CN110119671A - Underwater cognitive method based on artificial side line visual image - Google Patents
Underwater cognitive method based on artificial side line visual image Download PDFInfo
- Publication number
- CN110119671A CN110119671A CN201910231013.2A CN201910231013A CN110119671A CN 110119671 A CN110119671 A CN 110119671A CN 201910231013 A CN201910231013 A CN 201910231013A CN 110119671 A CN110119671 A CN 110119671A
- Authority
- CN
- China
- Prior art keywords
- pressure difference
- side line
- pressure
- underwater
- artificial side
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01L—MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
- G01L11/00—Measuring steady or quasi-steady pressure of a fluid or a fluent solid material by means not provided for in group G01L7/00 or G01L9/00
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
Underwater cognitive method of the present invention based on artificial side line visual image, it is proposed with from the point of view of vision to lateral-line system, neural network is imported by the pressure difference image generated and carries out class indication, quickly and accurately to carry out flow field perception and obstacle recognition, the application limitation by improving existing side line technology provides good basis for the engineering application of artificial lateral-line system.Method includes following steps: 1) measuring the array of pressure sensors data under different working conditions；2) pressure data for measuring step 1) is arranged in pressure difference matrix；3) the pressure difference matrix that step 2) obtains is depicted as pressure difference RGB image；4) the pressure difference RGB image input convolutional neural networks obtained step 3) carry out Classification and Identification, obtain each parameter of submerged flow field and classify.
Description
Technical field
Artificial side line technology, which is based on, the present invention relates to a kind of submarine navigation device carries out flow field operating condition sensing and obstacle recognition
Method, specifically using visual processes technology carry out data classification and identification, belong to submarine navigation device field of intelligent control.
Background technique
Existing submarine navigation device in carrying out environment sensing and navigation procedure under complex working conditions, generally there are
More apparent limitation, concentrated reflection are perceiving the side such as inaccurate to local environment and aircraft itself posture information
Face.
Underwater environment is perceived using optical system, not only process is cumbersome and also more limited in perceived distance.
Complicated underwater environment to aircraft motion control, environment sensing and in terms of cause huge interference, mesh
Before have using the sound system (such as acoustic Doppler fluid velocity profile instrument, ADCP) such as sonar etc, to make up short distance perception light
The shortcomings that system is applied.It increases although being run to a certain extent to aircraft with work, ADCP is difficult to detect
To local flow field environment and its have the shortcomings that expensive, instrument is heavy and energy consumption is higher, so can not be applied to small-sized
Submarine navigation device.
With constantly improve for global positioning system (GPS), submarine navigation device generally uses GPS to carry out in underwater environment
Navigation and positioning, but under the shielding action of water body meeting so that aircraft is not still able to satisfy reality there are regular hour error
The fast reaction of current situation portion and accurately navigation and positioning.
For the above-mentioned prior art, technological achievement newer at present is the artificial side line cognition technology based on bionic principle.
Side line is the important hydrodynamic force perceptual organ found with fish and amphibian, and fish can be helped to detect and handle respectively
Kind hydrodynamic force situation.Underwater Pressure perception is carried out using lateral-line system, is risen emphatically in fish and amphibian life-form structure
The effect wanted.The bionical side line produced according to above-mentioned animal side line structure and physiologic sensor performance, has been progressively applied to
In submarine navigation device, to improve the sensing capability of underwater partial fluid environment, while navigator fix is provided and is become apparent from
Booster action.
At present for the technical research and application of artificial lateral-line system, need to establish each underwater State corresponding mould
Type, versatility are poor.Meanwhile lacking necessary classified identification function, measured pressure in terms of perception flow field and barrier
Difference data directly, accurately can not identify flow velocity, flow direction, barrier simultaneously, and whether there is or not the relevant informations such as, barrier size.It is another
Aspect is perceived with navigator fix sensitivity, quick-reaction capability not enough in complex environment under water.
In view of this, special propose present patent application.
Summary of the invention
Underwater cognitive method of the present invention based on artificial side line visual image is that solving the above-mentioned prior art deposits
The problem of and to lateral-line system from the point of view of proposition with vision, i.e., by generate pressure difference image import neural network carry out
Class indication, quickly and accurately to carry out flow field perception and obstacle recognition, the application by improving existing side line technology is limited to
And good basis is provided for the engineering application of artificial lateral-line system.
