Utility model content
To solve the above problem of the existing technology, the utility model to design it is a kind of can situation complex water areas under water
Automatic obstacle avoiding and the target identification tracking system that the unmanned boat that specific objective identification in sea is tracked is carried out on the basis of automatic obstacle avoiding
System reduces the risk of operation to enrich the operations offshore content of unmanned boat.
To achieve the goals above, the technical solution of the utility model is as follows:A kind of target identification tracking system of unmanned boat
System, including data acquisition unit, perception unit, data processing unit and main control unit;
The data acquisition unit includes 3D laser radars, transmitting transducer, a pair of of hydrophone and integrated navigation system, institute
It includes GPS navigation system and IMU Inertial Measurement Units to state integrated navigation system;The transmitting transducer and a pair of of hydrophone with
Hydrophone processor connects;
The 3D laser radars are mounted in the middle part of unmanned boat top platform;Transmitting transducer is mounted on unmanned boat lower central
Position is less than the draft of floating drum;A pair of of vector sound wave hydrophone is mounted on the inside of unmanned boat lower buoy;GPS navigation system
System and IMU Inertial Measurement Units are mounted in the caisson of unmanned boat;
The perception unit includes 3D laser radars and two wide angle cameras;The 3D laser radars and data acquisition unit
In 3D laser radars share;Two wide angle cameras are separately mounted at left and right sides of unmanned boat front platform;
The data processing unit is the industrial personal computer of four cores, is mounted in the caisson of unmanned boat, and be equipped with large capacity
Internal storage;
The main control unit, which uses, passes through wireless data sending figure transmission module between STM32F4 processors, with bank base host computer
It is communicated.
Further, the hydrophone is vector sound wave hydrophone.
Further, the integrated navigation system is the integrated navigation system based on MEMS technology.
Compared with prior art, the utility model has the advantages that:
1, the utility model is scanned underwater surface barrier using 3D laser radars, vector hydrophone, industrial personal computer
Water hazard object location is parsed, hydrophone processor determines underwater obstruction object location, according to the Obstacle Position of acquisition, industry control
Machine takes obstacle avoidance algorithm to carry out avoidance to barrier, that is, realizes unmanned boat to the barrier avoidance of underwater surface, having reached can
The waters of situation complexity carries out the purpose of operation under water.
2, the target identification tracing system that devises unmanned boat of the utility model on the basis of unmanned boat automatic obstacle avoiding,
It realizes the identification to water surface specific objective and tracking task, unmanned boat is made to have autonomous judgement;In traditional visible sensation method
On the basis of use 3D laser radars and wide angle camera image fusion technology, reduce the image-region of wrong report, improve calculating
The efficiency of machine vision.
Specific implementation mode
The utility model is further described through below in conjunction with the accompanying drawings.As shown in figures 1 to 6, a kind of target of unmanned boat
Identify tracing system, including data acquisition unit, perception unit, data processing unit and main control unit 10;
The data acquisition unit includes 3D laser radars 5, transmitting transducer 2, a pair of of hydrophone 1 and integrated navigation system
System, the integrated navigation system includes GPS navigation system 11 and IMU Inertial Measurement Units 12;The transmitting transducer 2 and a pair
Hydrophone 1 is connect with hydrophone processor 9;
The 3D laser radars 5 are mounted in the middle part of unmanned boat top platform;Transmitting transducer 2 is mounted in unmanned boat lower part
Position is entreated, the draft of floating drum is less than;A pair of of vector sound wave hydrophone 1 is mounted on the inside of unmanned boat lower buoy;GPS navigation
System 11 and IMU Inertial Measurement Units 12 are mounted in the caisson 7 of unmanned boat;
The perception unit includes 3D laser radars 5 and two wide angle cameras 3;The 3D laser radars 5 are acquired with data
3D laser radars 5 in unit share;Two wide angle cameras 3 are separately mounted at left and right sides of unmanned boat front platform;
The data processing unit is the industrial personal computer 8 of four cores, is mounted in the caisson 7 of unmanned boat, and be equipped with large capacity
Internal storage;
The main control unit 10 passes mould using between STM32F4 processors, with bank base host computer by wireless data sending figure
Block 4 is communicated.
