.It is made up of working cap 10, minisize pick-up head 11, infrared illumination source 12 and signal transmssion line 13.Minisize pick-up head 11, infrared illumination source 12 are installed on the shade of working cap 10, and direction is aimed at the user eye.Infrared illumination source 12 is used for strengthening under the dark condition at night clear picture figure.The structure of working cap 10 as shown in Figure 2, working cap 10 shade portion materials can see through Infrared.Minisize pick-up head 11 is selected the CMOS photographic head of infrared ray responsive for use, and the left eye or the right eye of user pointed at its center of finding a view, and viewfinder range is to be first face at center with these eyes.Picture frame resolution 640 * 480 (300,000 pixel).One group of near infrared luminescence diode is as the lighting source under the night dark condition, and illumination zone is identical with the photographic head viewfinder range.The components and parts and the software code of minisize pick-up head 11, infrared illumination source 12, tired recognition device 3 parts at first are assemblied on the multilayer board, are installed on the shade of working cap 10 again.
The using method of heart rate and breathing rate harvester 2 is similar to belt, lies in the waist and belly of user.As shown in Figure 3, it is made up of webbing 20, pressure transducer 21, union joint 22, Connection Block 23 etc.Pressure transducer 21 is fixed together with webbing 20.Union joint 22 and Connection Block 23 are separately fixed at the two ends of webbing 20, and union joint 22 and Connection Block 23 hasps use, and make webbing 20 be adjacent to the user health.Its pressure transducer 21 signal amplitude peaks generally can reach 2 volts, are connected with tired recognition device 3 by holding wire.
Tired recognition device 3 comprises that as Fig. 4, Fig. 7, shown in Figure 8 digital signal processor 30, picture signal interface 31, pressure signal interface 32, eye messaging software 33, heart rate and breathing rate process software 34, training template form parts such as software 35, analysis of fatigue identification software 36.The data that picture signal interface 31 receives from eyes image harvester 1, and carry out data according to the picture frame form and keep in, its circuit diagram is as shown in Figure 9.Digital signal processor 30 operation eye messaging softwares 33 as handling unit, at first become black white image to image transitions by the two-value method with a two field picture, then image are corroded and expansion process, filter out the clutter in the image.Carry out level and vertical histogram computing, judge the band of position of eyes.Further the image to eyes carries out level and vertical histogram analysis, judges the closure of eyes, the size of eyelid aperture pupil.After in succession frame data are carried out relevant treatment, calculate catacleisis frequency, catacleisis time, average eyelid aperture size.Pressure signal interface 32 will convert digital form to from the analog pressure signal of heart rate and breathing rate harvester 2.Signal aroused in interest is the impulse wave that contains higher frequency components, and breathing transducing signal is the very low impulse wave of frequency.Heart rate and breathing rate process software 34 extract low-frequency component and radio-frequency component from digital pressure signal, and further calculate the breathing rate and the hrv parameter of user.
When training control interface 5 was in physical training condition, digital signal processor 30 operation training templates formed software 35.Gather eyes image, heart rate, the breathing rate information of (mental status is good) under the user normal operating conditions, behind above-mentioned signal processing, parameter extraction, further handle, form training template characteristic parameter through statistical method.
System is under the online alert status, digital signal processor 30 operation analysis of fatigue identification softwares 36.The criterion of carrying out fatigue state coupling identification mainly contains: in certain persistent period, the increase frequency of catacleisis, the time lengthening of catacleisis, eyelid aperture diminish, heart rate and breathing rate slows down, yawn increases etc.Utilization feature fusion technology improves the sane degree of identification.The relative variation that adopts characteristic parameter carries out fatigue tendency ONLINE RECOGNITION as coupling tolerance.When occurring a plurality of characteristic matching simultaneously, judge that fatigue state appears in user.Recognition result outputs to alarm device 4.
Alarm device 4 adopts digital speech storage and play-back technology.As shown in Figure 5, form by voice storage and playing circuit 40, output amplifier 41, speaker 42.Store multiple information and warning message in advance in voice storage and the playing circuit 40, and accept the control of alarm control interface 37.The voice reminder or the alarm sound signal of the output of this circuit are reduced to sound through amplifying circuit rear drive speaker, and the prompting user is on your toes, concentrate one's energy.Alarm device 4 is connected with tired recognition device 3 by holding wire.
Battery 6 provides power supply for other parts of this equipment.Alarm device 4 and battery 6 are assembled in the housing, and the profile of housing is similar to mobile phone handsets.According to environment for use, housing can be worn on the belt of user or hangs on the body of user, also can place on work chair or the workbench position.Fig. 7 has provided electrical schematic diagram of the present invention.
Training control function is realized by one group of master cock.After control entered physical training condition, the template characteristic data began to form automatically., reach after regulation requires as the basic foundation of convergence with the variable quantity of short-time average value, system provides training and finishes information.After training process was finished, equipment can change the online alert status of weariness working over to.
Among Fig. 6, the online early warning system method for early warning of weariness working of the present invention is:
A, system power-up start, and carry out initialization;
B, gather the user eyes image by the eyes image harvester, with the picture frame is unit, utilizes software that the user eyes image is handled, and calculates position, pupil size, the folding degree of eyes, catacleisis time, catacleisis frequency change next step then over to; As not gathered a two field picture, then directly change next step over to;
C, gather the heart rate and the breathing rate pressure data of user, as gathered, then utilize computed in software heart rate and breathing rate respiratory frequency and variation thereof by heart rate and breathing rate harvester; As not gathered, then directly change next step over to;
D, judge whether to be in physical training condition at present, as be in physical training condition, then generate the size that the driver status template parameter comprises eyes, closed frequency, the size of pupil, respiratory frequency by software; As be not in physical training condition, and then judge software by tired tendency, carry out the fatigue state pattern recognition according to driver's eyes image and breathing rate parameter;
E, according to recognition result, if user is in fatigue state then gives the alarm by alarm device; As be not in fatigue state, and then go back among the step b, detect judgement again.