To realize above-mentioned purpose of design, the underwater cognitive method based on artificial side line visual image includes following
Step:
1) the array of pressure sensors data under different working conditions are measured；
2) pressure data for measuring step 1) is arranged in pressure difference matrix；
3) the pressure difference matrix that step 2) obtains is depicted as pressure difference RGB image；
4) the pressure difference RGB image input convolutional neural networks obtained step 3) carry out Classification and Identification, are flowed under water
Each parameter in field is simultaneously classified.
Underwater cognitive method of the application based on artificial side line visual image, improves itself versatility without single
Corresponding modeling, i.e., by the analysis to different pressures difference image, while class indication go out such as flow velocity, flow direction, barrier whether there is or not,
The environmental informations such as barrier size.In the underwater environment of complex working condition, more sensitive perception energy can get with visual angle
Power.
Further, in the step 1), along the artificial side line of submarine navigation device and aircraft head and manually
Side line same level is laid with several sensors on section.
In the step 2), every a line of pressure difference matrix indicates the pressure difference distribution at some time point, row with
It is time difference Δ t between row；In pressure difference matrix, horizontal axis indicates the pressure difference of each adjacent monitoring point, when containing current
The pressure distributed intelligence at quarter；The longitudinal axis indicates time series, contains the time state in flow field.
Pressure difference matrix is generated the data zooming such as following formula during RGB image by the step 3):
Wherein, α indicates limiting factor, and taking α is 256.
In the step 4), the convolutional neural networks have four layers of convolutional neural networks of class Lenet-5, packet
It has included the first convolutional layer (conv1), the second convolutional layer (conv2), full articulamentum (fc1) and softmax layers；
Data carry out evening up operation after the first convolutional layer and the second convolutional layer, then pass through full articulamentum, finally arrive
Up to softmax layers, prediction output is obtained.
To sum up content, the underwater cognitive method based on artificial side line visual image have the advantage, that
1, based on lateral-line system by the way that differential pressure signal is converted into picture signal, i.e., pressure difference is regarded into pixel,
It can be suitable for various submerged applications environment, efficiently avoid respectively each classification or identification scene individually establishes one
The case where kind mathematical model or algorithm.
2, it is based on underweater vision cognitive method, flow field perception and obstacle recognition can be quickly and accurately carried out, to be
Currentlyying propel artificial lateral-line system, flow field perception and navigation provide more advanced technical support under water.
Detailed description of the invention
Fig. 1 is the pressure distribution schematic diagram measured under different flow field working conditions；
Fig. 2 is the schematic diagram for laying sensor；
Fig. 3 is the submarine navigation device structural schematic diagram for being laid with sensor；
Fig. 4 is the schematic diagram that one-dimensional colour band indicates two-dimensional level flow field；
Fig. 5 is the schematic diagram that two dimensional image indicates three-dimensional flow field；
Fig. 6 is pressure difference matrix schematic diagram；
Fig. 7 is the colour band schematic diagram of pressure difference；
Fig. 8 is the pressure difference image of different angle, friction speed；
Fig. 9 is differential pressure data RGB image under different operating conditions；
Figure 10 is convolutional neural networks model；
Figure 11 is pressure difference image pattern building schematic diagram；
Have in Fig. 3, head 1, connection ring 2, pulls pipe 3, flange screw 4, tail portion 5, watertight connector 6, cylinder 7, electricity
Pond 8, battery case 9, mounting plate 10, sensor 11, number adopt circuit board 12.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples.
Embodiment 1, as shown in figures 1 to 6, the underwater cognitive method based on artificial side line visual image includes
Have:
1) the array of pressure sensors data under different working conditions are measured；
2) pressure data for measuring step 1) is arranged in pressure difference matrix；
3) the pressure difference matrix that step 2) obtains is depicted as pressure difference RGB image；
4) the pressure difference RGB image input convolutional neural networks obtained step 3) carry out Classification and Identification, are flowed under water
Each parameter in field is simultaneously classified.
Specifically, as shown in Figure 1, the pressure that the submarine navigation device for being loaded with artificial side line measures under different flow field working conditions
Power distribution schematic diagram.There it can be seen that there is corresponding relationships for pressure distribution (or water velocity distribution) and flow field parameter.
As shown in Fig. 2, being cut along the artificial side line of submarine navigation device and aircraft head and the artificial side line same level
On face, continuously it is laid with several sensors (sensor spacing, which may be considered, is infinitely close to 0).
