Summary of the invention
The utility model is at the deficiencies in the prior art, a kind of electronic evidence-collecting system that can tightly collect evidence to the vehicle horning against traffic regulation phenomenon is proposed, and the electronic police system that a kind of this kind electronic evidence-collecting system adopts proposed, electronic evidence-collecting system and electronic police system that utilization the utility model proposes can carry out automatic evidence-collecting to horning against traffic regulation.
The utility model proposes the technical scheme of the electronic police device of this system of employing:
A kind of electronic evidence-collecting device comprises Mike's battle array, audio collection module, camera, video acquisition module, information transmission modular, wherein Mike's battle array is made of two or more microphones, links to each other with information transmission modular by the audio collection module; At least one camera links to each other with information transmission modular by video acquisition module; Information transmission modular will send host computer to from the acoustic information of Mike's battle array collection and image information sequence and the acoustic information sequence of receiving in real time.
As replenishing of technique scheme, the camera control module can also be followed the trail of by the further details that information transmission modular is subjected to the control of host computer to carry out vehicles peccancy and be taken.
In order to make the whistling auditory localization more accurate, the electronic evidence-collecting device can also comprise wind transducer and air velocity transducer, and host computer is to the whistling auditory localization time, and the wind speed and direction information of utilizing the time period of blowing a whistle to gather compensates auditory localization.
The utility model is based on above-mentioned electronic evidence-collecting device, the complete violation whistling electronic police system for vehicle of one cover is proposed, this system comprises Mike's battle array, audio collection module, camera, video acquisition module, information transmission interface, host computer, wherein Mike's battle array is made of two or more microphones, links to each other with information transmission modular by the audio collection module; At least one camera links to each other with information transmission modular by video acquisition module; The acoustic information sequence of Mike's battle array collection and receive image information sequence real-time Transmission and give host computer.
Realize being difficult to degree and the consideration of equipment cost aspect from technology, whistling identification and auditory localization also have the image labeling function to be finished by host computer: by information transmission modular with real-time image information sequence and acoustic information sequence transmission to host computer, host computer carries out auditory localization according to the relative position relation between Mike's battle array and the camera with from the acoustic information of Mike's battle array collection when recognizing whistling, the whistling sound source position is labeled on the image information of being gathered, and the image information behind the storage mark.
In order to make auditory localization more accurate, vehicle horning against traffic regulation electronic evidence-collecting system can also comprise wind transducer and air velocity transducer, microprocessor module or host computer are to the whistling auditory localization time, and the wind speed and direction information of utilizing the time period of blowing a whistle to gather compensates auditory localization.
The utlity model has following beneficial effect:
The utility model has thoroughly solved the difficult problem of horning against traffic regulation punishment evidence obtaining, the evidence that collects horning against traffic regulation that not only can strict logic, and automatic violation whistling electronic evidence-collecting system is provided.A lot of traffic hazards all are accompanied by the sound of blowing a whistle, so the violation whistling electronic evidence-collecting system that the utility model provides also has indirect accident record function.
Embodiment
Mike's battle array comprises a plurality of microphone arrangement with certain predetermined layout, and these microphones are used for catching from all directions and originate from the sound wave of the different point in space, and Mike's battle array is generally used for speech processing system.Speech processing system is when going back original sound, needs for elimination neighbourhood noise etc., the needs basis is carried out auditory localization from the voice messaging of each sound channel of Mike's battle array collection, seeking the direction of pointing to sound source has play a part important in spatial filtering, in some cases, the direction of sensing sound source is used to the aftertreatment of the sound signal of speaker's tracking and record.Carry out auditory localization by Mike's battle array, prior art has had multiple application mode.For example patent of invention 02800828.6 discloses a kind of method and apparatus that is used for video conference system, and it adopts the array that comprises two microphones and stillcamera automatically to find out speaker's position and electricity operation video image to produce the effect of movable translation-inclination-convergent-divergent (" PTZ ") camera.Adopt computer vision algorithms make to detect, locate and follow the trail of personage in the visual field of the stillcamera that is in wide-angle.Employing includes only the microphone array of the microphone of opening two horizontal intervals as the auditory localization device, and the estimation sound lag of utilizing this microphone array to obtain is selected just the talker.The multichannel audio information that the utility model also utilizes Mike's battle array to collect is carried out auditory localization.
