CN106327477A - Ice flood observation system based on technology of computer vision - Google Patents
Ice flood observation system based on technology of computer vision Download PDFInfo
- Publication number
- CN106327477A CN106327477A CN201610775730.8A CN201610775730A CN106327477A CN 106327477 A CN106327477 A CN 106327477A CN 201610775730 A CN201610775730 A CN 201610775730A CN 106327477 A CN106327477 A CN 106327477A
- Authority
- CN
- China
- Prior art keywords
- image
- ice
- data
- computer vision
- system based
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
Abstract
The invention discloses an ice flood observation system based on the technology of computer vision. A system which consists of an observation device, data processing equipment and terminal equipment is designed, wherein the observation device and the data processing equipment are in wireless signal connection. Moreover, a data collection and processing method based on the above system is also designed, and the method comprises the steps: automatically collecting image data through a camera mechanism, and transmitting the image data to the data processing equipment through a wireless signal; carrying out the image space conversion of the image collected by the camera mechanism through the data processing equipment, thereby achieving the quantitative analysis of ice flood distribution image data of a river. The beneficial effects of the invention are that the structure is simple; the design is reasonable; the system reads the ice flood data of engineering observation points in a region from one display according to the collection time; the system is high in data reading precision, is high in timeliness, is convenient and quick, can save a large amount of observers, alleviates the labor intensity, and eliminates the personal potential safety hazards.
Description
Technical field
The present invention relates to hydrological monitoring device, particularly relate to a kind of ice flood observation system based on computer vision technique.
Background technology
Northern China Areas along Yellow River, winter severe cold, special Lower Reaches of The Yellow River people across the Straits are blocked the flood caused deeply by ice slush
The hardship of overbank, dredging river, preventing ice flood from causing disaster is the unremitting pursuit of the people of harnessing the Yellow River over the years, therefore, carries out ice slush sight in winter
Survey is to grasp one of river course important means insulting the phase change firsthand information and preventing and treating flood, and ice slush observation is typically all and is dropping down
Ling Qi, freezing period of expansion, pay close attention in three periods of the phase of constructing a canal, for guaranteeing that the Yellow River spends ice flood in comfort, the personnel that reduce the menace of ice run are closed closely
Note ice slush change, the Ling Qing change in each observation station staff round-the-clock close attention river course, the narrowest river course and bend etc.
Dangerous situation section easily occurs, in terms of river course river surface width, stream Ling Midu, ice thickness, floating ice area, shore ice situation etc., is observed note
Record, and these work all lean on observation personnel be observed record by rule of thumb with eye estimating method and calculate, the same day, the data of observation were carried out
Arranging and report and will write by armrest, the arduous and inefficient work of these tradition has been not suitable with the electronic information of Modern High-Speed development
Technology requirement, and this eye estimating method error of observation data is relatively big, takies personnel many, the observation inconvenience of wind and snow sky, there are personnel not
Safety factors etc..
Summary of the invention
The present invention be directed to the deficiency existing for prior art, and provide a kind of simple in construction, reasonable in design, save
Cost of labor, provide again work quality and work efficiency a kind of based on computer vision technique ice flood observation system and
Image data acquiring and processing method.
The technical scheme is that and devise the observation device and data handling equipment connected by wireless signal, and
The system of terminal unit composition;And devise data acquisition based on said system and processing method, automatic by image mechanism
Gather view data, and by view data transmission of wireless signals to data handling equipment, then data handling equipment is to shooting
The image that mechanism gathers makees image space conversion process, it is achieved the quantitative analysis to river course ice slush distributed image data, at image
In process, use based on sub-pix vision technique frozen water partition method, effectively reduce threshold value and select difficulty.Analysis and research conventional images
Partitioning algorithm, in conjunction with ice slush feature of image, for the on-site actual situations in river course, write be applicable to ice slush image procossing based on
Opencv ice slush parser, utilizes the technical researches such as contour detecting, rim detection, image gray processing, image two-value differentiation
Ice slush and river part in the ice slush view data of system acquisition can be split by the realization of the clear parser of river surface ice, it is achieved
To river course ice slush density, maximum ice cube, real-time, the science of ice thickness, effectively observe.Wherein, entire image is being ensured
Entirety and the color of local and the distribution of brightness degree and feature on the premise of, the amount of calculation of gray-scale map becomes few.For
Whole image presents obvious black and white effect, needs to be carried out by picture binary conversion treatment, namely by the point on image again
Gray value be set to 0 or 255.Then image is carried out edge detection algorithm process, by arranging parameter threshold and matrix, adopt
General operator is drawn to carry out edge treated with sobel operator with Gauss.According to the result of rim detection, carry out contour detecting.
Specifically, a kind of ice flood observation system based on computer vision technique, including observation device and terminal unit, institute
Stating observation device and had data handling equipment by wireless signal connection, described data handling equipment signal connects described terminal unit;
Described observation device includes that new energy device, described electric supply installation are electrically connected with data acquisition unit and signal transmitting and receiving
Device;
It is described that described data acquisition unit includes that image mechanism and cradle head mechanism, described image mechanism and cradle head mechanism signal connect
Signal receiving/transmission device.
Wherein, described cradle head mechanism includes internally-arranged type DC controller and hemispherical protective cover, described hemispherical protective cover
Inside it is provided with described image mechanism;
Described controller includes the cradle head control module for controlling cloud platform rotation, and described cradle head control module by signal connects two
Platform motor is respectively used to realize the rotation of The Cloud Terrace left and right directions and above-below direction;
Also include for limiting the vertical limited bolt of cloud platform rotation scope, Horizontal limiting bolt and two cooperations described the most spacing
Bolt, the microswitch of Horizontal limiting bolt.
Described solaode includes 30-40 block solar battery sheet.
Described radio transmission apparatus uses cellular mobile communications networks.
Described observation device also includes the time relay.