The function of each device is as follows:
Geographical location and unmanned boat posture information of the data acquisition unit based on integrated navigation system to unmanned boat
It is acquired, GPS navigation system 11 obtains the real-time position information of unmanned boat, and IMU Inertial Measurement Units 12 obtain unmanned boat
Real time acceleration, speed and course angle, transmitting transducer 2 reflect underwater obstacle to underwater emission acoustic signals, hydrophone 1
The acoustic signals returned are received to determine the orientation of barrier;
The wide angle camera 3 is acquired unmanned boat image information in front;
The 3D laser radars 5 perceive unit and are shared with data acquisition unit.3D laser radars 5 are adopted for data acquisition unit
Collect peripheral obstacle information, and is used for avoidance;3D laser radars 5 are that perception unit is detected the front water surface, obtain detection
It identifies target, is merged with wide angle camera 3 with realizing;
The bank base host computer receives digital signal, picture signal and the water surface acquisition that unmanned boat is sent back in the water surface
Information, when unmanned boat independently judges to go wrong, the telecontrol antenna 6 of unmanned boat receives bank base host computer and sends instruction,
It controls the work of unmanned boat or makes a return voyage.
The industrial personal computer 8 carries out the data of 3D laser radars 5 location information that parsing obtains barrier, and and hydrophone
The location information of 1 underwater obstacle navigated to is merged, according to the environmental information of acquisition and the real-time attitude of unmanned boat
Real-time environmental modeling is carried out, area of feasible solutions is chosen;Visual information is handled using industrial personal computer 8, by 5 figure of 3D laser radars
Picture and 3 image of wide angle camera are merged, and image procossing is done to blending image, extract area-of-interest, and acquisition tracks and identifies mesh
Mark;Industrial personal computer 8 needs the larger memory headroom occupied after carrying out visual processes and the parsing of 5 data of 3D laser radars, therefore needs
The internal storage of large capacity.
The hydrophone processor 9 uses STM32F4 series processors, the signal transmitting of control transmitting transducer 2 to be added
The signal that location algorithm receives hydrophone 1 does algorithm process, realizes to underwater obstacle ranging localization.
The main control unit 10, as bottom control, is controlled unmanned boat motor using STM32F4 series processors
System, the area of feasible solutions obtained according to data processing unit and identification tracking target, realize unmanned boat the automatic obstacle avoiding on the basis of
Target identification is tracked.
Further, the hydrophone 1 is vector sound wave hydrophone.
Further, the integrated navigation system is the integrated navigation system based on MEMS technology.
The working method of the utility model, includes the following steps:
A, data initialization
Initialize bank base host computer operating system and unmanned boat device driver.
B, data information acquisition
IMU Inertial Measurement Units 12 obtain the real-time attitude of unmanned boat, and the real-time attitude includes acceleration, speed and boat
To angle;GPS navigation system 11 obtains unmanned boat real time position;3D laser radars 5 are scanned front three dimensions, before acquisition
Obstacle information on the square water surface;A pair of of vector hydrophone 1 is based on digital method to underwater obstruction by acoustic signals
Object carries out positioning distance measuring;Data transmission is realized by I/0 mouthfuls, RS232 serial communication protocols, at data information transfer to data
Manage unit;
C, visual information acquires
Acquisition of the unit based on the realization of camera communication protocols to image information is perceived, and acquisition image information is passed through
The connection of USB2.0 standards is transmitted to industrial personal computer 8;Wide angle camera 3 has the visual angle of bigger, obtains larger range of Surface Picture, 3D
Laser radar 5 scanned while realizing avoidance in front of waters, obtain detection target, the scan data of 3D laser radars 5 passes through
Serial communication reaches data processing unit;
D, data processing
Industrial personal computer 8 parses the radar data scanned, obtains the depth information of the barrier on sea, determines sea
The coordinate of barrier on face, hydrophone 1 receive the acoustic signals that barrier is reflected back, by signal processing circuit by hydrophone
1 acoustic signals received are converted into digital signal and are sent to hydrophone processor 9, and hydrophone processor 9 is fixed based on the time difference
Position algorithm determines the Azimuth & Range of underwater underwater obstacle;
Industrial personal computer 8 is according to the collected unmanned boat real-time attitude of data acquisition unit and location information and 3D laser radars
5 and 1 collected water hazard object information of hydrophone to the front water surface carry out environmental modeling, led based on GPS during avoidance
Boat system 11 realizes global path planning, according to the collected obstacle information of data acquisition unit, is based on vector field histogram
Obstacle avoidance algorithm carries out part avoidance, that is, VFH*, realizes local avoidance and the front prediction of unmanned boat, the method based on intensified learning
Optimize the avoid-obstacle behavior of unmanned boat;
The processing of visual information is completed in the industrial personal computer 8, and wide angle camera 3 establishes ideal pin-hole model, and industrial personal computer 8 is by 3D
Laser radar 5 and 3 image co-registration of wide angle camera, extract image outline, for the scanning on 3D radar images from bianry image
The region arrived carries out visual processes, carries out feature extraction for interested region, determines the target of recognition and tracking, is used in combination visual
The green circle changed is got up;
E, bottom control
Data processing unit is by the water hazard object information that industrial personal computer 8 is handled and the underwater barrier that hydrophone processor 9 is handled
The information of object is hindered to be transmitted to main control unit 10, main control unit 10 is based on Kalman filtering by the target identification tracked information of vision
It is merged with environmental modeling information, controls the real-time attitude of unmanned boat, when programming, to evading than vision for barrier
Target identification tracking has higher priority.