The detector that the utility model adopts as shown in Figure 1,3 cameras (3,4,5) and 4 microphones (6,7,8,9) be installed in jointly on the same support 1,4 microphones are connected with support 1 by link frame 2 respectively, wherein microphone 6, microphone 7, microphone 8 horizontal positioned, the central point 80cm of distance bracket 1, common formation Mike battle array, camera 3, camera 4, camera 5 also is installed on the support 1, camera 3, camera 4, camera 5, microphone 6, microphone 7, microphone 8 is staggered the installation on same surface level, each adjacent camera becomes 60 degree angles respectively with microphone link frame 2, and microphone 9 is installed in the central point vertical direction 80cm place of support 1.
In the present embodiment, 3 cameras and Mike's battle array are integrated on the same support, in the practical application, also can distinguish independent installation.But install in any case, system when carrying out auditory localization and image labeling, the relative position information of each microphone and the relative position information of Mike's battle array and each camera in the Mike's battle array that all needs to know for sure.The utility model is also installed air velocity transducer 10 and wind transducer 11 in cantilever tip.The wind speed and direction sensor is not the necessary electron device of the utility model embodiment, and utilizing their collection wind directions and wind velocity signal purpose is in order more accurately to carry out the auditory localization compensation.
Fig. 2 is the structured flowchart of vehicle horning against traffic regulation electronic evidence-collecting of the present utility model system.Vehicle horning against traffic regulation electronic evidence-collecting of the present utility model system comprises the electronic evidence-collecting device that can be installed in the road monitoring point and is used for computing, determines sound source position and carries out computing machine (or being called host computer) two parts of image labeling that the electronic evidence-collecting device comprises Mike's battle array, audio collection module, one or more camera, video acquisition module, microprocessor module, storer etc.Mike's battle array links to each other with microprocessor module by the audio collection module; One or more cameras link to each other with microprocessor module by video acquisition module.
Microprocessor module is according to the acoustic information Real time identification automobile whistling of gathering from Mike's battle array, and image information sequence and the acoustic information sequence that receives in each time period of blowing a whistle stored in the storer respectively.The definition of the time period of blowing a whistle here is: to recognize the moment that whistling violating the regulations exists is time reference, one period blink of intercepting after this moment, as one minute, perhaps before this moment and intercept one period blink afterwards, each 30 second before and after for example getting.Microprocessor module can have multiple way of realization, for example, utilizes single-chip microcomputer as master controller, utilizes DSP or voice recognition chip to finish identification to the vehicle horning against traffic regulation whistling.
The sound identifying function of microprocessor module is to realize like this, at first the acoustical signature of blowing a whistle of various vehicles is extracted and set up the voice recognition model database of blowing a whistle, and it is pre-deposited the microprocessor that is used for whistling identification in the electronic evidence-collecting device, in DSP or voice recognition chip.When the microprocessor that is used for whistling identification carries out real-time whistling identification, whether the sound of the surrounding environment that at least one microphone in the Mike's battle array that collects is picked up is compared with the voice recognition model database of blowing a whistle, have vehicle to blow a whistle in the surrounding environment this moment thereby judge.If not then continue identification, if have, it then is benchmark with the current time, the image of this time period and the wind speed and direction information (if using the wind speed and direction sensor) of acoustic information (the multichannel digital sound information of Mike's battle array) and air velocity transducer and wind transducer record are stored in the storer, continue then the acoustic information in the surrounding environment is carried out Real time identification.