The image data acquiring of a kind of ice flood observation system based on computer vision technique and processing method, its feature exists
In, comprise the following steps:
Step 1: use image mechanism to gather image;
Step 2: the image collected is sent to data handling equipment;
Step 3: data handling equipment, by the data being made up of 3 components red, green and blue of point each in image, turns
It is changed to the data of gray scale;
Step 4: gradation of image equalization processing;
Step 5: by gaussian filtering method noise reduction process;
Step 6: carrying out image threshold segmentation;
Step 7: calculate ice slush density and the area of maximum ice cube.
Step 8: by temperature, ice slush density and maximum ice cube areal calculation ice thickness.
New energy device uses solar storage battery electric power system.Owing to system design goal is out of office for being capable of
Outer Yellow River Channel long-term stable operation, and these locations do not have the power supplys such as civil power to come for system power supply more, so system design choosing
Select new energy device to provide power supply for system long-term stable operation.River course ice condition ice slush image monitoring system is operated in Huang
Bank, river course, river, it is contemplated that the situation of river course unpowered and the field condition of river course abundance at open sunshine, choosing during system design
It is system power supply by solar storage battery electric power system.I.e. ice hydrological telemetry system and cradle head control ice slush image capturing system is equal
By storage battery power supply, and control solar panels to accumulator charging by controller for solar.Solar panel is by solar energy
Cell piece (being equivalent to the battery of joint an about 0.5V) forms, and is to utilize photovoltaic generation principle to convert the solar into electric energy, continues
And be stored in accumulator or make its work for load supplying.In solar panel, the quantity of cell piece is come certainly by system voltage
Fixed, typically take 1.4-1.5 times of system voltage.The quality of whole solar electric power supply system and cost are largely by solar energy
The quality of cell panel and cost determination, and the number of solar battery sheet decides the price of solar panel.Thus be
The requirement of system long-term stable operation in the wild should be disclosure satisfy that during system design during solar panel type selecting, meet too again
The sun energy cell panel i.e. solar battery sheet of area minimum is minimum so that its price is minimum.
Transmission of wireless signals is used by the frozen water feelings of collection in worksite, the timing of ice slush view data or to send in real time to monitoring
Center.The master-plan structure of system, totally sees and the Yellow River field data collecting part and Surveillance center's information can be divided into receive
Process and control two parts.Field data collecting part completes the collection of ice condition ice slush view data and sends, Surveillance center's information
Reception processes and processes with the reception and later stage controlling the responsible data of part, connects the communication that two-part tie is exactly data, because of
And the selection of communication network is particularly important, it is to ensure that field data real-time stabilization passes the basis of Surveillance center back.Consider
System is operated in the field that nobody stands fast at, so wireless communication technology must be selected to complete the transmission to system data.
Native system uses cloud computing mode, and the Yellow River field data collecting part is exactly observation device part, and Surveillance center believes
Breath reception process is exactly data handling equipment and terminal unit with control part.The outdoor running and comparing of image mechanism is ripe steady
Fixed, fault be generally present in as data process by industry control processor on, as used cloud computing mode to be omitted from strange land dimension
The work protected.Cloud computing mode is industry control processor to be put back to office as cloud computing center, the shooting of each control point
Mechanism is as acquisition terminal, after each control point image mechanism has gathered image, beams back cloud computing center in real time and is analyzed meter
Calculate and draw data.The most not only save strange land maintenance work, and save each control point and industry control processor is installed
Expense, reaches maximizing the benefits.
Transmission of wireless signals uses 3G transport module.Different according to communication distance, communication mainly has VHF, penetrates
Frequently RF, cellular communication (GSM, GPRS, 3G) and satellite communication, the communication distance of different communication mode is different, wherein VHF and
RF has the restriction of farthest communication distance, and also has the most harsh requirement to working environment, easily by landform and weather
Impact.It is littoral that observation device is operated in Yellow River Channel, and Monitoring Data to be passed back apart from far data handling equipment, its distance
It is far longer than 40Km, thus VHF and RF radio communication mode all can not meet system work requirements.Review communication distance unrestricted
The satellite of system, the rate of its 0.6 yuan/SOB are the hugest for the view data of youngster KB at least, are that system cost is not caned
Bear, thus satellite communication is not the most at the row of selection.Through on-the-spot investigation, HUANGHE ESTUARY monitoring station covers model in Cellular Networks
In enclosing, cellular mobile communications networks (3G) is therefore used to become the first-selection of this ice slush image monitoring system design.
Original image data are acquired, process and computing by observation device, and by initial data and process data storage
In SD card, send order remotely control The Cloud Terrace by laboratory Surveillance center host computer and realize the rotation in direction, upper and lower, left and right
Turn and camera lens is realized the operations such as zoom, zoom, aperture opening and closing.Thus realize comprehensive many to Yellow River Channel ice slush image
Angle monitor.Observation device is selectively mounted to away from river surface limit, river course nearby when installing, and before lock, pumping plant is ideal peace
Dress place, not only can guarantee that safety but also is easy to operation maintenance, must be fetched by instrument container in advance before constructing a canal.Lower margin is passed through during installation
The support of photographic head and instrument container is first fixed by screw, then will photographic head and instrument container be fixed on support, such one again
Come, both achieved the long-time steady operation of system, also ensure that to a certain extent instrument container and photographic head safety.
The image mechanism running of observation device is wanted to meet the remote control and regulation monitoring of whole river cross-section multi-angle;
Also want to carry out with the change of monitored area, river course image scene definition the control of remotely focusing of focal length, be therefore provided with The Cloud Terrace machine
Structure, is installed on image mechanism in cradle head mechanism, sends remote control command by Surveillance center's terminal unit, passes through 3G network
Deliver to on-the-spot ice slush image data acquiring device, control cradle head mechanism and drive about image mechanism realization, rotate up and down, and take the photograph
The far and near regulation of camera structure focal length, thus realize the on-demand monitoring of far and near multi-angle to river cross-section ice slush image.