Under different working conditions, always there is the numerical value of a sensor to be the largest.From this maximum value sensor to
Any one direction is set out, and the numerical value of other sensors is always in decline；And from some minimum value sensor to any one
Direction is set out, and the numerical value of other sensors is always in rising.
If in conjunction with shown in Fig. 3, on the direction that the outer lateral edge side line of aircraft cylinder 7, side line and head extend under water, cloth
If multiple groups sensor 11 is covered with the pressure value acquisition point of aircraft whole body to be formed, so that it may uniquely measure the operating condition ginseng in flow field
Number, includes but are not limited to flow velocity and flow direction.
In conjunction with shown in Fig. 2, Fig. 3 and Fig. 4, for the side line that the sensor of horizontal distribution forms, if by aircraft
The two-dimensional curve on head is launched into one-dimensional straight line, then the two-dimensional flow field of horizontal plane can be indicated with one-dimensional colour band, passes through
The position of different colours and/or brightness can clearly give expression to the angle of carrier in colour band, then by color in colour band or bright
The depth of degree can give expression to the pressure size of the position.
The middle section of colour band indicates carrier header region, and left part indicates that the odd side of carrier, right part indicate
The even number side of carrier.
For the operating condition that carrier is 0 ° relative to the angle in flow field, colour band middle position is white bosom, that is, is said
For bright Pressure maximum value in head center position, two sides are black bosom, i.e. pressure minimum.White and the depth of black are not
Together, just indicate that pressure value is different, that is, flow field velocity is different, as shown in Fig. 4 (a), the part (b).
For different angles (this angle is angle of the carrier relative to flow field, same as below), white bosom region
It can be used to the difference of expressive perspective with the position in black bosom region.White bosom region is located at right side, indicates angle
For positive value, white bosom region is located at left side, and expression angle is negative value.White bosom region is in the position of right side reference axis
Reacted the size of angle, position is more kept right, and shows that angle is bigger, and position more keeps left, show that angle is smaller, position in centre,
Indicate that angle is 0 °, as shown in Fig. 4 (c), the part (d).
In conjunction with shown in Fig. 2, Fig. 3 and Fig. 5, for three-dimensional situation, if by the development of a sphere of carrier header at a circle, that
The different location in white bosom region and black bosom region just can indicate the angle of carrier and water flow.
If carrier is in the horizontal plane, the white maximum of points and black minimum point of carrier are always located in 0 ° -180 °
On horizontal line, horizontal different location indicate carrier in the horizontal plane with the angle of water (flow) direction.
Each diameter line in X-Y scheme both corresponds to an one-dimensional colour band.And if carrier is located at other face
Interior, then the straight line that white bosom and black bosom are linked to be can pass through this angle with horizontal plane shape γ at a certain angle
Just can expression vector space angle.
If can determine that this angle from two dimension, it is restored in three-dimensional space, we can represent carrier
With the spatial position of water flow.
It is indicated with factors such as distribution of color, depth by pressure value size, although theoretically feasibility, this side
There are the numerical value that an apparent defect is pressure easily there is noise jamming under real conditions for method.
In this regard, the application proposes the visual perception of pressure difference method, i.e. " common mode interference " by pressure difference value removal centainly,
And adjacent pressure monitoring point distance is closer, common mode interference is smaller.
Further, the state change in flow field is expressed using visual pressure difference matrix.
Every a line of pressure difference matrix indicates the pressure difference distribution at some time point, and capable is time difference Δ between row
t.For pressure difference matrix, horizontal axis indicates the pressure difference of each adjacent monitoring point, contains the pressure distribution at current time
Information；The longitudinal axis indicates time series, contains the time state in flow field.
As shown in fig. 6, wherein t indicates time series, arrow indicates downwards the pressure difference vector of different time sequence；Example
Such as, d2123 indicates that p21-p23, d911 indicate that p9-p11, d13 indicate p1-p3, and so on.
It is 0 by pressure difference in colour bar pattern as shown in fig. 7, pressure difference matrix is showed in the form of RGB image
Color lump is known as No. 0 color lump.On the right of No. 0 color lump, coloration is brighter to illustrate that pressure difference is bigger, and pressure difference is positive value, also illustrates pressure
Value is declining；On No. 0 color lump left side, color is darker, and expression pressure difference is bigger, and pressure difference is negative value, and pressure value is becoming larger.