The electronic evidence-collecting device is installed in road monitoring point place, data communication interface between electronic evidence-collecting device and the computing machine can be designed to various ways, for example, if near road monitoring point, be provided as the industrial computer of host computer, electronic evidence-collecting device and industrial computer can pass through RS233 between the two, communications such as USB, receive the acoustic information and the image information that collect in each time period of blowing a whistle that the transmission of electronic evidence-collecting device comes by industrial computer, and carry out whistling auditory localization and image labeling, at this moment, industrial computer is called host computer for the microprocessor module of electronic evidence-collecting device.Industrial computer carries out data communication by the control center of cable network or wireless network and vehicle supervision department again.If host computer is not set near the road monitoring point, as the computer installation of host computer control center in vehicle supervision department, the electronic evidence-collecting device can also pass through industrial field bus, Ethernet interface communication or adopt communication and compunication, the image information sequence and the acoustic information sequence that receive in the time period is blown a whistle in the computing machine transmission can be handled also transmission process at times in real time.In this case by the computer realization auditory localization and the image labeling function that are arranged in the control center of vehicle supervision department.
Host computer is handled the blow a whistle information of time period of each group, the process of handling comprises two steps, at first the whistling sound source that goes out to blow a whistle of the installation site information calculations by Mike's battle array and the three dimensions relative position relation of Mike's battle array, the location algorithm principle that the utility model embodiment adopts is as follows:
Four identical microphone S
1, S
2, S
3, S
4Be placed in the measured zone, these sensor location coordinates are respectively S
1(x
1, y
1, z
1), S
2(x
2, y
2, z
2), S
3(x
3, y
3, z
3), S
4(x
4, y
4, z
4).(when z) having located the sound source generation, sound wave passes to microphone to arbitrfary point T in measured zone, produces four acoustical signal S respectively for x, y
1(t), S
2(t), S
3(t), S
4(t).Microprocessor on each microphone is at first carved t
1, t
2, t
3, t
4, and give host computer.The real time that sound source arrives four microphones is t
p+ t
1, t
p+ t
2, t
p+ t
3, t
p+ t
4T wherein
pFor sound event began to four times that microphone begins to sample simultaneously of supposition, four microphones need have same sampling time benchmark.Host computer is gathered simultaneously from the other wind speed of measured zone, the signal of wind transducer, and obtains local wind vector V
w(v
x, v
y, v
z).Also can obtain local sound propagation velocity C by suitable sensor host computer.So following system of equations (1) is arranged to be set up
(x
1-x)
2+(y
1-y)
2+(z
1-z)
2=[(c+v
1)·(t
p+t
1)]
2
(x
2-x)
2+(y
2-y)
2+(z
2-z)
2=[(c+v
2)·(t
p+t
2)]
2
(x
3-x)
2+(y
3-y)
2+(z
3-z)
2=[(c+v
3)·(t
p+t
3)]
2
(x
4-x)
2+(y
4-y)
2+(z
4-z)
2=[(c+v
4)·(t
p+t
4)]
2
(1)
V wherein
1, v
2, v
3, v
4Be respectively wind vector V
w(v
x, v
y, v
z) at vector
TS
1, TS
2, TS
3, TS
4On projection.They are calculated by relational expression (2).
(2)
System of equations (1) is carried out cancellation t
pProcessing, obtain system of equations (3)
T wherein
21=t
2-t
1, t
31=t
3-t
1, t
41=t
4-t
1, constitute sound source time corresponding difference vector Δ T
m(t
21, t
31, t
41).
System of equations (3) is about variable x, y, z, equation of higher degree group, the mistiming vector Δ T of the corresponding sound source position that obtains by observation
m(t
21, t
31, t
41) can in the hope of the position T of sound source (x, y, y).For four microphone position coordinates arbitrarily, wind speed V
wAnd local velocity of sound C, find the solution with numerical solution.
Above equation passes through numerical methods of solving by host computer, draw the sound source position coordinate, the acoustical signal of being accepted on each microphone with Mike's battle array carry out suitable digital signal processing and calculate signal characteristic peaks constantly the substitution equation all by host computer automatically operation finish.When considering implementation cost air velocity transducer and wind transducer can be installed, wind speed is brought calculating into 0 value.
After obtaining the position coordinates of whistling sound source, use three dimensional virtual technique to come out the locus of point source of sound is virtual by the relative position relation between Mike's battle array and the camera, utilize camera position information and angle lens information to make the Softcam visual angle, use virtual technology with the virtual sound source position point that comes out of virtual camera visual angle shooting, draw this two dimensional image on the camera photographic images, obtain this two-dimensional coordinate on the camera photographic images, this coordinate is marked on the image of the camera real scene shooting of (calculating the moment of the employed acoustic information of sound source position) with constantly.The vehicle of whistling sound source position mark is exactly the vehicle of blowing a whistle at that time on the image.The time and space consistance that Here it is utilizes acoustics position judgment and optical position to judge.
