CN108363958A - A kind of oil tank detection method based on high-resolution optical remote sensing image - Google Patents
A kind of oil tank detection method based on high-resolution optical remote sensing image Download PDFInfo
- Publication number
- CN108363958A CN108363958A CN201810054662.5A CN201810054662A CN108363958A CN 108363958 A CN108363958 A CN 108363958A CN 201810054662 A CN201810054662 A CN 201810054662A CN 108363958 A CN108363958 A CN 108363958A
- Authority
- CN
- China
- Prior art keywords
- oil tank
- region
- remote sensing
- tank
- image
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 230000003287 optical effect Effects 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 12
- 238000012706 support-vector machine Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 6
- 238000005192 partition Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 2
- 239000000284 extract Substances 0.000 abstract description 2
- 230000011218 segmentation Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 244000144992 flock Species 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
-
- 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
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a kind of oil tank detection techniques based on high-resolution optical remote sensing image, belong to remote sensing image object detection field.Oil tank area to be tested is chosen first, and by supporting vector machine model binaryzation and oil tank suspicious region is shown to image to be detected；Then oil tank suspicious region is further determined whether as oil tank, if the length-width ratio in the region is judged as oil tank region, is otherwise judged as non-tank farm domain close to there are other white pixel unicom regions around 1 and its in a certain range.The geographical location of oil tank is finally mapped to by picture position.The method of the present invention is not necessarily to manual intervention, can detect oil tank automatically, accurately extract oil tank, greatly improves accuracy of detection, meets and applies needs；Efficiently quickly, accuracy rate is high, can detect mass remote sensing image data with the short time.
Description
Technical field
The invention belongs to remote sensing image object detection fields, specifically, referring to that one kind being based on high-resolution optical remote sensing
The oil tank detection technique of image.
Background technology
Currently, with the development of remote sensing technology, the resolution ratio of optical remote sensing image is higher and higher, faces the high-resolution of magnanimity
Rate remote sensing image data needs more efficient processing method.And oil tank is based on distant as important strategic materials reserve means
The oil tank detection of sense image is then of great significance.
The existing general ship detection method based on high-resolution optical remote sensing image mostly uses greatly two ways：1)
Method based on visual interpretation manually chooses the position of oil tank in remote sensing image, can so obtain the accurate location of oil tank；
2) method based on Threshold segmentation differentiates each pixel of image generally according to some gray threshold, then passes through one
A little simple iconology analyses obtain the position of oil tank.
Although first method accuracy of detection is high, detection efficiency is low, extremely relies on manpower, it is difficult to answer on a large scale
With, it is difficult to the remotely-sensed data of magnanimity is handled, the human cost of input is far above the benefit of oil tank detection, it is difficult to which large-scale promotion is answered
With.
The automatic detection and identification of oil tank may be implemented in second method, but its detection result is very poor, it is difficult to reach
Satisfactory precision；And for different remote sensing images, angle, illumination, radiation when being imaged all change, and lead to it
Applicable segmentation threshold is different, and so results in being difficult to the detection that selected unified segmentation threshold carries out oil tank.Such as gynophore
To each all selected different threshold value, that automatic detection and identification for being still difficult to carry out oil tank on a large scale.
Invention content
In order to solve problem above, the present invention proposes a kind of oil tank detection side based on high-resolution optical remote sensing image
Method achievees the purpose that oil tank detects automatically by pattern-recognition and image analysis technology.
The oil tank detection method based on high-resolution optical remote sensing image of the present invention, realization include the following steps：
Step 1 chooses oil tank area to be tested；
After detecting River Bank-Line and port area in remote sensing image, intercept along near river and port area 1000
The image of rice is as oil tank region of interest to be detected.
Step 2 to image to be detected binaryzation and obtains oil tank suspicious region；
The supporting vector machine model that oil tank region and non-tank farm regional partition are opened in training, image to be detected is thrown pixel-by-pixel
Enter into trained vector machine model to differentiate that it is oil tank region or non-tank farm domain, the gray value in oil tank region is assigned to
The gray value of white, non-tank farm domain is assigned to black, so as to obtain the binary map containing only oil tank region and non-tank farm domain
Picture.Unicom regional analysis is carried out to bianry image, each independent white pixel unicom region is oil tank suspicious region.
