CN106770087A - Greasy dirt remote sensing module, system and method - Google Patents
Greasy dirt remote sensing module, system and method Download PDFInfo
- Publication number
- CN106770087A CN106770087A CN201611053096.3A CN201611053096A CN106770087A CN 106770087 A CN106770087 A CN 106770087A CN 201611053096 A CN201611053096 A CN 201611053096A CN 106770087 A CN106770087 A CN 106770087A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- greasy dirt
- region
- oil
- 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
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
- G01N21/6428—Measuring fluorescence of fluorescent products of reactions or of fluorochrome labelled reactive substances, e.g. measuring quenching effects, using measuring "optrodes"
- G01N21/643—Measuring fluorescence of fluorescent products of reactions or of fluorochrome labelled reactive substances, e.g. measuring quenching effects, using measuring "optrodes" non-biological material
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Health & Medical Sciences (AREA)
- Immunology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Optics & Photonics (AREA)
- Molecular Biology (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a kind of greasy dirt remote sensing module, system and method, the module includes:Long wave ultraviolet line source, the long wave ultraviolet line source can produce the long wave ultraviolet detectable signal to water surface irradiation to be measured;It is capable of the view data of the Real-time Collection water surface to be measured in image data acquiring end, the image data acquiring end;And greasy dirt identification end, greasy dirt identification end can be it is determined that carrying out image procossing and identifying the surface of the water surface to be measured and/or whether there is greasy dirt under water after image type corresponding to acquired image data.The present invention extracts the feature of oil contaminants in image by being loaded with the greasy dirt identification end of Ultraluminescence oil spilling Intelligent Recognition algorithm and carries out accurate, high-sensitivity intelligent oil identification, with on the efficient identification water surface and underwater greasy dirt and its have and be easy to modularization, it is lightweight, small volume, can be embedded in easily in aircraft pod, in boat-carrying monitoring system and in bank base monitoring system the advantages of.
Description
Technical field
The present invention relates to greasy dirt Detection Techniques, particularly relate on a kind of airborne, boat-carrying and the bank base water surface, underwater
Etc. greasy dirt remote sensing module, the system and method for occasion oil overflowing remote sense monitoring.
Background technology
With quickly propelling for global industry process, human society is growing day by day to the demand of oil, its exploitation rule
Mould expands rapidly, and offshore field quantity and oil transportation at sea amount are sharply increased.Made by reasons such as operation on the sea, transport and equipment
Into offshore spilled oil event be also continuously increased, this exacerbates petroleum pollution in ocean, and then compromises the existence ring of marine organisms
Border, seriously threatens the marine eco-environment and fishery resources.
Airborne offshore oil overflowing remote sense monitoring system product has the AIREYE (infrared/visible ray) in the U.S., auspicious in the world at present
Ocean monitoring system MSS of allusion quotation etc., the MSS5000 of Aerospace PLC, BAe of foremost Yao Shu Sweden;And boat-carrying spilled oil monitoring system
System has SeaDarQ radar surveillance systems, and bank base oil spill monitoring system has " water-surface oil spilling monitoring alarm system ".Spain, Sweden,
Many developed countries such as western cured, Dutch, Finland, U.S. are all equipped with above-mentioned monitoring system respectively.
But the hardware of said system or detector be all using technology maturation and more advanced radar system, it is infrared/
Ultraviolet imagery system and contactless Ultraluminescence measuring system, and these systems can use with integrated, it is also possible to it is used alone,
With oil spilling effect on the preferably detection water surface.But greasy dirt detection then all failure of these systems to underwater.
The content of the invention
In view of the defect that prior art is present, the invention aims to provide a kind of greasy dirt remote sensing module, the oil
Dirty remote sensing module can be embedded into existing remote sense monitoring system, with the greasy dirt of on the Real time identification water surface and underwater.
To achieve these goals, technical scheme:
A kind of greasy dirt remote sensing module, it is characterised in that including:
Long wave ultraviolet line source, the long wave ultraviolet line source can be produced and detected to the long wave ultraviolet of water surface irradiation to be measured
Signal;
It is capable of the view data of the Real-time Collection water surface to be measured in image data acquiring end, the image data acquiring end;
And greasy dirt identification end, greasy dirt identification end can be from the image type determined corresponding to acquired image data
Image procossing is carried out afterwards and is identified the surface of the water surface to be measured and/or be whether there is greasy dirt under water.
Further, as preferred scheme of the invention
The long wave ultraviolet line source is light-focusing type long wave ultraviolet lamp.
Further, as preferred scheme of the invention
The greasy dirt identification end includes:Can it is determined that after image type corresponding to acquired image data, according to
Identified image type carries out the image processing part of respective image treatment and can be based on the figure that image processing part is obtained
The surface of the water surface to be measured is identified as processing data and/or under water with the presence or absence of the greasy dirt identification part of greasy dirt;Wherein at described image
Reason portion includes:
View data reading unit, the view data reading unit can in real time read what image data acquiring end was gathered
View data;
Image takes frame unit, and the image takes frame unit and can take frame mode with constant duration and described image data are carried out
Frame extraction process;
Image pre-processing unit, the image pre-processing unit can be to carrying out figure through each view data after frame extraction process
As pretreatment;
Graphics processing unit, the graphics processing unit can determine the figure corresponding to current view data after pretreatment
It is, then using the blue channel component of evening images as source images, and the source images to be entered as whether type is evening images
Row threshold division treatment, using be partitioned into from the source images as the presence oil spilling of target area image-region and as
The oil-free image-region of background area;Otherwise determine that the image type corresponding to current view data after pretreatment is daytime
Image, and image threshold treatment is carried out to the image based on oil spilling area pixel characteristic, it is partitioned into the presence of excessive with from described image
The image-region and oil-free image-region of oil.