The detector of present embodiment adopts the Mike's battle array that is made of four microphones, and in actual applications, the number of microphone is not fixed, the minimum Mike's battle array that is made of two microphones can be arranged to, patent of invention 02800828.6 as mentioned above.In the prior art, the number that is used for the microphone of auditory localization has had 6,8 etc., and its space layout mode also is diversified.Generally speaking, distance should not be too near between the microphone in Mike's battle array, and during apart from each other, antijamming capability is stronger, and it is many more that the number of microphone is provided with, the acoustic information that may obtain also many more, auditory localization also can be more accurate.
In the present embodiment, the electronic evidence-collecting device just is used for discerning whistling, and an acoustic information and an image information of gathering with each time period of blowing a whistle be transferred in the host computer, is finished the function of auditory localization and image labeling by the host computer machine.In actual applications, if the microprocessor module of electronic evidence-collecting device designs the DSP that special execution auditory localization and image labeling function are arranged in addition, if add controlled zoom camera and controlled The Cloud Terrace, the electronic evidence-collecting device just can be controlled camera at that time sound source was carried out track up more accurately, detail characteristic with the vehicle that obtains more to blow a whistle, more help the law enforcement of traffic control department, what store in the storer so is exactly more accurate sound source position information and be not only multichannel digital audio signal, can be transferred to host computer by various data transfer meanses equally certainly and carry out image labeling and handle.
Traffic control department can punish the vehicle that is marked according to the mark on each camera image in accordance with the law, seeks the detail characteristics such as license plate numbers that are marked vehicle from image.
Physical circuit is realized referring to seeing Fig. 4-8.
Present embodiment adopts the dual micro processor structure: master controller is C8051F040, utilizes voice recognition chip HL7300 to carry out voice recognition, and storage chip is FLASH, and the USB communication chip is C8051F340.
1) P4.7 of C8051F040~4.0 pin, P5.0~5.2 foot control system voice recognition chips operation, P5.4~P5.7, P6.0~6.7 are read voice recognition data and are deposited single-chip microcomputer in.P1.0~P1.7 is a FLASH storer control port, and P2.0~2.7 are the FLASH FPDP.P0.0~0.1 is and C8051F340 serial communication interface, receive No. four microphones (MIC0, MIC1, MIC2, MIC3) acquired signal and carry out AD by interface (18~21) and be converted to digital signal and handle and store, the data of single-chip microcomputer (52~54 pin) receiver, video capture card are handled and are stored.
2) speech recognition circuit is formed with master chip HL7003, and the speech recognition circuit that HL7003 forms has two kinds of main operator schemes: word logging mode and word recognition mode.Except these two kinds of patterns, HL7003 can also detect button and produce output signal in signal.All operation is all handled by the HL7003 internal processes, any in No. four microphones leaded up to circuit and inserted 27 pin of HL7003, P4.7~4.0 ports by single-chip microcomputer C8051F040, P5.0~5.2 port controlling voice recognition chips operation, C8051F040 also controls the storage of signal by the recognition result of HL7003.
3) FLASH is a mass-memory unit, is used to store the sound data file relevant with the whistling that recognizes.
4) the C8051F340 chip is used for the USB communication, by USB interface with as the computing machine communications exchange data of host computer.
Need to prove, the utility model belongs to a kind of component invention, be a kind of innovation and creation that on technology is used, have novelty, wherein related Mike's battle array, camera, wind direction and air velocity transducer, video acquisition module, image capture module, voice recognition technology, auditory localization technology, image labeling technology etc. all belong to prior art and existing various ways and application.Here just with the embodiment of the present utility model detailed explanation that has been center deployment, the imbody of described optimal way or some characteristic, should be understood to this instructions only is to describe the utility model by the mode that provides embodiment, in fact on some details of forming, construct and using, can change to some extent, comprise the combination and the assembly of parts, these distortion and application all should belong in the scope of the present utility model.