Step 3 further determines whether as oil tank oil tank suspicious region according to following criterion；
The criterion is：For oil tank suspicious region, if the length-width ratio in the region certain model close to around 1 and its
There are other white pixel unicom regions in enclosing, is then judged as oil tank region, is otherwise judged as non-tank farm domain.
Step 4 is mapped to geographical location by picture position；
After obtaining the pixel coordinate of oil tank in the picture, according to reflecting for pixel coordinate in remote sensing image and geographical coordinate
Relationship is penetrated, the geographical location of each oil tank is positioned, completes the automatic detection of oil tank.
The present invention innovatively proposes a kind of multi-stage cascade oil tank detection method from thick to thin, is utilized first in step 2
Supporting vector machine model carries out the oil tank two-value classification of pixel scale, extracts doubtful oil tank region；Then step 3 is again first
Begin accurately to extract oil tank using morphological image method on the basis of extracting, efficiently quickly, accuracy rate is high.With existing method phase
Than the method for the present invention includes the following advantages and benefits：
1) it is not necessarily to manual intervention, oil tank can be detected automatically；
2) the iconologies analysis method such as the recognition methods of support vector machines isotype and the analysis of unicom area is used, inspection is greatly improved
Precision is surveyed, meets and applies needs；
3) efficiently quickly, mass remote sensing image data can be detected with the short time.
Description of the drawings
Fig. 1 is the oil tank overhaul flow chart the present invention is based on high-resolution optical remote sensing image.
Specific implementation mode
Technical scheme of the present invention is described further with reference to the accompanying drawings and detailed description.
A kind of oil tank detection method based on high-resolution optical remote sensing image of the present invention, as shown in Figure 1, distant with a scape
For feeling image, each step of oil tank of the present invention detection is described in detail.
Step 1: choosing oil tank interest region to be detected in remote sensing image.
It is general near river, harbour to there is a large amount of oil tank to assemble, oil tank is detected in remote sensing image, it is necessary first to
It detects river and port area, after completing the detection at river and harbour, generally extends along the vertical direction in river
Including 1000 meters or so can include all oil tanks, therefore present invention interception is along 1000 meters near river and port area
Remote sensing image as oil tank region of interest to be detected.
In this embodiment, according to water front and harbour information, oil tank area to be tested in remote sensing image, such as Fig. 1 are intercepted out
In (1) shown in.
Step 2: carrying out image binary segmentation to image to be checked using support vector machines (SVM), bianry image is joined
Logical regional analysis, obtains oil tank suspicious region.
This step carries out binary segmentation using trained SVM models to area to be tested image, obtains doubtful tank farm
Domain and non-tank farm domain, detailed process are as follows：
201), the SVM models that training can open oil tank region and non-tank farm regional partition first：It chooses representative
Remote sensing images, therefrom choose out tank section and other parts respectively, record their gray value, then utilize this two
The gray value sample training divided obtains SVM models.The output result of SVM models is oil tank region or non-tank farm domain.
202) after, obtaining trained SVM models, image to be detected is put into pixel-by-pixel in trained SVM models
Differentiate that it is oil tank region or non-tank farm domain, the gray value in oil tank region is assigned to white, the gray value in non-tank farm domain
It is assigned to black, so as to obtain the bianry image containing only oil tank region and non-tank farm domain.
203) unicom regional analysis, is carried out to above-mentioned bianry image, it can be deduced that the area that same grayscale value flocks together
Domain, that is, obtain all some regions for being judged as oil tank region and connecting into, it is each independent as shown in (2) in Fig. 1
Logical region all may be an oil tank.
Step 3: formulating criterion, judge whether unicom region is oil tank.