It is illustrated in figure 2 the logical construction block diagram of the data processing of hydrophone 1, number is executed using hydrophone processor 9
Word signal acquisition and Digital Signal Processing take TTL to be connected in series, hydrophone between transmitting transducer 2 and hydrophone processor 9
The frequency of the control transmitting pulse signal of processor 9, once emits pulse number, width at the repetition period.Hydrophone processor 9 is controlled
Power amplifier processed emits amplified signal, and power amplifier is connect with transmitting transducer 2, and transmitting transducer 2 is to underwater emission sound
Wave signal, the acoustic signals that barrier is reflected back are received by hydrophone 1, are located at a pair of of vector hydrophone 1 of unmanned boat both sides
Successively receive two acoustic signals, the acoustic signals that transmitting transducer 2 emits meeting during being received back next by hydrophone 1
Be weakened, and be received back in the acoustic signals come and will include other underwater noises, i.e., the acoustic signals of energy converter transmitting by
To damage, so a prime amplifier has been used, the impaired acoustic signals that processing hydrophone 1 receives, to impaired
Acoustic signals are amplified, and are filtered to unnecessary frequency.Hydrophone processor 9 by take digital method come
Determine the azimuth-range of barrier.It is connected in series by TTL and the orientation of barrier is mapped to main control unit 10.
The prime amplifier includes being not intended to for reading instrument amplifier from 1 signal of hydrophone and for filtering out
The fixed frequencies such as the bandpass filter, such as propeller, motor of frequency signal signal.
Digital method, that is, the DTOA is estimated by measuring the time difference that two hydrophones 1 receive signal
Range difference, and then barrier is positioned.
It places obstacles and the coordinate of object point is hindered to be (x, y, 0), the coordinate of two receiving hydrophones 1 is respectively (x1,y1,z1)、(x2,y2,
z2), time difference of second signal and first signal that measure is t, then second hydrophone 1 and first hydrophone 1
Transmission distance deviation RS1=RS-R1, can be established an equation according to transmission distance deviation and acquire the position coordinates (x, y, 0) of barrier.
Rs:The distance between i-th point and barrier;
(xiyi,zi):The coordinate of i-th of hydrophone 1;
It willAbove formula is substituted into, is enabledxsi=xs-
xiAbbreviation obtains
X, y-coordinate, the coordinate of acquired disturbance object are solved using above formula.
It is illustrated in figure 3 the flow chart of processing of vision, the process of the visual processes includes radar image and camera
The fusion of image obtains target template, pyramid algorith sample detecting, the matching of moving target, motion estimate tracking.
The method that the process of the visual processes uses a variety of basic image procossings, including noise reduction, medium filtering, figure
Image intensifying, edge detection, feature extraction etc.;
The 3D laser radars 5 and the image fusion technology of wide angle camera 3 are by interested point in 3D laser radars 5
Position is converted to the coordinate of 3 pixel of wide angle camera, and by the way that the point of 3D laser radars 5 is converted to Descartes's XYZ coordinate, 3D is swashed
The origin of optical radar 5 is converted to the origin of wide angle camera 3, and calculates pixel value by the intrinsic model of wide angle camera 3 to complete
, camera has higher resolution ratio, realizes mapping of the depth point to frame of pixels.The depth information of desired region is carried out at threshold value
Reason, and connected remaining point to form continuum using morphological operation.3D laser radars 5 melt with 3 image of wide angle camera
Conjunction technology realizes the frame of pixels that the co-ordinate system location of object is transformed into wide angle camera 3, thus by laser radar detection object,
Interested region is extracted, reduces the image-region that may be reported by mistake, improves the efficiency of vision system.
The utility model is not limited to the present embodiment, any equivalent concepts in the technical scope that the utility model discloses
Or change, it is classified as the scope of protection of the utility model.