Further, as preferred scheme of the invention
The image pixel Distribution value that described image processing unit is based on corresponding to current view data after pretreatment is special
Levy determine its corresponding to image type.
Further, as preferred scheme of the invention
It is right to obtain that described image processing unit can enter row threshold division to source images based on pre-stored Otsu algorithms
The bianry image answered, and in bianry image pixel value to confirm as target area in 255 corresponding image regions, pixel value is 0
Background area is confirmed as in region corresponding image region, the image-region and oil-free image district that there is oil spilling to be effectively partitioned into
Domain;Described Otsu algorithms enter row threshold division process to source images:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the pixel appearance of i
Probability pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target
Class pixel set C1, wherein C0Pixel composition by gray value between [0, k], C1By picture of the gray value between [k+1, L-1]
Element composition,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), it is determined that working asCorresponding k values are optimal point when maximum
Threshold value is cut, row threshold division is entered to image finally according to the threshold value k for obtaining, obtain bianry image.
Further, as preferred scheme of the invention
Described image processing unit can carry out pixel characteristic analysis to the day images, i.e., successively in image
Each pixel is compared judgement, if there is in pixel rgb value a certain component in image with respect to other two points
More than set threshold value, then it is the pixel corresponding to the image-region that there is oil spilling to confirm as the pixel to amount, with effective
It is partitioned into the image-region and oil-free image-region that there is oil spilling.
The purpose of the present invention also provides a kind of greasy dirt remote sensing system, and it includes:Existing remote sense monitoring system with
And it is embedded in the greasy dirt remote sensing module of the remote sense monitoring system.
Separately, the purpose of the present invention also provides a kind of greasy dirt remote detecting method, it is characterised in that comprise the following steps:
S1, produce long wave ultraviolet detectable signal to water surface irradiation to be measured;
The view data of S2, the Real-time Collection water surface to be measured;
Image procossing is carried out after S3, the image type for determining corresponding to acquired image data and identify the water surface to be measured
Surface and/or under water whether there is greasy dirt.
Further, as preferred scheme of the invention
The S3 includes:
S31, in real time reading acquired image data;
S32, frame mode is taken with constant duration frame extraction process is carried out to described image data;
S33, to carrying out image preprocessing through each view data after frame extraction process;
Whether the image type corresponding to S34, the current view data after pretreatment of determination is evening images, then will be
The blue channel component of evening images enters row threshold division treatment to the source images as source images, with from the source figure
The image-region as the presence oil spilling of target area and the oil-free image-region as background area are partitioned into as in;Otherwise
It is determined that the image type corresponding to current view data after pretreatment is day images, and based on oil spilling area pixel characteristic pair
The image carries out image threshold treatment, to be partitioned into the image-region and the oil-free image district that there is oil spilling from described image
Domain;
S35, the surface that the water surface to be measured is identified based on the image processing data that S34 is obtained and/or under water with the presence or absence of oil
It is dirty.
Further, as preferred scheme of the invention
Determine that its institute is right by the image pixel value distribution characteristics corresponding to current view data after pretreatment in S34
The image type answered.
Further, as preferred scheme of the invention
Otsu algorithms in S34 by being pre-stored enter row threshold division to source images to obtain corresponding bianry image, and
By pixel value in bianry image for target area is confirmed as in 255 corresponding image regions, pixel value is 0 region corresponding image
Background area is confirmed as in region, the image-region and oil-free image-region that there is oil spilling to be effectively partitioned into;Described Otsu
Algorithm enters row threshold division process to source images:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the pixel appearance of i
Probability pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target
Class pixel set C1, wherein C0Pixel composition by gray value between [0, k], C1By picture of the gray value between [k+1, L-1]
Element composition,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), it is determined that working asCorresponding k values are optimal point when maximum
Threshold value is cut, row threshold division is entered to image finally according to the threshold value k for obtaining, obtain bianry image.
Further, as preferred scheme of the invention
S34 includes that the day images are carried out with pixel characteristic analysis is compared to each pixel in image sentences successively
It is disconnected, if there is threshold value of a certain component with respect to other two components more than set by pixel rgb value in image, confirm as
The pixel is the pixel corresponding to the image-region that there is oil spilling, exist to be effectively partitioned into oil spilling image-region and
Oil-free image-region.
Compared with prior art, beneficial effects of the present invention:
The present invention uses light-focusing type uviol lamp instead of ultraviolet laser as the excitation source for exciting the upper and lower oil spilling of the water surface,
And oily pollution in high definition camera colored UV fluorescence (blue green light) image is extracted using Ultraluminescence oil spilling Intelligent Recognition algorithm
The feature of thing is to realize accurate, high-sensitivity intelligent oil identification process, and equipment of the invention has and is easy to modularization, weight
Gently, small volume, can be embedded in easily in aircraft pod, in boat-carrying monitoring system and in bank base monitoring system the advantages of.
Brief description of the drawings
Fig. 1 is the structural representation of the upper and lower greasy dirt remote sensing module of the water surface of the invention.
Fig. 2 is Ultraluminescence oil spilling Intelligent Recognition algorithm flow chart of the invention.
Fig. 3-1 is the night spilled oil simulation image illustration of example of the present invention;
Fig. 3-2 is the evening images grey level histogram corresponding to Fig. 3-1;
Fig. 3-3 is the spilled oil simulation image illustration on daytime of example of the present invention;
Fig. 3-4 is the day images grey level histogram corresponding to Fig. 3-3;
Fig. 3-5 is the night blue channel analog image illustration of example of the present invention;
Fig. 3-6 is the blue channel histogram corresponding to Fig. 3-5;
Fig. 3-7 be example of the present invention a certain photograph intensity and shooting angle under obtain oil spilling image on daytime;
Fig. 3-8 be example of the present invention another photograph intensity and shooting angle under obtain oil spilling image on daytime;
Fig. 3-9 be Fig. 3-7 oil spilling image procossing after image;
Fig. 3-10 is without image after oil spilling image procossing;
Fig. 4-1 is the image entered to evening images after row threshold division of the present invention;
Fig. 4-2 is the image entered to day images after row threshold division of the present invention;
Fig. 4-3 is the image after Fig. 4-1 scannings;
Fig. 4-4 is the image after Fig. 4-2 scannings.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with the embodiment of the present invention in it is attached
Figure, is clearly and completely described to technical scheme, it is clear that described embodiment is that a part of the invention is real
Example is applied, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation
Property work under the premise of the every other embodiment that is obtained, belong to the scope of protection of the invention.
Fundamental design idea of the invention:Because long wave ultraviolet has good seawater penetration capacity and works as long wave ultraviolet
When line is radiated at the water surface to be measured, if its surface has the greasy dirts such as petroleum-type product, fluorescence can be produced;If its surface is cleaning
Or when being radiated at decoy (such as phytoplankton, water plant, sea grass under water), then will not produce fluorescence;In addition long wave
Ultraviolet is radiated at when on the water surface of cleaning, can be continued to advancing under water, when long wave ultraviolet runs into along certain refraction angle
Stay under water under water oil when can also produce fluorescence;Fluorescence then will not be equally produced when UVA Radiation is to decoy under water,
Therefore using the such feature of long wave ultraviolet, the large-scale water surface, under water a range of petroleum-type product objective are realized
Search work;It is to be capable of achieving the standard to water petrochina class product objective in conjunction with acquisition technology and image processing techniques
Really, high-sensitivity intelligent identification process.
Based on above-mentioned design philosophy, such as Fig. 1, greasy dirt remote sensing module of the present invention, it mainly includes following several
Individual part:
(1) long wave ultraviolet line source, the long wave ultraviolet line source can produce the long wave ultraviolet to water surface irradiation to be measured
Detectable signal;Specifically as preferred case of the invention, the long wave ultraviolet line source uses 365nm long wave ultraviolet lamps
(light-focusing type), using long wave ultraviolet lamp from water surface certain limit to be measured, such as on aircraft, on ship or bank eminence to
The water surface to be measured launches long wave ultraviolet.
(2) it is capable of the view data of the Real-time Collection water surface to be measured in image data acquiring end, the image data acquiring end;Specifically
Conduct preferred case of the invention, described image data acquisition end uses high definition camera, and the high definition camera is mountable to aircraft
On upper, ship or bank eminence etc., can be such as arranged on same position of platform with uviol lamp, to obtain respective image number
According to.The data at image data acquiring end can be by wireless transmission method, such as wireless transport module simultaneously.
(3) greasy dirt identification end, the greasy dirt recognizes end it is determined that the image type corresponding to acquired image data is laggard
Row image procossing simultaneously identifies the surface of the water surface to be measured and/or whether there is greasy dirt under water.
Specifically as preferred case of the invention, the greasy dirt identification end is made up of image processing part, greasy dirt identification part:
Wherein, image processing part is used for it is determined that after image type corresponding to acquired image data, according to really
Fixed image type carries out respective image treatment;Described image processing unit includes:
View data reading unit, the view data reading unit can in real time read what image data acquiring end was gathered
View data;
Image takes frame unit, and the image takes frame unit and can take frame mode with constant duration and described image data are carried out
Frame extraction process;
Image pre-processing unit, the image pre-processing unit can be to carrying out figure through each view data after frame extraction process
As pretreatment;Specifically as preferred case of the invention, described image pretreatment includes carrying out image to foregoing view data
Denoising, because noise is one of key factor of influence image segmentation, therefore suppresses noise, improves signal noise ratio (snr) of image
It is essential link in image preprocessing, present invention preferably employs image filtering method, such as neighborhood averaging filtering method enters
Row is pre-processed to noise.Neighborhood averaging filter method is a kind of the more commonly used filter processing method, its cardinal principle be by
The average gray value of the gray value neighborhood of pixels of image pixel is represented, to reach the purpose of smoothed image, removal noise, passed through
After neighborhood averaging filtering, the corresponding output G (j, k) of pixel F (j, k) of view data is:
In formula, F (j, k) represents the corresponding gray value of pixel of view data, and A represents the neighborhood centered on it, and L is represented
The number of pixel in neighborhood;If F (j, k) is noise spot, then have very big difference with it with its pixel grey scale closer to the distance,
Replace it using the average value of its neighborhood pixels, then can significantly weaken the influence of noise, become the gradation of image in neighborhood
Uniformly, and then the effect of smooth grey is reached.Simultaneously in order to reduce amount of calculation, arithmetic speed is improved, described image pretreatment is also
Including carrying out image Compression to foregoing view data, i.e., foregoing view data original image pixel is uniformly adopted at equal intervals
Sample realizes compression of images, to keep sampled pixel still to reflect the general picture feature of original image;Specifically such as set original
The size of view data is W*H, and the diminution factor of width and length is respectively k1And k2, then the corresponding sampling interval is:W/
k1, W/k2;That is artwork horizontal direction every W/k1, in vertical direction every W/k2Take a pixel.
The purpose for setting graphics processing unit is completed to be divided into oil spilling region and non-oil spilling region two parts
Image segmentation treatment, in order to greasy dirt identification part according to segmentation result, obtains corresponding oil spilling status information, finally to above-mentioned each
Whether information carries out comprehensive analysis, occurs the result of oil spilling to draw.That is graphics processing unit is needed to view data
Enter row threshold division so that view data is each divided into two parts, a part is regional aim, and another part is background area
Domain, then for the view data that there is oil spilling, its target area is then oil spilling region, and background area is non-oil spilling region, while
When view data does not exist oil spilling, then Zone Full is all background area.
Separately below to image type for evening images are specifically described with the technology of day images:
If image type is evening images, with following characteristics:In the case that at night, light is faint, if using ultraviolet
Fluorescent lamp replaces sunshine, and the illumination that ultraviolet fluorescent lamp sends is mapped on the oil film in oil spilling region, can be reflected back blue light, non-to overflow
Oily region does not have this phenomenon then;Similarly, the video figure that oil spilling region can be captured by the camera at image data acquiring end
As can also present the color characteristic different from non-oil spilling region (because under the irradiation of ultraviolet fluorescent lamp, the oil film meeting of oil spilling region
Blue pixel characteristic is presented, i.e., the target area and background area in acquired image data in its blue channel image are poor
Different maximum, is more beneficial for the segmentation of target area and background area) so, in follow-up image segmentation treatment, can select night
The blue channel of late image is used as dividing processing source images.
Based on above-mentioned principle, described image processing unit is required to determine that current view data institute after pretreatment is right
Whether the image type answered is evening images, is then using the blue channel component of evening images as source images, and to the source
Image enters row threshold division treatment, using the image-region as the presence oil spilling of target area is partitioned into from the source images with
And as the oil-free image-region of background area;Otherwise determine the image type corresponding to current view data after pretreatment
It is day images, and image threshold treatment is carried out to the image based on oil spilling area pixel characteristic, is partitioned into from described image
There is the image-region and oil-free image-region of oil spilling.Specifically as preferred case of the invention, described image treatment is single
The image pixel value distribution characteristics that unit can be based on corresponding to current view data after pretreatment determines the figure corresponding to it
As type.Such as, because evening images and day images have bigger difference in the image pixel value regularity of distribution, therefore can be based on
Image pixel value distribution characteristics determines evening images of the corresponding image type of image as shown in Fig. 3-1, Fig. 3-2, at night
In the case of not having sunshine, most of pixel value of its image is all relatively low, and the day images as shown in Fig. 3-3, Fig. 3-4,
Its image has very big difference on pixel Distribution value with evening images, also just says due to day and night because respective image is special
Property, the pixel value of day images therefore can calculate every apparently higher than the characteristic of evening images by substantial amounts of experimental data
The average pixel value of width image, then by counting the average pixel value of a large amount of day and night images, can find optimal picture
Element value cut-point (segmentation threshold), using the cut-point divide image type (i.e. determine night and day images, determine it is right
The image type answered, for the image procossing for carrying out follow-up is prepared.
Because most of region is all the water surface in evening images, with similar Pixel Information, for analog information
Region, their pixel distribution approach, can show on the histogram " crest ";As shown in Fig. 3-5,3-6, because night schemes
Most of region of picture is black, so pixel concentrates on low value region in correspondence histogram, oil spilling area distribution is in pixel value
Region more than 50 and the evening images for there is oil spilling, target area and background area have different Pixel Informations,
Can be presented on histogram different " crests ".In actual applications, when oil spill area reaches certain ratio with respect to entire image area
During example, distribution of the image pixel in histogram can just be rendered obvious by two effects of " crest ", in order to identify oil spilling, enter
And obtain oil spilling information.Therefore in view of distribution of the evening images pixel in histogram can be presented two " crests " i.e. double-peak feature
Effect, then image segmentation can be carried out to evening images blue channel component using Otsu algorithms, by target area and background area
Domain splits exactly.Otsu algorithms-maximum variance between clusters are a kind of conventional Research on threshold selection, and it is using between class
Variance maximum automatically determines threshold value and carries out image segmentation based on the threshold value.
Specifically as preferred case of the invention, described image processing unit can be based on pre-stored Otsu algorithms pair
Enter row threshold division as the blue channel component of the evening images of source images to obtain corresponding bianry image, and bianry image
To confirm as target area in 255 corresponding image regions, pixel value is that 0 region corresponding image region is confirmed as to middle pixel value
Background area, the image-region and oil-free image-region that there is oil spilling to be effectively partitioned into;Described Otsu algorithms are to source figure
It is as entering row threshold division process:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the pixel appearance of i
Probability pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target
Class pixel set C1, wherein C0Pixel composition by gray value between [0, k], C1By picture of the gray value between [k+1, L-1]
Element composition,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), it is determined that working asCorresponding k values are optimal point when maximum
Threshold value is cut, row threshold division is entered to image finally according to the threshold value k for obtaining, obtain bianry image.Then pixel value in bianry image
For 255 region be target area (oil spilling region), pixel value be 0 region be background area (aqua region).
If image type be day images, i.e., in the case where on daytime, light is strong, solar irradiation be mapped on oil film with
During other regions such as water surface, because oil film property reflects light, the video image captured by camera can be presented and it
The different color characteristic in his region, i.e., can be showed, such as such as Fig. 3-7,3-8 institutes in view data with different Pixel Informations
The oil spilling image on daytime obtained under the two kinds of different illumination intensities and different shooting angles shown, two images are in oil spilling region
Existing motley feature, and the pixel value in other regions is relatively.Although the oil spilling image of day images is in pixel simultaneously
Night oil spilling image is far longer than in information content and complexity, but, in same oil spilling image on width daytime, oil spilling region and
There is obvious Pixel Information difference in non-oil spilling region, according to the pixel characteristic of oil spilling on daytime, again may be by Threshold segmentation pair
Two regions are divided.Specifically as preferred case of the invention, described image processing unit can scheme to the daytime
As carrying out pixel characteristic analysis, i.e., judgement is compared to each pixel in image successively, if there is pixel rgb value in image
In a certain component with respect to other two components more than set threshold value, then it is the image district that there is oil spilling to confirm as the pixel
Pixel corresponding to domain, the image-region and oil-free image-region that there is oil spilling to be effectively partitioned into.As such as Fig. 3-9,3-
To image 3-7 process shown in 10 obtains Fig. 3-9, and the image for occurring without oil spilling, Fig. 3-10 is obtained after treatment.
Wherein, the image processing data that greasy dirt identification part is used to be obtained based on image processing part identifies the water surface to be measured
Surface and/or under water whether there is greasy dirt.Described image processing data includes the blueness to night oil spilling image by Otsu algorithms
Channel components enter the image that row threshold division is obtained, as shown in Fig. 4-1, and by pixel characteristic analyze be (rgb value feature and
Quantative attribute) enter the image that row threshold division is obtained, as shown in the Fig. 4-2, wherein Fig. 4-3 is the image after Fig. 4-1 scannings;Fig. 4-
4 is the image after Fig. 4-2 scannings.It should be noted that in the practical application of module, its treatment is video flowing, and video flowing
It is a series of images with temporal characteristics, therefore is carrying out when image (frame of video) is processed, it is necessary to be overflow to every piece image
Oil condition is marked, i.e., oil spilling status indication is 1 (having oil spilling), is otherwise 0 (without oil spilling).Simultaneously as it is one oil spilling occur
Individual lasting process, it is therefore desirable to record the oil spilling state in a period of time, that is, records several adjacent treatment frames;Greasy dirt
Whether identification part carries out comprehensive analysis for these oil spilling states, finally to draw be judged to oil spilling occur.If there is then sending out
Deliver newspaper alert data item;It otherwise is left intact.
It is specific as preferred case of the invention, greasy dirt identification part can by way of scan image processing data,
The local region for meeting oil spilling feature of statistics sets scan box, travels through entire image, and Rule of judgment is:When the area for meeting oil spilling
Whether domain reaches certain amount, is, can be determined that frame under process has oil spilling.As such as Fig. 3-3, Fig. 3-4, in Fig. 3-3,
When pixel value is that 255 pixel number reaches 1/2 of number of pixels in scan box in scan box, mark is this time scanned;When
After the end of scan, all labeled scanning times are counted, when number of times reaches certain quantity, judge that present frame is present and overflow
Oil.Similarly, treatment is scanned to Fig. 3-4, according to its discrimination standard, standard compliant scanning times is counted, when number of times reaches
During certain quantity, judge that present frame has oil spilling.The meaning of statistics is to exclude erroneous judgement, improves the degree of accuracy for judging.
Actual experiment is proved:The detectivity of greasy dirt remote sensing module of the present invention:Detection range is 100 meters and warp
Substantial amounts of feasibility Experiment shows that the depth of oil spilling under the module detection water surface of the invention depends on water quality, can be visited in running water
Survey underwater oil clot of the distance more than 3 meters, the oil clot in yellow water at 0.5 meter of detectable underwater distance, the water detection in Bohai Sea Gulf
Underwater distance is between 2.0~3.0 meters.
The present invention also provides one kind by greasy dirt remote sensing Module-embedding in rear structure in other existing remote sense monitoring systems
Into greasy dirt remote sensing system, the dirty remote sensing system can be believed by the IP address of camera, account, password and port numbers etc.
Breath, accesses camera, and video flowing is read in real time;Frame of video is taken from video flowing at set time intervals;Then it is directed to
Frame of video to be processed carries out the operation such as image preprocessing, image procossing, Threshold segmentation, and data are carried out after obtaining processing data
Comprehensive analysis processing, judges whether to meet the condition for oil spilling occur.If met, alarm signal is sent to web service
And store the data item of the warning message;Web service receive data item, warning window are ejected on platform, and show
Warning message.Conversely, not sending alarm signal then, subsequent video stream is continued with.Experiment is proved:The dirty remote sensing system can
Oil spilling, accuracy rate >=99% on the round-the-clock remote sensing Real time identification water surface;And oil film on the water surface that detects can be realized>2μm;Detection
The minimum target radial direction 5m for arriving;While the round-the-clock oil clot for real-time detecting underwater pawpaw size, accuracy rate >=95%.
Separately, such as Fig. 2, the present invention provides a kind of greasy dirt remote detecting method, it is characterised in that comprise the following steps:
S1, produce long wave ultraviolet detectable signal to water surface irradiation to be measured;
The view data of S2, the Real-time Collection water surface to be measured;
Image procossing is carried out after S3, the image type for determining corresponding to acquired image data and identify the water surface to be measured
Surface and/or under water whether there is greasy dirt.
Further, as preferred scheme of the invention
The S3 includes:
S31, in real time reading acquired image data;
S32, frame mode is taken with constant duration frame extraction process is carried out to described image data;
S33, to carrying out image preprocessing through each view data after frame extraction process;
Whether the image type corresponding to S34, the current view data after pretreatment of determination is evening images, then will be
The blue channel component of evening images enters row threshold division treatment to the source images as source images, with from the source figure
The image-region as the presence oil spilling of target area and the oil-free image-region as background area are partitioned into as in;Otherwise
It is determined that the image type corresponding to current view data after pretreatment is day images, and based on oil spilling area pixel characteristic pair
The image carries out image threshold treatment, to be partitioned into the image-region and the oil-free image district that there is oil spilling from described image
Domain;
S35, the surface that the water surface to be measured is identified based on the image processing data that S34 is obtained and/or under water with the presence or absence of oil
It is dirty.
Further, as preferred scheme of the invention
Determine that its institute is right by the image pixel value distribution characteristics corresponding to current view data after pretreatment in S34
The image type answered;Otsu algorithms in S34 by being pre-stored enter row threshold division to source images to obtain corresponding binary map
Picture, and by pixel value in bianry image for target area is confirmed as in 255 corresponding image regions, pixel value is right by 0 region
Image-region is answered to confirm as background area, the image-region and oil-free image-region that there is oil spilling to be effectively partitioned into;It is described
Otsu algorithms source images entered with row threshold division process be:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the pixel appearance of i
Probability pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target
Class pixel set C1, wherein C0Pixel composition by gray value between [0, k], C1By picture of the gray value between [k+1, L-1]
Element composition,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), it is determined that working asCorresponding k values are optimal point when maximum
Threshold value is cut, row threshold division is entered to image finally according to the threshold value k for obtaining, obtain bianry image;
S34 includes that the day images are carried out with pixel characteristic analysis is compared to each pixel in image sentences successively
It is disconnected, if there is threshold value of a certain component with respect to other two components more than set by pixel rgb value in image, confirm as
The pixel is the pixel corresponding to the image-region that there is oil spilling, exist to be effectively partitioned into oil spilling image-region and
Oil-free image-region.
In sum, the present invention excites the upper and lower oil spilling of the water surface by replacing ultraviolet laser to be used as with light-focusing type uviol lamp
Excitation source, the spy that oil contaminants in high definition camera colored UV fluorescence (blue green light) image are extracted using greasy dirt identification part
Levy and carry out accurate, high-sensitivity intelligent oil identification;And the present invention easily modularization, small volume such as can easily be embedded in airborne hanging
In cabin, in boat-carrying monitoring system and in bank base monitoring system;Power consumption is again small, is particularly suitable for unmanned plane and uses, and China's unmanned plane
Developed in recent years rapidly, the again cheap unmanned plane of technically reliable use is popularized application in an all-round way and provided for module of the invention
Solid foundation condition.
The above, the only present invention preferably specific embodiment, but protection scope of the present invention is not limited thereto,
Any one skilled in the art the invention discloses technical scope in, technology according to the present invention scheme and its
Inventive concept is subject to equivalent or change, should all be included within the scope of the present invention.
Claims (10)
1. a kind of greasy dirt remote sensing module, it is characterised in that including:
Long wave ultraviolet line source, the long wave ultraviolet line source can be produced and detect letter to the long wave ultraviolet of water surface irradiation to be measured
Number;
It is capable of the view data of the Real-time Collection water surface to be measured in image data acquiring end, the image data acquiring end;
And greasy dirt identification end, greasy dirt identification end can be it is determined that the image type corresponding to acquired image data be laggard
Row image procossing simultaneously identifies the surface of the water surface to be measured and/or whether there is greasy dirt under water.
2. greasy dirt remote sensing module according to claim 1, it is characterised in that:The long wave ultraviolet line source is optically focused
Type long wave ultraviolet lamp.
3. greasy dirt remote sensing module according to claim 1, it is characterised in that:
The greasy dirt identification end includes:Can it is determined that after image type corresponding to acquired image data, according to really
Fixed image type carries out the image processing part of respective image treatment and can be based at the image that image processing part is obtained
Reason data identify the surface of the water surface to be measured and/or under water with the presence or absence of the greasy dirt identification part of greasy dirt;Wherein described image processing unit
Including:
View data reading unit, the view data reading unit can in real time read image data acquiring end acquired image
Data;
Image takes frame unit, and the image takes frame unit and can take frame mode with constant duration and carry out frame to described image data and carries
Take treatment;
Image pre-processing unit, the image pre-processing unit can be to carrying out image by each view data after frame extraction process
Pretreatment;
Graphics processing unit, the graphics processing unit can determine the image class corresponding to current view data after pretreatment
Whether type is evening images, is, then using the blue channel component of evening images as source images, and to carry out threshold to the source images
Value dividing processing, realization is partitioned into as the image-region of the presence oil spilling of target area and as the back of the body from the source images
The oil-free image-region of scene area;Otherwise determine the image type corresponding to current view data after pretreatment for daytime schemes
Picture, and image threshold treatment is carried out to the image based on oil spilling area pixel characteristic, realization is partitioned into the presence of excessive from described image
The image-region and oil-free image-region of oil.
4. greasy dirt remote sensing module according to claim 3, it is characterised in that:
The image pixel value distribution characteristics that described image processing unit is based on corresponding to current view data after pretreatment is true
Determine the image type corresponding to it.
5. greasy dirt remote sensing module according to claim 3, it is characterised in that:
Described image processing unit can be based on pre-stored Otsu algorithms and source images are gone forward side by side with row threshold division to obtain correspondence
Bianry image, and by pixel value in bianry image for target area is confirmed as in 255 corresponding image regions, pixel value is 0
Background area is confirmed as in region corresponding image region, the image-region and oil-free image-region that there is oil spilling to be partitioned into;
Described Otsu algorithms enter row threshold division process to source images:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the general of the pixel appearance of i
Rate pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target class picture
Plain set C1, wherein C0Pixel composition by gray value between [0, k], C1By pixel groups of the gray value between [k+1, L-1]
Into,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), is determined and is worked asCorresponding k values are optimal segmentation when maximum
Source images are entered row threshold division by threshold value finally according to the threshold value k for obtaining, and obtain bianry image.
6. greasy dirt remote sensing module according to claim 3, it is characterised in that:
Described image processing unit can carry out pixel characteristic analysis to the day images, i.e., successively to each pixel in image
Be compared judgement, if exist in image a certain component in pixel rgb value be all higher than with respect to other two components it is set
Threshold value, then it is the pixel corresponding to the image-region that there is oil spilling to confirm as the pixel, there is oil spilling so as to be partitioned into
Image-region and oil-free image-region.
7. a kind of greasy dirt remote sensing system, it is characterised in that it includes:Greasy dirt remote sensing mould as claimed in claim 1
Block.
8. a kind of greasy dirt remote detecting method, it is characterised in that comprise the following steps:
S1, produce long wave ultraviolet detectable signal to water surface irradiation to be measured;
The view data of S2, the Real-time Collection water surface to be measured;
Image procossing is carried out after S3, the image type for determining corresponding to acquired image data and identify the table of the water surface to be measured
Face and/or under water whether there is greasy dirt.
9. greasy dirt remote detecting method according to claim 7, it is characterised in that:
The S3 includes:
S31, in real time reading acquired image data;
S32, frame mode is taken with constant duration frame extraction process is carried out to described image data;
S33, to carrying out image preprocessing through each view data after frame extraction process;
S34, determine whether the image type corresponding to current view data after pretreatment is evening images, be then by night
The blue channel component of image enters row threshold division treatment to the source images as source images, with from the source images
It is partitioned into the image-region as the presence oil spilling of target area and the oil-free image-region as background area;Otherwise determine
Image type corresponding to current view data after pretreatment is day images, and based on oil spilling area pixel characteristic to the figure
As carrying out image threshold treatment, to be partitioned into the image-region and oil-free image-region that there is oil spilling from described image;
S35, the surface that the water surface to be measured is identified based on the image processing data that S34 is obtained and/or under water whether there is greasy dirt.
10. greasy dirt remote detecting method according to claim 7, it is characterised in that:
S34 includes:Row threshold division is entered to source images based on pre-stored Otsu algorithms to obtain corresponding bianry image, and will
To confirm as target area in 255 corresponding image regions, pixel value is 0 region corresponding image to pixel value in the bianry image
Background area is confirmed as in region, the image-region and oil-free image-region that there is oil spilling to be effectively partitioned into;Described Otsu
Algorithm enters row threshold division process to source images:
The pixel count for setting a certain source images is N, and tonal range is taken as [0, L-1], then gray level is the general of the pixel appearance of i
Rate pi=ni/ N, i=0,1,2 ..., L-1, whileWherein niIt is the pixel count of gray level i;
The pixel of source images is divided into background classes pixel set C by gray value using segmentation threshold variable k0And target class picture
Plain set C1, wherein C0Pixel composition by gray value between [0, k], C1By pixel groups of the gray value between [k+1, L-1]
Into,
Intensity profile probability, the average of source images are based on simultaneouslyThen corresponding C0Average
C1AverageWherein pixel is classified into background classes pixel set C0Probability accumulation and
Pixel is classified into target class pixel set C1Probability accumulation and
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
Summary is various to obtain uk=w0u0+w1u1 (1)
Then its corresponding inter-class variance is
K values successively in the range of [0, L-1] are allowed based on formula (2), it is determined that working asCorresponding k values are optimal segmentation threshold when maximum
Value, row threshold division is entered finally according to the threshold value k for obtaining to image, obtains bianry image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611053096.3A CN106770087B (en) | 2016-11-24 | 2016-11-24 | Greasy dirt remote sensing module, system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611053096.3A CN106770087B (en) | 2016-11-24 | 2016-11-24 | Greasy dirt remote sensing module, system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106770087A true CN106770087A (en) | 2017-05-31 |
CN106770087B CN106770087B (en) | 2019-08-20 |
Family
ID=58911268
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611053096.3A Active CN106770087B (en) | 2016-11-24 | 2016-11-24 | Greasy dirt remote sensing module, system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106770087B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107514626A (en) * | 2017-09-01 | 2017-12-26 | 浙江大唐国际绍兴江滨热电有限责任公司 | A kind of boiler determines blowoff basin draining recovery method and device |
CN109640090A (en) * | 2018-12-18 | 2019-04-16 | 深圳市丰巢科技有限公司 | Picture compression method and device, equipment and storage medium |
CN109883905A (en) * | 2019-03-07 | 2019-06-14 | 南京理工大学 | Underwater partial size on-line measurement system and method |
CN110751645A (en) * | 2019-10-23 | 2020-02-04 | 大连海事大学 | Oil spilling monitoring system and method for oil transportation wharf |
CN110781860A (en) * | 2019-11-05 | 2020-02-11 | 大连海事大学 | Intelligent alarm system and method for monitoring oil spilling of offshore oil platform |
CN111476729A (en) * | 2020-03-31 | 2020-07-31 | 北京三快在线科技有限公司 | Target identification method and device |
WO2020232644A1 (en) * | 2019-05-21 | 2020-11-26 | 唐山哈船科技有限公司 | Oil spill monitoring device near oil well platform and operation method therefor |
WO2020232640A1 (en) * | 2019-05-21 | 2020-11-26 | 唐山哈船科技有限公司 | Sea surface oil spilling monitoring apparatus and operation method |
CN112964437A (en) * | 2021-03-11 | 2021-06-15 | 华能东莞燃机热电有限责任公司 | Oil microleakage recognition method |
CN113055642A (en) * | 2021-02-03 | 2021-06-29 | 天津东泉石油技术开发有限公司 | Oil spilling monitoring device with high sensitivity for ocean platform |
CN113221818A (en) * | 2021-05-27 | 2021-08-06 | 大连海事大学 | Intelligent detection method and system for sea surface oil spill characteristics |
CN113450303A (en) * | 2020-03-24 | 2021-09-28 | 刘金涛 | Oil leakage pre-judging system applying big data storage |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104297176A (en) * | 2014-09-17 | 2015-01-21 | 武汉理工大学 | Device, system and method for monitoring visibility of channel segments of Yangtze River in mountainous area in all-weather manner |
CN104616303A (en) * | 2015-02-11 | 2015-05-13 | 西安电子科技大学 | Ultraviolet image based water surface oil spill detection system and method |
-
2016
- 2016-11-24 CN CN201611053096.3A patent/CN106770087B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104297176A (en) * | 2014-09-17 | 2015-01-21 | 武汉理工大学 | Device, system and method for monitoring visibility of channel segments of Yangtze River in mountainous area in all-weather manner |
CN104616303A (en) * | 2015-02-11 | 2015-05-13 | 西安电子科技大学 | Ultraviolet image based water surface oil spill detection system and method |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107514626A (en) * | 2017-09-01 | 2017-12-26 | 浙江大唐国际绍兴江滨热电有限责任公司 | A kind of boiler determines blowoff basin draining recovery method and device |
CN109640090A (en) * | 2018-12-18 | 2019-04-16 | 深圳市丰巢科技有限公司 | Picture compression method and device, equipment and storage medium |
CN109883905A (en) * | 2019-03-07 | 2019-06-14 | 南京理工大学 | Underwater partial size on-line measurement system and method |
WO2020232644A1 (en) * | 2019-05-21 | 2020-11-26 | 唐山哈船科技有限公司 | Oil spill monitoring device near oil well platform and operation method therefor |
WO2020232640A1 (en) * | 2019-05-21 | 2020-11-26 | 唐山哈船科技有限公司 | Sea surface oil spilling monitoring apparatus and operation method |
CN110751645A (en) * | 2019-10-23 | 2020-02-04 | 大连海事大学 | Oil spilling monitoring system and method for oil transportation wharf |
CN110751645B (en) * | 2019-10-23 | 2023-12-19 | 大连海事大学 | Oil overflow monitoring system and method for oil delivery wharf |
CN110781860A (en) * | 2019-11-05 | 2020-02-11 | 大连海事大学 | Intelligent alarm system and method for monitoring oil spilling of offshore oil platform |
CN110781860B (en) * | 2019-11-05 | 2024-01-19 | 大连海事大学 | Intelligent oil spill monitoring alarm system and method for offshore oil platform |
CN113450303B (en) * | 2020-03-24 | 2023-01-13 | 缘边智联(南京)智能科技有限公司 | Oil leakage pre-judging system applying big data storage |
CN113450303A (en) * | 2020-03-24 | 2021-09-28 | 刘金涛 | Oil leakage pre-judging system applying big data storage |
CN111476729B (en) * | 2020-03-31 | 2023-06-09 | 北京三快在线科技有限公司 | Target identification method and device |
CN111476729A (en) * | 2020-03-31 | 2020-07-31 | 北京三快在线科技有限公司 | Target identification method and device |
CN113055642A (en) * | 2021-02-03 | 2021-06-29 | 天津东泉石油技术开发有限公司 | Oil spilling monitoring device with high sensitivity for ocean platform |
CN112964437A (en) * | 2021-03-11 | 2021-06-15 | 华能东莞燃机热电有限责任公司 | Oil microleakage recognition method |
CN113221818A (en) * | 2021-05-27 | 2021-08-06 | 大连海事大学 | Intelligent detection method and system for sea surface oil spill characteristics |
CN113221818B (en) * | 2021-05-27 | 2024-03-29 | 大连海事大学 | Intelligent detection method and system for sea surface oil spill characteristics |
Also Published As
Publication number | Publication date |
---|---|
CN106770087B (en) | 2019-08-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106770087B (en) | Greasy dirt remote sensing module, system and method | |
CN106407927B (en) | The significance visual method suitable for underwater target detection based on polarization imaging | |
CN103442209B (en) | Video monitoring method of electric transmission line | |
CN101586956B (en) | River water level monitoring method based on monocular camera | |
CN104851086B (en) | A kind of image detecting method for cable surface defect | |
CN108647648A (en) | A kind of Ship Recognition system and method under visible light conditions based on convolutional neural networks | |
CN104182992B (en) | Method for detecting small targets on the sea on the basis of panoramic vision | |
CN107240079A (en) | A kind of road surface crack detection method based on image procossing | |
CN106446926A (en) | Transformer station worker helmet wear detection method based on video analysis | |
CN104537651B (en) | Proportion detecting method and system for cracks in road surface image | |
CN108734111A (en) | SAR image surface vessel recognition methods | |
CN108197583A (en) | The building change detecting method of optimization and image structure feature is cut based on figure | |
CN102654464A (en) | Copper strip surface defect detection system based on multi-feature fuzzy recognition | |
CN109815807A (en) | A kind of ship detecting method of pulling in shore based on edge line analysis and converging channels feature | |
CN104616303A (en) | Ultraviolet image based water surface oil spill detection system and method | |
CN111340951A (en) | Ocean environment automatic identification method based on deep learning | |
CN106339657A (en) | Straw incineration monitoring method and device based on monitoring video | |
CN109801327A (en) | A kind of urban waterlogging depth of accumulated water information extracting method based on video data | |
CN108765406A (en) | A kind of snow mountain detection method based on infrared remote sensing image | |
CN106909925A (en) | A kind of underwater picture object detection method | |
CN105893970A (en) | Nighttime road vehicle detection method based on luminance variance characteristics | |
CN103488992A (en) | Oil spilling detection method for complex SAR image scenes | |
CN103456123B (en) | A kind of video smoke detection method based on flowing with diffusion characteristic | |
CN105184308A (en) | Remote sensing image building detection and classification method based on global optimization decision | |
Marcos et al. | Automated benthic counting of living and non-living components in Ngedarrak Reef, Palau via subsurface underwater video |
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 |