Cradle head mechanism is the support equipment fixing, installing image mechanism, can be divided into different by the difference of criteria for classification
Type.It is divided into indoor type and outdoor version according to using environment difference.Ball-type and plain edition can be divided into again, wherein according to profile difference
Spherical pan head is to be placed in video camera etc. in a hemispherical protective cover to prevent dust from disturbing camera lens.According to running voltage not
With being divided into direct current height variable speed model and exchanging constant speed type.It is divided into internally-arranged type and externally positioned type according to the installation of cradle head controllor.
Native system data acquisition unit is operated in Yellow River Channel bank, field, uses solar storage battery electric power system to supply
Electricity, thus exchange constant speed The Cloud Terrace can not be selected, and it is capable of remotely controlling of The Cloud Terrace and remotely focusing of camera lens.In addition
Relative to controller externally positioned type The Cloud Terrace, internally-arranged type lead-out wire is less, cabling the most on circuit boards, system run all right.To sum up, it is
System is designed as the built-in spherical pan head of DC controller.Cradle head control mainly includes motor control and the big module of lens control two.Mirror
Head controls the parameter of major control camera lens, is subdivided into zoom, zoom, aperture etc..And motor control is subdivided into knee-action control
System, left and right action control.
Cradle head control main circuit The Cloud Terrace to be realized, in left and right directions, the rotation of above-below direction, can enter electricity by two
Machine realizes this function.Entering motor is a kind of electromagnetic and mechanical device, and electric pulse can be converted into corresponding displacement of the lines or angular displacement.It
Start and stop are quick, and when motor dynamics torque is more than or equal to motor load, flashy the starting or stoping of motor can be by input
Pulse controls.Enter the rotating speed of motor and elongation be general only by input pulse frequency influence, and with ambient pressure, temperature,
The factors such as vibration, load change and voltage pulsation are unrelated.Therefore, motor may be used for the control in cloud platform rotation direction,
Thus the control to cloud platform rotation just becomes the control entering motor working method.
In the design, for controlling the scope of cloud platform rotation, the angle limit that The Cloud Terrace is vertical, horizontally rotate i.e. is controlled, firmly
Consider during part design that design is vertical, Horizontal limiting bolt, realized vertically by two microswitch respectively, Horizontal limiting function.
When the horizontal or vertical rotational angle of The Cloud Terrace reaches to set the limit, touching limited bolt microswitch, microswitch disconnects, and cuts off electricity
Source, The Cloud Terrace is not rotated further by this direction.So far the design on cradle head control point road is completed.
Now timing with can the ice slush automatic image data collecting of favored area be limited by system power dissipation, ice slush image uses fixed
Time acquisition mode, set 8:00,12:00,14:00,16:00 period every day as ice slush image acquisition time.Section between at this moment, be
Uniting to photographic head, 3G wireless video module and complete ice slush image acquisition task, remaining period system is in low-power consumption.With
Time, send order by host computer and realize the long-range control of The Cloud Terrace, it is achieved the ice slush image monitoring to selection area.
During cradle head control ice slush Design of Image Acquisition System, for ensureing that system is under solar storage battery powering mode
Long-time stable works, and needs farthest to consider to reduce system power dissipation, so during hardware circuit design, design adds relay
Device control circuit, enters motor, video camera and 3G wireless video module regularly to The Cloud Terrace in system and powers on every day, other times
System is in resting state, reduces system power dissipation with this.
Observation device is provided with mastery routine and Interruption power-up routine.
In order to realize the quantitative analysis to river course ice slush distributed image data, first the image of camera collection is made image
Space conversion and noise reduction process.Being uneven owing to on-the-spot illumination changes over, photographic head is even during working long hours
You there will be under-exposed phenomenon so that target object ice slush and the brightness irregularities of background river in image, when causing segmentation
The selection difficulty of threshold value.Therefore, in Image semantic classification, use based on sub-pixel precision contour detecting vision technique, effectively drop
Low profile selects difficulty.Effectively achieve ice by carrying out image threshold segmentation technology, water separates, and completes the calculating of ice slush data.
Image semantic classification mainly includes the processes such as greyscale image transitions, gray balance and image filtering.Image capture device
In the view data obtained, the data of each point are made up of red (R), green (G) and blue (B) 3 components, for ease of
Follow-up process and raising processing speed, 3 components of each point are converted into gray scale (I) information, pass through by gradation conversion process
Use based on sub-pixel precision contour detecting technology, improve picture contrast, containing a lot of in image after above-mentioned process
Noise spot, has had a strong impact on successive image segmentation effect, it is necessary to filter as far as possible, uses gaussian filtering method herein, effectively rejects
Noise.
After Image semantic classification, solve brightness irregularities phenomenon, thus reduce the difficulty of follow-up threshold value
The process of Threshold segmentation image can describe with following math equation:
(1)
In formula (1), g (x, y) represent image segmentation result, f (x, y) be pending image herein refer to obtain ice slush image, T is
Threshold value, any meet f (point of x, y) >=T is referred to as object-point, is the ice slush part in image herein, other point then be referred to as background
River part in point, referred to image.From formula (1), ice to be realized, water properly separate, and choose suitable segmentation threshold T
It it is key value.Selecting iterative method, step is summarized as follows:
1. selecting initial threshold T (j), the average gray value of general image generally can be selected as initial threshold, j is for repeatedly
Generation number.J=0 time initial.
2. image is split with T (j).Image is divided into 2 regions C1 (j) and C2 (j).
3. the average gray value in two regions is calculated.Wherein N1 (j), N2 (j) are the pixel of iteration j time domain C1 and C2
Number.(x y) represents (x, y) gray value put in image to f.
(2)
(3)
Calculate new threshold value the most again, it may be assumed that
(4)
5. j=j+1 repetitive (2)-(4) are made until the difference of T (j+1) and T (j) is less than setting.As seen from Figure 3, top
Partial image is that image directly carries out Threshold segmentation design sketch without pretreatment, owing to brightness of image is uneven, causes image to divide
After cutting, ice, water are lumped together, are not carried out properly separating of frozen water.Lower part image effect is that image first passes through pretreatment, so
After carry out the design sketch of image segmentation, it can be seen that after image binaryzation, it is achieved that ice, water perfection separate, for follow-up ice slush number
Just data is provided to support according to statistics.