If Fig. 8 is the pressure difference image under different angle, friction speed.Therefrom it can be seen that, for v=0.1m/s,
Pressure difference is substantially close with No. 0 color lump at any angle.Pressure difference image by comparing same angle, friction speed can be seen
Out, the variation of pressure difference size has significantly been reacted in the variation of brightness.In addition, by comparing same speed, the pressure of different angle
Power difference image is it can also be seen that the variation of angle has also been reacted in the movement of brightness.
As shown in figure 9, being the pressure difference RGB image under different working conditions.
Pressure difference matrix is showed in the form of RGB image.RGB image is made of three data channel, respectively
For the channel R, the channel G and channel B, each channel are made of 8 data, and range is [0~255].
In the third party library matplotlib using Python, during differential pressure data is converted into image, because
Pressure difference has positive and negative values, these values can zoom between 0~1, and the pressure difference size between operating condition each in this way can not be presented
Clearly compare.For example, in v=0.1m/s, it is assumed that pressure difference is up to 10, minimum -10, then matplotlib meeting
It is 1 by 10 normalizeds, and is 0 by -10 normalizeds；And in v=0.5m/s, it is assumed that pressure difference is up to 100, most
Small is -100, then 100 normalizeds can be 1 by matplotlib, is normalized to processing 0 for -100.In this way, then two figure
The possible numerical value difference of the same point of middle brightness will be very big, lose comparative.
In order to guarantee that image comparison body shows differentiation under different working conditions and clear expression, the application use following number
According to scaling formula:
Wherein, r is to indicate normalized value, that is, the number between 0-1；D is pressure difference；α indicates limiting factor, takes α
It is 256, discovery is directed to different working conditions in simulation analysis process because before, and pressure difference absolute value is simultaneously no more than 256.
As shown in Figure 10, the Classification and Identification of different pressures difference image is carried out using convolutional neural networks algorithm.
Wherein, four layers of convolutional neural networks for having built a class Lenet-5 include the first convolutional layer (conv1), the
Two convolutional layers (conv2), full articulamentum (fc1) and softmax layers.
In each convolutional layer, data can be by convolution, batch standardization (bn) and pondization processing.
The core size of first convolutional layer is 5 × 5, quantity 16；Pond window size is 2, the moving step length of every dimension
It also is 2, activation primitive selects relu function.
The core size of second convolutional layer is 5 × 5, quantity 32；Pond layer parameter is identical as convolution 1, activation primitive selection
Relu function.
Full articulamentum shares 16 nodes, and activation primitive is relu function.
Softmax layers of interstitial content is identical as the number to be classified.
Data carry out evening up operation (corresponding to the Flatten in figure), so after the first convolutional layer and the second convolutional layer
Afterwards by full articulamentum, softmax layers are finally reached, obtains prediction output.
The label of prediction output and sample is compared and acquires loss loss, gradient is then calculated according to Adam optimizer
And undated parameter.
Particularly, using small batch (minibatch) training, the sample of each batch during training (train)
Number is 128.
Before carrying out network training, need to divide data set.The application is by test data according to pressure difference square
The form of battle array is arranged, wherein every a line represents the pressure difference of the adjacent sensors at a certain moment；Then, sample moment is constructed
Battle array, each matrix includes 22 rows altogether, and time interval is approximately equal to 2 seconds, that is to say, that each sample reflects the pressure in 2 second time
The variation of power difference.Then, a sample is created every a time point, that is, is differed between each sample 0.1 second approximate.
Sample as shown in figure 11 creates process, and each of figure box represents a sample matrix, and time interval is close
Approximately equal to 2 seconds, Δ t was approximately equal to 0.1 second.
Each operating condition is created and builds 1500 samples, and creates corresponding label.After having established all data sets,
Data set is upset into training set and test set, and guarantees that training set and test set meet same distribution.
Finally, pressure difference RGB image is identified by the convolutional neural networks, result is with higher accurate
Degree, demonstrate the application has more reliable performance in terms of perception flow field change and barrier.
, can be by the analysis to different pressures difference RGB image using the application, while obtaining flow velocity, flow direction, barrier and have
The process for identifying different information of flow is uniformly arrived " pressure difference RGB image " this tool by the information such as nothing, barrier size
In, without establishing corresponding mathematical model respectively for different information of flow.
It should be understood that for those of ordinary skills, it can be modified or changed according to the above description,
And all these modifications and variations should all belong to the protection domain of appended claims of the present invention.