Since oil tank generally is round and all flocks together in certain rule, so these unicom for being formed by oil tank
The length-width ratio in region should be close to 1, and those are independent, around be also less likely without the part in other unicom regions
For oil tank, so judging those length-width ratios close to there is other white pixel unicom areas around 1 and its in a certain range in the present invention
The white pixel unicom area in domain is oil tank.
Unicom regional analysis is carried out to bianry image, unicom region independent of each other is obtained, in the length according to unicom region
Width than with whether close to each other judge that it, whether really for oil tank, obtains oil tank testing result, as shown in (3) in Fig. 1.
Step 4: corresponding to its geographical coordinate by the image coordinate of oil tank.
In remote sensing image, the mapping relations of pixel coordinate and geographical coordinate are saved, the oil obtained according to step (3)
Tank testing result is mapped to geographical coordinate from its pixel coordinate, as shown in (4) in Fig. 1, completes the automatic detection of oil tank.
Claims (2)
1. a kind of oil tank detection method based on high-resolution optical remote sensing image, which is characterized in that including：
Step 1, oil tank area to be tested is chosen；
After detecting River Bank-Line and port area in remote sensing image, intercept along 1000 meters near river and port area
Image is as oil tank region of interest to be detected；
Step 2, binaryzation is carried out to image to be detected and obtains oil tank suspicious region；
The supporting vector machine model that oil tank region and non-tank farm regional partition are opened in training, image to be detected is put into pixel-by-pixel
Differentiate that it is oil tank region or non-tank farm domain in trained vector machine model, the gray value in oil tank region is assigned to white
The gray value of color, non-tank farm domain is assigned to black, obtains the bianry image containing only oil tank region and non-tank farm domain；To binary map
As carrying out unicom regional analysis, each independent white pixel unicom region is oil tank suspicious region；
Step 3, oil tank suspicious region is further determined whether as oil tank according to following criterion；
The criterion is：For oil tank suspicious region, if the length-width ratio in the region is close to around 1 and its in a certain range
There are other white pixel unicom regions, be then judged as oil tank region, is otherwise judged as non-tank farm domain；
Step 4, geographical location is mapped to by picture position；
After obtaining the pixel coordinate of oil tank in the picture, closed according to the mapping of pixel coordinate in remote sensing image and geographical coordinate
System, positions the geographical location of each oil tank, completes the automatic detection of oil tank.
2. according to the method described in claim 1, it is characterized in that, in the step 2, when Training Support Vector Machines model,
It chooses out tank section and other parts respectively from remote sensing images, records two-part gray value, then utilize two parts
Gray value sample training supporting vector machine model, the output result of model is oil tank region or non-tank farm domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810054662.5A CN108363958B (en) | 2018-01-19 | 2018-01-19 | Oil tank detection method based on high-resolution optical remote sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810054662.5A CN108363958B (en) | 2018-01-19 | 2018-01-19 | Oil tank detection method based on high-resolution optical remote sensing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108363958A true CN108363958A (en) | 2018-08-03 |
| CN108363958B CN108363958B (en) | 2021-10-01 |
Family
ID=63006421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810054662.5A Active CN108363958B (en) | 2018-01-19 | 2018-01-19 | Oil tank detection method based on high-resolution optical remote sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108363958B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210453A (en) * | 2019-06-14 | 2019-09-06 | 中国资源卫星应用中心 | A kind of oil tank amount of storage based on Characteristics of The Remote Sensing Images determines method and system |
| CN113808134A (en) * | 2021-11-19 | 2021-12-17 | 中科星睿科技(北京)有限公司 | Oil tank layout information generation method, oil tank layout information generation device, electronic apparatus, and medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140363084A1 (en) * | 2013-06-05 | 2014-12-11 | Digitalglobe, Inc. | Oil tank farm storage monitoring |
| CN104217196A (en) * | 2014-08-26 | 2014-12-17 | 武汉大学 | A method for detecting automatically a circular oil tank with a remote sensing image |
| US20160026848A1 (en) * | 2014-07-25 | 2016-01-28 | Digitalglobe, Inc. | Global-scale object detection using satellite imagery |
| CN105488481A (en) * | 2015-12-04 | 2016-04-13 | 清华大学 | Detection method |
-
2018
- 2018-01-19 CN CN201810054662.5A patent/CN108363958B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140363084A1 (en) * | 2013-06-05 | 2014-12-11 | Digitalglobe, Inc. | Oil tank farm storage monitoring |
| US20160026848A1 (en) * | 2014-07-25 | 2016-01-28 | Digitalglobe, Inc. | Global-scale object detection using satellite imagery |
| CN104217196A (en) * | 2014-08-26 | 2014-12-17 | 武汉大学 | A method for detecting automatically a circular oil tank with a remote sensing image |
| CN105488481A (en) * | 2015-12-04 | 2016-04-13 | 清华大学 | Detection method |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210453A (en) * | 2019-06-14 | 2019-09-06 | 中国资源卫星应用中心 | A kind of oil tank amount of storage based on Characteristics of The Remote Sensing Images determines method and system |
| CN110210453B (en) * | 2019-06-14 | 2021-06-29 | 中国资源卫星应用中心 | Remote sensing image feature-based oil tank storage capacity determination method and system |
| CN113808134A (en) * | 2021-11-19 | 2021-12-17 | 中科星睿科技(北京)有限公司 | Oil tank layout information generation method, oil tank layout information generation device, electronic apparatus, and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108363958B (en) | 2021-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113469177B (en) | Deep learning-based drainage pipeline defect detection method and system | |
| CN108734143A (en) | A kind of transmission line of electricity online test method based on binocular vision of crusing robot | |
| CN112149667B (en) | Automatic reading method of pointer instrument based on deep learning | |
| Kulkarni et al. | Automatic number plate recognition (anpr) system for indian conditions | |
| CN105512666A (en) | River garbage identification method based on videos | |
| CN111583671B (en) | Millimeter wave radar intersection traffic flow monitoring method and system | |
| CN109977191A (en) | Problem map detection method, device, electronic equipment and medium | |
| CN101576956A (en) | On-line character detection method based on machine vision and system thereof | |
| CN108921099A (en) | Moving ship object detection method in a kind of navigation channel based on deep learning | |
| CN103353988A (en) | Method for evaluating performance of heterogeneous SAR (synthetic aperture radar) image feature matching algorithm | |
| CN101533466B (en) | Image processing method for positioning eyes | |
| CN111914767A (en) | Scattered-pollution enterprise detection method and system based on multi-source remote sensing data | |
| CN108363958A (en) | A kind of oil tank detection method based on high-resolution optical remote sensing image | |
| CN102855485A (en) | Automatic wheat earing detection method | |
| Hu et al. | A study of automatic recognition and localization of pipeline for ground penetrating radar based on deep learning | |
| CN114639064A (en) | Water level identification method and device | |
| CN113378642B (en) | Method for detecting illegal occupation buildings in rural areas | |
| CN117058133A (en) | Hydraulic system oil leakage detection method based on image data | |
| Perry et al. | Panel segmentation: A python package for automated solar array metadata extraction using satellite imagery | |
| CN103745476A (en) | Mobile phone clapboard sand detection method based on line scanning local peak analysis | |
| CN116612461A (en) | Target detection-based pointer instrument whole-process automatic reading method | |
| CN113781513B (en) | Leakage detection method and system for water supply pipeline of power plant | |
| CN115690602A (en) | Mesoscale vortex identification method and system | |
| CN112581424B (en) | Classification extraction method for surface and subsurface defects of optical element | |
| CN112036246B (en) | Construction method of remote sensing image classification model, remote sensing image classification method and system |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211207 Address after: Floor 9, block a, Jiahao International Center, 116 Zizhuyuan Road, Haidian District, Beijing 100097 Patentee after: CHINA HIGHWAY ENGINEERING CONSULTING Corp. Patentee after: Zhongzi Data Co., Ltd Address before: 100089 courtyard 17, Changyun palace, West Third Ring Road, Haidian District, Beijing Patentee before: CHINA HIGHWAY ENGINEERING CONSULTING Corp. |
|
| TR01 | Transfer of patent right |