Algorithm based on opencv.The picture of photographic head shooting, carries out the gray-value image process of 8, is processed into gray scale
Figure.On the premise of the entirety of guarantee entire image with the color of local and the distribution of brightness degree and feature, the meter of gray-scale map
Calculate quantitative change must lack.In order to whole image presents obvious black and white effect, need picture to be carried out binary conversion treatment, also again
It is exactly that the gray value of the point on image is set to 0 or 255.Then image is carried out edge detection algorithm process, by arranging
Parameter threshold and matrix, use sobel operator and Gauss to draw general operator to carry out edge treated.Process knot according to rim detection
Really, contour detecting is carried out.
Profile calculates and is divided into both of which: (1) only retrieves the profile of ragged edge, and they are organized as two-layer.(2) retrieval
All of profile, and reconstruct the whole level of nested profile.A total of how many ice cubes can be calculated by the calculating of profile,
The shape of each ice cube, the area of ice cube, the thickness of ice cube, the density of ice cube, the maximum area of ice cube, and bergy is big
Little, and the percentage ratio of whole river surface width, and by incoming cloud computing platform real-time for GPRS, cloud computing platform is according to up-to-date
Monitoring situation, carries out big data analysis, chain rate and carry out data comparison and analysis on year-on-year basis.
According to the threshold values arranged, carry out Business Processing.After these Business Processing, cloud computing platform can Based Intelligent Control set
Standby duty.Such as: shoot specific, increasing parameter value and threshold values, definite obtains current river surface situation, Yi Jilian
The photographic head that dynamic whole river surface is deployed troops on garrison duty, carries out combining taking pictures, and carries out the data analysis of overall river surface, with and according to weather information,
Historical information and algorithm model, produce form and warning message, in order to accomplish that ice condition finds in time, quickly processes.
Algorithm above solves and calculates ice slush density and the method for maximum ice slush, require to report in flood control office
In also have " ice thickness " data.The development of ice thickness, change have certain rule within a period of time, and its quantitative factor is mainly gas
Temperature, temperature is to affect the factor that ice thickness is most sensitive.In same section, when other conditions are identical, temperature is the lowest, accumulation angrily temperature
Lasting the longest, ice thickness value is bigger, and ice thickness also exists certain functional relationship with accumulation angrily temperature.Meanwhile, the density of ice slush and
Maximum ice cube and ice thickness are proportional, and density is the biggest, and ice cube is the thickest.Ice cube is the biggest, and thickness is the biggest.Ice thickness according to impact is fixed
The analysis of amount factor, can set up ice thickness function expression is:, in formula, d is ice thickness, and m is ice slush density;K is maximum
Ice cube;T is temperature, DEG C, the i.e. stable turn of per day angrily temperature of negative accumulation.
By field data analysis for many years, screening, ice thickness d the get involved impact of day of year averagely angrily temperature t in river course is relatively big, for
This, ice thickness formula can be expressed simply as:
In formula, k is ice thickness coefficient;For from daily mean temperature in autumn stably turn negative from the per day angrily temperature of accumulation calculated, DEG C.
Wherein, k can be painted by point to survey ice thickness d and survey accumulation the warmest per day graph of a relation analysis and determine (slope), and t is from net
The temperature that network obtains.
The beneficial effect of this programme can be learnt according to the narration of such scheme, and simple in construction is reasonable in design, according to collection
Time reads local set engineering values measuring point ice slush observation data on a display, have reading data precision high, time
Intersexuality is strong, convenient and swift, can save substantial amounts of observation personnel, reduce labor intensity, and eliminates personal safety hidden danger, and efficiency is better than people
Work, saves human and material resources, financial resources, and this equipment investment is few, instant effect, improve work efficiency.This project implementation, can be in real time
Recognize the truth that ice slush changes, provide the firsthand information for grasping ice slush dangerous situation in time, effectively promote digitized yellow
River construction process, improves the level of scientific and technical innovation development.There is provided scientific basis for decision-making of reducing the menace of ice run from now on, be greatly improved the Yellow River
Reduce the menace of ice run level of office automation.
Accompanying drawing explanation
Fig. 1 is cloud computing mode figure of the present invention;
Fig. 2 is observation device main program flow chart of the present invention;
Fig. 3 is that observation device Interruption of the present invention powers on flow chart;
Detailed description of the invention
For the technical characterstic of this programme can be clearly described, below by detailed description of the invention, this programme is illustrated.
A kind of ice flood observation system based on computer vision technique, including observation device and terminal unit, its feature exists
In, described observation device is connected by wireless signal data handling equipment, and described data handling equipment signal connects described terminal
Equipment;
Described observation device includes that new energy device, described electric supply installation are electrically connected with data acquisition unit and signal transmitting and receiving
Device;
It is described that described data acquisition unit includes that image mechanism and cradle head mechanism, described image mechanism and cradle head mechanism signal connect
Signal receiving/transmission device.
Wherein, described cradle head mechanism includes internally-arranged type DC controller and hemispherical protective cover, described hemispherical protective cover
Inside it is provided with described image mechanism;
Described controller includes the cradle head control module for controlling cloud platform rotation, and described cradle head control module by signal connects two
Platform motor is respectively used to realize the rotation of The Cloud Terrace left and right directions and above-below direction;
Also include for limiting the vertical limited bolt of cloud platform rotation scope, Horizontal limiting bolt and two cooperations described the most spacing
Bolt, the microswitch of Horizontal limiting bolt.
Described solaode includes 30-40 block solar battery sheet.
Described radio transmission apparatus uses cellular mobile communications networks.
Described observation device also includes the time relay.
The image data acquiring of a kind of ice flood observation system based on computer vision technique and processing method, its feature exists
In, comprise the following steps:
Step 1: use image mechanism to gather image;
Step 2: the image collected is sent to data handling equipment;
Step 3: data handling equipment, by the data being made up of 3 components red, green and blue of point each in image, turns
It is changed to the data of gray scale;
Step 4: gradation of image equalization processing;
Step 5: by gaussian filtering method noise reduction process;
Step 6: carrying out image threshold segmentation;
Step 7: calculate ice slush density and the area of maximum ice cube.
Step 8: by temperature, ice slush density and maximum ice cube areal calculation ice thickness.
New energy device uses solar storage battery electric power system.Owing to system design goal is out of office for being capable of
Outer Yellow River Channel long-term stable operation, and these locations do not have the power supplys such as civil power to come for system power supply more, so system design choosing
Select new energy device to provide power supply for system long-term stable operation.River course ice condition ice slush image monitoring system is operated in Huang
Bank, river course, river, it is contemplated that the situation of river course unpowered and the field condition of river course abundance at open sunshine, choosing during system design
It is system power supply by solar storage battery electric power system.I.e. ice hydrological telemetry system and cradle head control ice slush image capturing system is equal
By storage battery power supply, and control solar panels to accumulator charging by controller for solar.Solar panel is by solar energy
Cell piece (being equivalent to the battery of joint an about 0.5V) forms, and is to utilize photovoltaic generation principle to convert the solar into electric energy, continues
And be stored in accumulator or make its work for load supplying.In solar panel, the quantity of cell piece is come certainly by system voltage
Fixed, typically take 1.4-1.5 times of system voltage.The quality of whole solar electric power supply system and cost are largely by solar energy
The quality of cell panel and cost determination, and the number of solar battery sheet decides the price of solar panel.Thus be
The requirement of system long-term stable operation in the wild should be disclosure satisfy that during system design during solar panel type selecting, meet too again
The sun energy cell panel i.e. solar battery sheet of area minimum is minimum so that its price is minimum.
Transmission of wireless signals is used by the frozen water feelings of collection in worksite, the timing of ice slush view data or to send in real time to monitoring
Center.The master-plan structure of system, totally sees and the Yellow River field data collecting part and Surveillance center's information can be divided into receive
Process and control two parts.Field data collecting part completes the collection of ice condition ice slush view data and sends, Surveillance center's information
Reception processes and processes with the reception and later stage controlling the responsible data of part, connects the communication that two-part tie is exactly data, because of
And the selection of communication network is particularly important, it is to ensure that field data real-time stabilization passes the basis of Surveillance center back.Consider
System is operated in the field that nobody stands fast at, so wireless communication technology must be selected to complete the transmission to system data.
As it is shown in figure 1, native system uses cloud computing mode, the Yellow River field data collecting part is exactly observation device part,
Surveillance center's information reception process and control part are exactly data handling equipment and terminal unit.The outdoor operating ratio of image mechanism
More mature and stable, fault be generally present in as data process by industry control processor on, as use cloud computing mode will save
Go the work that strange land is safeguarded.Cloud computing mode is industry control processor to put back to office as cloud computing center, each monitoring
The image mechanism of point, as acquisition terminal, after each control point image mechanism has gathered image, is beamed back cloud computing center in real time and is entered
Row analytical calculation also draws data.The most not only save strange land maintenance work, and save each control point and industry control is installed
The expense of processor, reaches maximizing the benefits.
Transmission of wireless signals uses 3G transport module.Different according to communication distance, communication mainly has VHF, penetrates
Frequently RF, cellular communication (GSM, GPRS, 3G) and satellite communication, the communication distance of different communication mode is different, wherein VHF and
RF has the restriction of farthest communication distance, and also has the most harsh requirement to working environment, easily by landform and weather
Impact.It is littoral that observation device is operated in Yellow River Channel, and Monitoring Data to be passed back apart from far data handling equipment, its distance
It is far longer than 40Km, thus VHF and RF radio communication mode all can not meet system work requirements.Review communication distance unrestricted
The satellite of system, the rate of its 0.6 yuan/SOB are the hugest for the view data of youngster KB at least, are that system cost is not caned
Bear, thus satellite communication is not the most at the row of selection.Through on-the-spot investigation, HUANGHE ESTUARY monitoring station covers model in Cellular Networks
In enclosing, cellular mobile communications networks (3G) is therefore used to become the first-selection of this ice slush image monitoring system design.
Original image data are acquired, process and computing by observation device, and by initial data and process data storage
In SD card, send order remotely control The Cloud Terrace by laboratory Surveillance center host computer and realize the rotation in direction, upper and lower, left and right
Turn and camera lens is realized the operations such as zoom, zoom, aperture opening and closing.Thus realize comprehensive many to Yellow River Channel ice slush image
Angle monitor.Observation device is selectively mounted to away from river surface limit, river course nearby when installing, and before lock, pumping plant is ideal peace
Dress place, not only can guarantee that safety but also is easy to operation maintenance, must be fetched by instrument container in advance before constructing a canal.Lower margin is passed through during installation
The support of photographic head and instrument container is first fixed by screw, then will photographic head and instrument container be fixed on support, such one again
Come, both achieved the long-time steady operation of system, also ensure that to a certain extent instrument container and photographic head safety.
The image mechanism running of observation device is wanted to meet the remote control and regulation monitoring of whole river cross-section multi-angle;
Also want to carry out with the change of monitored area, river course image scene definition the control of remotely focusing of focal length, be therefore provided with The Cloud Terrace machine
Structure, is installed on image mechanism in cradle head mechanism, sends remote control command by Surveillance center's terminal unit, passes through 3G network
Deliver to on-the-spot ice slush image data acquiring device, control cradle head mechanism and drive about image mechanism realization, rotate up and down, and take the photograph
The far and near regulation of camera structure focal length, thus realize the on-demand monitoring of far and near multi-angle to river cross-section ice slush image.
Cradle head mechanism is the support equipment fixing, installing image mechanism, can be divided into different by the difference of criteria for classification
Type.It is divided into indoor type and outdoor version according to using environment difference.Ball-type and plain edition can be divided into again, wherein according to profile difference
Spherical pan head is to be placed in video camera etc. in a hemispherical protective cover to prevent dust from disturbing camera lens.According to running voltage not
With being divided into direct current height variable speed model and exchanging constant speed type.It is divided into internally-arranged type and externally positioned type according to the installation of cradle head controllor.
Native system data acquisition unit is operated in Yellow River Channel bank, field, uses solar storage battery electric power system to supply
Electricity, thus exchange constant speed The Cloud Terrace can not be selected, remotely controlling of The Cloud Terrace to be capable of and remotely focusing of camera lens.In addition
Relative to controller externally positioned type The Cloud Terrace, internally-arranged type lead-out wire is less, cabling the most on circuit boards, system run all right.To sum up, it is
System is designed as the built-in spherical pan head of DC controller.Cradle head control mainly includes motor control and the big module of lens control two.Mirror
Head controls the parameter of major control camera lens, is subdivided into zoom, zoom, aperture etc..And motor control is subdivided into knee-action control
System, left and right action control.
Cradle head control main circuit The Cloud Terrace to be realized, in left and right directions, the rotation of above-below direction, can enter electricity by two
Machine realizes this function.Entering motor is a kind of electromagnetic and mechanical device, and electric pulse can be converted into corresponding displacement of the lines or angular displacement.It
Start and stop are quick, and when motor dynamics torque is more than or equal to motor load, flashy the starting or stoping of motor can be by input
Pulse controls.Enter the rotating speed of motor and elongation be general only by input pulse frequency influence, and with ambient pressure, temperature,
The factors such as vibration, load change and voltage pulsation are unrelated.Therefore, motor may be used for the control in cloud platform rotation direction,
Thus the control to cloud platform rotation just becomes the control entering motor working method.
In the design, for controlling the scope of cloud platform rotation, the angle limit that The Cloud Terrace is vertical, horizontally rotate i.e. is controlled, firmly
Consider during part design that design is vertical, Horizontal limiting bolt, realized vertically by two microswitch respectively, Horizontal limiting function.
When the horizontal or vertical rotational angle of The Cloud Terrace reaches to set the limit, touching limited bolt microswitch, microswitch disconnects, and cuts off electricity
Source, The Cloud Terrace is not rotated further by this direction.So far the design on cradle head control point road is completed.
Now timing with can the ice slush automatic image data collecting of favored area be limited by system power dissipation, ice slush image uses fixed
Time acquisition mode, set 8:00,12:00,14:00,16:00 period every day as ice slush image acquisition time.Section between at this moment, be
Uniting to photographic head, 3G wireless video module and complete ice slush image acquisition task, remaining period system is in low-power consumption.With
Time, send order by host computer and realize the long-range control of The Cloud Terrace, it is achieved the ice slush image monitoring to selection area.
During cradle head control ice slush Design of Image Acquisition System, for ensureing that system is under solar storage battery powering mode
Long-time stable works, and needs farthest to consider to reduce system power dissipation, so during hardware circuit design, design adds relay
Device control circuit, enters motor, video camera and 3G wireless video module regularly to The Cloud Terrace in system and powers on every day, other times
System is in resting state, reduces system power dissipation with this.
Observation device is provided with mastery routine and Interruption power-up routine, mastery routine and the flow process of Interruption power-up routine
As shown in Figures 2 and 3.
In order to realize the quantitative analysis to river course ice slush distributed image data, first the image of camera collection is made image
Space conversion and noise reduction process.Being uneven owing to on-the-spot illumination changes over, photographic head is even during working long hours
You there will be under-exposed phenomenon so that target object ice slush and the brightness irregularities of background river in image, when causing segmentation
The selection difficulty of threshold value.Therefore, in Image semantic classification, use based on sub-pixel precision contour detecting vision technique, effectively drop
Low profile selects difficulty.Effectively achieve ice by carrying out image threshold segmentation technology, water separates, and completes the calculating of ice slush data.
Image semantic classification mainly includes the processes such as greyscale image transitions, gray balance and image filtering.Image capture device
In the view data obtained, the data of each point are made up of red (R), green (G) and blue (B) 3 components, for ease of
Follow-up process and raising processing speed, 3 components of each point are converted into gray scale (I) information, pass through by gradation conversion process
Use based on sub-pixel precision contour detecting technology, improve picture contrast, containing a lot of in image after above-mentioned process
Noise spot, has had a strong impact on successive image segmentation effect, it is necessary to filter as far as possible, uses gaussian filtering method herein, effectively rejects
Noise.
After Image semantic classification, solve brightness irregularities phenomenon, thus reduce the difficulty of follow-up threshold value
The process of Threshold segmentation image can describe with following math equation:
(1)
In formula (1), g (x, y) represent image segmentation result, f (x, y) be pending image herein refer to obtain ice slush image, T is
Threshold value, any meet f (point of x, y) >=T is referred to as object-point, is the ice slush part in image herein, other point then be referred to as background
River part in point, referred to image.From formula (1), ice to be realized, water properly separate, and choose suitable segmentation threshold T
It it is key value.Selecting iterative method, step is summarized as follows:
1. selecting initial threshold T (j), the average gray value of general image generally can be selected as initial threshold, j is for repeatedly
Generation number.J=0 time initial.
2. image is split with T (j).Image is divided into 2 regions C1 (j) and C2 (j).
3. the average gray value in two regions is calculated.Wherein N1 (j), N2 (j) are the pixel of iteration j time domain C1 and C2
Number.(x y) represents (x, y) gray value put in image to f.
(2)
(3)
Calculate new threshold value the most again, it may be assumed that
(4)
5. j=j+1 repetitive (2)-(4) are made until the difference of T (j+1) and T (j) is less than setting.As seen from Figure 3, top
Partial image is that image directly carries out Threshold segmentation design sketch without pretreatment, owing to brightness of image is uneven, causes image to divide
After cutting, ice, water are lumped together, are not carried out properly separating of frozen water.Lower part image effect is that image first passes through pretreatment, so
After carry out the design sketch of image segmentation, it can be seen that after image binaryzation, it is achieved that ice, water perfection separate, for follow-up ice slush number
Just data is provided to support according to statistics.
Algorithm based on opencv.The picture of photographic head shooting, carries out the gray-value image process of 8, is processed into gray scale
Figure.On the premise of the entirety of guarantee entire image with the color of local and the distribution of brightness degree and feature, the meter of gray-scale map
Calculate quantitative change must lack.In order to whole image presents obvious black and white effect, need picture to be carried out binary conversion treatment, also again
It is exactly that the gray value of the point on image is set to 0 or 255.Then image is carried out edge detection algorithm process, by arranging
Parameter threshold and matrix, use sobel operator and Gauss to draw general operator to carry out edge treated.Process knot according to rim detection
Really, contour detecting is carried out.
Profile calculates and is divided into both of which: 1, only retrieve the profile of ragged edge, and they are organized as two-layer.2, retrieval institute
Some profiles, and reconstruct the whole level of nested profile.A total of how many ice cubes can be calculated, respectively by the calculating of profile
The shape of individual ice cube, the area of ice cube, the thickness of ice cube, the density of ice cube, the maximum area of ice cube, and bergy is big
Little, and the percentage ratio of whole river surface width, and by incoming cloud computing platform real-time for GPRS, cloud computing platform is according to up-to-date
Monitoring situation, carries out big data analysis, chain rate and carry out data comparison and analysis on year-on-year basis.
According to the threshold values arranged, carry out Business Processing.After these Business Processing, cloud computing platform can Based Intelligent Control set
Standby duty.Such as: shoot specific, increasing parameter value and threshold values, definite obtains current river surface situation, Yi Jilian
The photographic head that dynamic whole river surface is deployed troops on garrison duty, carries out combining taking pictures, and carries out the data analysis of overall river surface, with and according to weather information,
Historical information and algorithm model, produce form and warning message, in order to accomplish that ice condition finds in time, quickly processes.
Algorithm above solves and calculates ice slush density and the method for maximum ice slush, require to report in flood control office
In also have " ice thickness " data.The development of ice thickness, change have certain rule within a period of time, and its quantitative factor is mainly gas
Temperature, temperature is to affect the factor that ice thickness is most sensitive.In same section, when other conditions are identical, temperature is the lowest, accumulation angrily temperature
Lasting the longest, ice thickness value is bigger, and ice thickness also exists certain functional relationship with accumulation angrily temperature.Meanwhile, the density of ice slush and
Maximum ice cube and ice thickness are proportional, and density is the biggest, and ice cube is the thickest.Ice cube is the biggest, and thickness is the biggest.Ice thickness according to impact is fixed
The analysis of amount factor, can set up ice thickness function expression is:, in formula, d is ice thickness, and m is ice slush density;K is maximum
Ice cube;T is temperature, DEG C, the i.e. stable turn of per day angrily temperature of negative accumulation.
By field data analysis for many years, screening, ice thickness d the get involved impact of day of year averagely angrily temperature t in river course is relatively big, for
This, ice thickness formula can be expressed simply as:
In formula, k is ice thickness coefficient;For from daily mean temperature in autumn stably turn negative from the per day angrily temperature of accumulation calculated, DEG C.
Wherein, k can be painted by point to survey ice thickness d and survey accumulation the warmest per day graph of a relation analysis and determine (slope), and t is from net
The temperature that network obtains.
The present invention can pass through or use prior art to realize without the technical characteristic described, and does not repeats them here, certainly,
Described above is not limitation of the present invention, and the present invention is also not limited to the example above, the ordinary skill of the art
Change that personnel are made in the essential scope of the present invention, retrofit, add or replace, also should belong to the protection model of the present invention
Enclose.
Claims (6)
1. an ice flood observation system based on computer vision technique, including observation device and terminal unit, it is characterised in that
Described observation device is connected by wireless signal data handling equipment, and described data handling equipment signal connects described terminal and sets
Standby;
Described observation device includes that new energy device, described electric supply installation are electrically connected with data acquisition unit and signal transmitting and receiving
Device;
It is described that described data acquisition unit includes that image mechanism and cradle head mechanism, described image mechanism and cradle head mechanism signal connect
Signal receiving/transmission device.
A kind of ice flood observation system based on computer vision technique the most according to claim 1, it is characterised in that described
Cradle head mechanism includes internally-arranged type DC controller and hemispherical protective cover, is provided with described video camera in described hemispherical protective cover
Structure;
Described controller includes the cradle head control module for controlling cloud platform rotation, and described cradle head control module by signal connects two
Platform motor is respectively used to realize the rotation of The Cloud Terrace left and right directions and above-below direction;
Also include for limiting the vertical limited bolt of cloud platform rotation scope, Horizontal limiting bolt and two cooperations described the most spacing
Bolt, the microswitch of Horizontal limiting bolt.
A kind of ice flood observation system based on computer vision technique the most according to claim 1, it is characterised in that described
Solaode includes 30-40 block solar battery sheet.
A kind of ice flood observation system based on computer vision technique the most according to claim 1, it is characterised in that described
Radio transmission apparatus uses cellular mobile communications networks.
A kind of ice flood observation system based on computer vision technique the most according to claim 1, it is characterised in that described
Observation device also includes the time relay.
6. one kind by the ice flood observation system based on computer vision technique described in claim 1-5 image data acquiring and
Processing method, it is characterised in that comprise the following steps:
Step 1: use image mechanism to gather image;
Step 2: the image collected is sent to data handling equipment;
Step 3: data handling equipment, by the data being made up of 3 components red, green and blue of point each in image, turns
It is changed to the data of gray scale;
Step 4: gradation of image equalization processing;
Step 5: by gaussian filtering method noise reduction process;
Step 6: carrying out image threshold segmentation;
Step 7: calculate ice slush density and the area of maximum ice cube;
Step 8: by temperature, ice slush density and maximum ice cube areal calculation ice thickness.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610775730.8A CN106327477A (en) | 2016-08-31 | 2016-08-31 | Ice flood observation system based on technology of computer vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610775730.8A CN106327477A (en) | 2016-08-31 | 2016-08-31 | Ice flood observation system based on technology of computer vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106327477A true CN106327477A (en) | 2017-01-11 |
Family
ID=57789139
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610775730.8A Pending CN106327477A (en) | 2016-08-31 | 2016-08-31 | Ice flood observation system based on technology of computer vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106327477A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109660770A (en) * | 2019-01-12 | 2019-04-19 | 田华 | Intelligent scene monitors system and method |
CN109784224A (en) * | 2018-12-28 | 2019-05-21 | 太原理工大学 | It is a kind of for river frost flower feature and the image detecting system of ice sheet Thickness Analysis |
CN109819319A (en) * | 2019-03-07 | 2019-05-28 | 重庆蓝岸通讯技术有限公司 | A kind of method of video record key frame |
CN112365509A (en) * | 2020-11-10 | 2021-02-12 | 山东大学 | Ice early warning method and system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101358863A (en) * | 2008-08-21 | 2009-02-04 | 华中电网有限公司 | Low-temperature ice-coating test method for on-line monitoring device for transmission line |
CN101574567A (en) * | 2009-06-08 | 2009-11-11 | 南京航空航天大学 | Computer vision technique based method and system for detecting and extinguishing fire disaster intelligently |
US20160125586A1 (en) * | 2014-11-03 | 2016-05-05 | The Johns Hopkins University | Ice analysis based on active and passive radar images |
-
2016
- 2016-08-31 CN CN201610775730.8A patent/CN106327477A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101358863A (en) * | 2008-08-21 | 2009-02-04 | 华中电网有限公司 | Low-temperature ice-coating test method for on-line monitoring device for transmission line |
CN101574567A (en) * | 2009-06-08 | 2009-11-11 | 南京航空航天大学 | Computer vision technique based method and system for detecting and extinguishing fire disaster intelligently |
US20160125586A1 (en) * | 2014-11-03 | 2016-05-05 | The Johns Hopkins University | Ice analysis based on active and passive radar images |
Non-Patent Citations (3)
Title |
---|
侯启真: "《机场安全防范技术》", 30 April 2013, 中国民航出版社 * |
陈兵旗: "《实用数字图像处理与分析 第2版》", 2 February 2014, 中国农业大学出版社 * |
马东晓,刘航东,张海波: "黄河冰凌自动观测新认识与探索", 《2016第八届全国河湖治理与水生态文明发展论坛论文集》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109784224A (en) * | 2018-12-28 | 2019-05-21 | 太原理工大学 | It is a kind of for river frost flower feature and the image detecting system of ice sheet Thickness Analysis |
CN109784224B (en) * | 2018-12-28 | 2022-06-24 | 太原理工大学 | Image detection system for analyzing ice flower characteristics and ice cover thickness of river channel |
CN109660770A (en) * | 2019-01-12 | 2019-04-19 | 田华 | Intelligent scene monitors system and method |
CN109819319A (en) * | 2019-03-07 | 2019-05-28 | 重庆蓝岸通讯技术有限公司 | A kind of method of video record key frame |
CN112365509A (en) * | 2020-11-10 | 2021-02-12 | 山东大学 | Ice early warning method and system |
CN112365509B (en) * | 2020-11-10 | 2023-10-27 | 山东大学 | Ice early warning method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112422783B (en) | Unmanned aerial vehicle intelligent patrol system based on parking apron cluster | |
CN108109385B (en) | System and method for identifying and judging dangerous behaviors of power transmission line anti-external damage vehicle | |
CN106327477A (en) | Ice flood observation system based on technology of computer vision | |
CN102252623B (en) | Measurement method for lead/ground wire icing thickness of transmission line based on video variation analysis | |
CN102638013B (en) | The target image identification power transmission state monitoring system of view-based access control model attention mechanism | |
KR101529678B1 (en) | Method for hybrrid solar tracking to maximize photovoltaic power generation, apparatus and photovoltaic blind system using the same | |
CN102156180B (en) | System and method for monitoring and forecasting regional visibility | |
CN103630238B (en) | A kind of shallow lake water surface imaging spectral acquisition system and synchronous automatic monitoring method | |
WO2017193153A1 (en) | Solar power forecasting | |
CN103353952A (en) | Photovoltaic power prediction method based on ground-based cloud chart | |
CN100434883C (en) | Solar energy in-situ detection method and system | |
CN106210613A (en) | A kind of video frequency monitoring method, Apparatus and system | |
CN111079073A (en) | Building three-dimensional solar potential calculation method | |
CN112996090B (en) | Energy-saving management system and method | |
CN109711348A (en) | Intelligent monitoring method and system based on the long-term real-time architecture against regulations in hollow panel | |
CN111766643B (en) | Automatic laser radar switching device based on weather monitoring and operation method | |
CN106770289A (en) | A kind of node for distributed leaf area index hemisphere IMAQ | |
CN107133685A (en) | The Forecasting Methodology and forecasting system of photovoltaic generating system generated energy | |
CN116631158A (en) | Landslide early warning device based on image recognition and computer vision | |
CN102854138A (en) | Visibility measuring system and method based on digital camera shooting method | |
CN112257340A (en) | Control method and control system of photovoltaic cell panel defrosting robot | |
CN103220503A (en) | Wild video surveillance system | |
Zhen et al. | Transmission tower protection system based on Internet of Things in smart grid | |
CN113160311A (en) | Solar cell panel adjusting method and device, sun tracking system and storage equipment | |
CN116192005A (en) | Photovoltaic tracking method, system and medium based on micro space-time scale irradiation prediction |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170111 |