Claims (5)
1. a kind of underwater cognitive method based on artificial side line visual image, it is characterised in that: it include following steps,
1) the array of pressure sensors data under different working conditions are measured；
2) pressure data for measuring step 1) is arranged in pressure difference matrix；
3) the pressure difference matrix that step 2) obtains is depicted as pressure difference RGB image；
4) the pressure difference RGB image input convolutional neural networks obtained step 3) carry out Classification and Identification, and it is each to obtain submerged flow field
Parameter is simultaneously classified.
2. the underwater cognitive method according to claim 1 based on artificial side line visual image, it is characterised in that: in institute
In the step 1) stated, the cloth along the artificial side line of submarine navigation device and aircraft head and artificial side line same level section
Equipped with several sensors (11).
3. the underwater cognitive method according to claim 1 based on artificial side line visual image, it is characterised in that: in institute
In the step 2) stated, every a line of pressure difference matrix indicates the pressure difference distribution at some time point, and capable is the time between row
Poor Δ t；
In pressure difference matrix, horizontal axis indicates the pressure difference of each adjacent monitoring point, contains the pressure distribution letter at current time
Breath；The longitudinal axis indicates time series, contains the time state in flow field.
4. the underwater cognitive method according to claim 1 based on artificial side line visual image, it is characterised in that: in institute
In the step 3) stated, the data zooming such as following formula of RGB image is generated,
Wherein, α indicates limiting factor, and taking α is 256.
5. the underwater cognitive method according to claim 1 based on artificial side line visual image, it is characterised in that: in institute
In the step 4) stated, it includes the first convolution that the convolutional neural networks, which have four layers of convolutional neural networks of class Lenet-5,
Layer (conv1), the second convolutional layer (conv2), full articulamentum (fc1) and softmax layers；
Data carry out evening up operation after the first convolutional layer and the second convolutional layer, then pass through full articulamentum, finally reach
Softmax layers, obtain prediction output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910231013.2A CN110119671B (en) | 2019-03-26 | 2019-03-26 | Underwater sensing method based on artificial lateral line visual image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910231013.2A CN110119671B (en) | 2019-03-26 | 2019-03-26 | Underwater sensing method based on artificial lateral line visual image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110119671A true CN110119671A (en) | 2019-08-13 |
| CN110119671B CN110119671B (en) | 2022-10-04 |
Family
ID=67520626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910231013.2A Active CN110119671B (en) | 2019-03-26 | 2019-03-26 | Underwater sensing method based on artificial lateral line visual image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110119671B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113283443A (en) * | 2020-02-20 | 2021-08-20 | 中国石油天然气股份有限公司 | Working condition identification method and device, computer equipment and storage medium |
| CN115493740A (en) * | 2022-11-14 | 2022-12-20 | 长江勘测规划设计研究有限责任公司 | Method and system for measuring pressure pulsation of internal flow passage of water turbine by using machine vision |
| CN117538011A (en) * | 2023-11-01 | 2024-02-09 | 中国北方车辆研究所 | Underwater flow field sensing and identifying method for amphibious vehicle and acquisition and terminal equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150285035A1 (en) * | 2014-04-02 | 2015-10-08 | Onesubsea Ip Uk Limited | Controlled pressure equalization |
| CN105333988A (en) * | 2015-11-25 | 2016-02-17 | 中国海洋大学 | Artificial lateral line pressure detection method |
| CN107145105A (en) * | 2017-05-24 | 2017-09-08 | 北京大学 | A kind of artificial lateral-line system based on pressure transducer array |
| CN109061101A (en) * | 2018-06-29 | 2018-12-21 | 东北大学 | Thickener underflow concentration, mud layer height, internal mine amount hard measurement device and method |
| US10161822B1 (en) * | 2016-03-10 | 2018-12-25 | The United States Of America, As Represented By The Secretary Of The Navy | Differential pressure measurement system with solenoid coupled reference reservoir |
-
2019
- 2019-03-26 CN CN201910231013.2A patent/CN110119671B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150285035A1 (en) * | 2014-04-02 | 2015-10-08 | Onesubsea Ip Uk Limited | Controlled pressure equalization |
| CN105333988A (en) * | 2015-11-25 | 2016-02-17 | 中国海洋大学 | Artificial lateral line pressure detection method |
| US10161822B1 (en) * | 2016-03-10 | 2018-12-25 | The United States Of America, As Represented By The Secretary Of The Navy | Differential pressure measurement system with solenoid coupled reference reservoir |
| CN107145105A (en) * | 2017-05-24 | 2017-09-08 | 北京大学 | A kind of artificial lateral-line system based on pressure transducer array |
| CN109061101A (en) * | 2018-06-29 | 2018-12-21 | 东北大学 | Thickener underflow concentration, mud layer height, internal mine amount hard measurement device and method |
Non-Patent Citations (3)
| Title |
|---|
| GUIJIE LIU, ANYI WANG, XINBAO WANG, AND PENG LIU: "A Review of Artificial Lateral Line in Sensor Fabrication and Bionic Applications for Robot Fish", 《APPLIED BIONICS AND BIOMECHANICS》 * |
| RISTROPH L, LIAO J C, ZHANG J.: "Lateral line layout correlates with the differential hydrodynamic pressure on swimming fish", 《PHYSICAL REVIEW LETTERS》 * |
| 刘贵杰，宫华耀，吴乃龙，闫茹，李蒙蒙: "基于鱼类侧线感知机理的流场辨识方法及仿真研究", 《机械工程学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113283443A (en) * | 2020-02-20 | 2021-08-20 | 中国石油天然气股份有限公司 | Working condition identification method and device, computer equipment and storage medium |
| CN115493740A (en) * | 2022-11-14 | 2022-12-20 | 长江勘测规划设计研究有限责任公司 | Method and system for measuring pressure pulsation of internal flow passage of water turbine by using machine vision |
| CN115493740B (en) * | 2022-11-14 | 2023-02-28 | 长江勘测规划设计研究有限责任公司 | Method and system for measuring pressure pulsation of internal flow passage of water turbine by using machine vision |
| CN117538011A (en) * | 2023-11-01 | 2024-02-09 | 中国北方车辆研究所 | Underwater flow field sensing and identifying method for amphibious vehicle and acquisition and terminal equipment |
| CN117538011B (en) * | 2023-11-01 | 2024-10-01 | 中国北方车辆研究所 | Underwater flow field sensing and identifying method for amphibious vehicle and acquisition and terminal equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110119671B (en) | 2022-10-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cong et al. | Underwater robot sensing technology: A survey | |
| CN110414396B (en) | Unmanned ship perception fusion algorithm based on deep learning | |
| Han et al. | Marine organism detection and classification from underwater vision based on the deep CNN method | |
| CN110119671A (en) | Underwater cognitive method based on artificial side line visual image | |
| CN111856448A (en) | Marine obstacle identification method and system based on binocular vision and radar | |
| CN109932708B (en) | Method for classifying targets on water surface and underwater based on interference fringes and deep learning | |
| Liu et al. | Detection and pose estimation for short-range vision-based underwater docking | |
| CN110108704A (en) | A kind of automatic monitoring and pre-alarming method of cyanobacteria and its automatic monitoring and alarming system | |
| CN106570485B (en) | A kind of raft culture remote sensing images scene mask method based on deep learning | |
| CN111637874B (en) | Multi-AUV layered detection system and detection method for red tide sea area | |
| CN109859202A (en) | A kind of deep learning detection method based on the tracking of USV water surface optical target | |
| CN111144208A (en) | Automatic detection and identification method for marine vessel target and target detector | |
| CN109284663A (en) | A kind of sea obstacle detection method based on normal state and uniform Mixture Distribution Model | |
| CN117214904A (en) | Intelligent fish identification monitoring method and system based on multi-sensor data | |
| CN116595870A (en) | Dynamic platform-based wake identification and flow field detection method and system | |
| Liu et al. | Research on artificial lateral line perception of flow field based on pressure difference matrix | |
| CN116823838B (en) | Ocean ship detection method and system with Gaussian prior label distribution and characteristic decoupling | |
| CN109579845A (en) | A kind of ship tracking localization method based on electric field detecting array | |
| CN117454680A (en) | Ocean search cluster design system and method | |
| CN116242584A (en) | Floating ocean platform along with ship wave measuring device based on BP neural network | |
| CN114202587A (en) | Visual feature extraction method based on shipborne monocular camera | |
| Yang et al. | Wind measurement by computer vision on unmanned sailboat | |
| CN112326917A (en) | Water environment pollution traceability system | |
| CN110895680A (en) | Unmanned ship water surface target detection method based on regional suggestion network | |
| CN118072206B (en) | Unmanned aerial vehicle remote sensing image target recognition method based on whale habitat selection model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |