CN104077777B - Sea surface vessel target detection method - Google Patents
Sea surface vessel target detection method Download PDFInfo
- Publication number
- CN104077777B CN104077777B CN201410317291.7A CN201410317291A CN104077777B CN 104077777 B CN104077777 B CN 104077777B CN 201410317291 A CN201410317291 A CN 201410317291A CN 104077777 B CN104077777 B CN 104077777B
- Authority
- CN
- China
- Prior art keywords
- sea
- land
- remote sensing
- connected region
- region
- 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.)
- Active
Links
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a sea surface vessel target detection method which comprises the following steps that (1) a sea-land template automatic partitioning method based on scanning line detecting is used, and a sea-land partitioning template with the same size as an original remote sensing image is generated; (2) the sea-land partitioning template is used for being matched with an original port remote sensing image, and a minimum enclosing rectangle of each communication zone is obtained; and (3) the minimum enclosing rectangles of the communication zones obtained from the step (2) are subjected to screening, and a sea surface vessel target is determined. According to the sea surface vessel target detection method, the obtained sea-land partitioning template is matched with the original remote sensing image, sea surface target separation can be well carried out, sea surface vessel target detection is achieved quickly and accurately, the method is suitable for quick extraction of high-definition remote sensing images under a complex sea-land background, and the problem of invalid pixels caused by image correction in the prior art is avoided. The sea surface vessel target detection method can be widely used in a sea surface vessel target detection process in high-definition port remote sensing images in various civil and military fields.
Description
Technical field
The present invention relates to a kind of detection method, automatically split especially with regard to a kind of extra large land template based on scan line detection
The surface vessel object detection method of method.
Background technology
Utilize remotely-sensed data to carry out Ship Target Detection and suffer from huge realistic meaning in civilian and military field.Along with
The data retrieval capabilities of remote sensing images constantly strengthens and the raising of resolution, utilizes and develops remote sensing image interpretation technology
Extremely the most urgent.Comprehensive domestic and international ongoing research, the naval vessel testing process of optically-based remote sensing images mainly wraps
Include pretreatment (sea land segmentation, cloud and mist rejecting etc.), object candidate area extracts and false-alarm targets rejects three main steps
Suddenly.Wherein the template segmentation of land, sea is important preprocessing means, and i.e. the segmentation automatically of land, sea template is harbour Ship Target
The basis identified.
The existing surface vessel object detection method automatically split based on sea land template is when realizing land, sea and separating, typically
All use based on Threshold segmentation or method based on Texture Segmentation.Enrich in process high-resolution, background complexity, details
During remote sensing images, extra large land based on Threshold segmentation template automatic division method existence sea land boundary alignment precision is low, sea
The problems such as hole easily occur with land area；And extra large land based on Texture Segmentation template is automatically segmented in textural characteristics and carries
When taking, speed is very slow, there is also that positioning precision is poor, problem that hole easily occur in sea and land area simultaneously.Therefore
The remote sensing images enriched for high-resolution, land, sea background complexity, details, existing based on Threshold segmentation with based on texture
Segmentation extra large land template dividing method tend not to obtain preferable result, the most existing detection method can not quickly,
Realize the detection of the surface vessel target of high definition remote sensing images under the background of Complex Sea land exactly.
Summary of the invention
For the problems referred to above, it is an object of the invention to provide one and can realize quickly and accurately under the background of Complex Sea land
The surface vessel object detection method of high definition remote sensing images.
For achieving the above object, the present invention takes techniques below scheme: a kind of surface vessel object detection method, it wraps
Include following steps:
1) use extra large land template automatic division method based on scan line detection, generate an equal amount of with former remote sensing images
1. land, sea segmentation template, comprising: input High Resolution Visible Light harbour remote sensing images to be detected, according to based on picture
Vegetarian noodles amasss relevant resampling processing method, carries out resampling process；2. the harbour remote sensing figure after processing for resampling
Picture, uses water area based on scan line seed points detection method, detection water area seed points；3. according to step
2. the water area initial seed point obtained, utilize region growing algorithm search all gray values in connected region meet with
The point of lower condition: this pixel grey scale with adjoin seed points gray scale difference value within 2 and with initial seed point gray scale difference
Value is within 8, and it is water area that labelling meets the pixel of this condition, and other region is set to land area；4. to step
Rapid 3) region, extra large land obtained carries out binaryzation, will be set to 255 by land area gray value, by water area gray value
Territory is set to 0；5. binary conversion treatment result is carried out Morphological scale-space, i.e. one time etching operation and an expansive working,
Obtain the extra large land segmentation template after harbour remote sensing images use resampling to process；6. to the resampling 5. obtained by step
Extra large land segmentation template after reason, according to the resampling processing method interpolation relevant based on elemental area, obtains and former remote sensing
Image an equal amount of sea land segmentation template；
2) step 1 is utilized) former harbour remote sensing images mate by the extra large land that obtains segmentation template, obtain each connection
Region minimum enclosed rectangle, comprising: 1. by being labeled as the part on land in land, sea segmentation template, in the remote sensing of former harbour
The land area gray value that image is corresponding is set to 0；And the part on sea will be labeled as in land, sea segmentation template, at Yuan Gang
The water area gray value that mouth remote sensing images are corresponding keeps constant, thus obtains water area image；2. in water area
On image, use region growing algorithm that sea water part is marked；3. by binaryzation, the sea water region to labelling
Gray value is set to 0, and other parts gray value is set to 255, obtains the image of sea surface drag；4. connected region is utilized to process
Method, obtains each candidate's sea-surface target connected region；5. the minimum enclosed rectangle method searching connected region is utilized,
Obtain each connected region minimum enclosed rectangle；
3) to through step 2) obtain each connected region minimum enclosed rectangle, screen, determine surface vessel mesh
Mark, comprising: 1. using width, length and the length-width ratio parameter of minimum enclosed rectangle as the parameter of connected region；②
Set the threshold value limit of three characteristic parameters of shape of the width of connected region minimum enclosed rectangle, length-width ratio and connected region
System, the sea separation target connected region meeting three above screening parameter threshold restriction is surface vessel target.
Wherein step 1) step 2. in, the process of described detection water area seed points comprises the following steps: a, from
Upper and lower progressive scan resampling process after harbour remote sensing images, first scan the first row；B, judgement scan in this row
The number of inactive pixels point, wherein N is this row pixel count: if the number of inactive pixels point is more than or equal to N/10, then
This row is set to inactive line, scans next line, return to step b；If the number of inactive pixels point is less than N/10, the most right
This row gray scale is by doing difference after forward direction, i.e. current pixel point gray value deducts the gray value of later pixel, its difference
Absolute value be difference value, be set to 0 by often going the difference value of last pixel；C, investigation difference result, poor
In the result divided, if difference value is less than 2, it is set to 0；D, judge whether the contiguous pixels number that difference value is 0 surpasses
Cross N/5: if it exceeds, then it is assumed that continuous print flat site occurs, is water area, take pixel conduct now
Water area initial seed point, enter step 1) step 3.；Otherwise, scan next line, return to step b, until
Complete the scanning of last column.
Wherein step 3) step 1. or step 2. in, form parameter F of described connected region is defined as follows:
F=| | B | |2/4πA；Wherein, B is the girth of connected region, and A is the area of connected region.
Wherein step 3) step 2. in, width, length-width ratio and the company of described setting connected region minimum enclosed rectangle
The threshold restriction of three characteristic parameters of shape in logical region, is the priori according to Ship Target.
Due to the fact that and take above technical scheme, it has the advantage that 1, the present invention is in surface vessel target detection
During propose a kind of based on scan line detection extra large land template automatic division method, in the method the present invention should
The method detecting seed points with the water area of high definition remote sensing images under the background of Complex Sea land, calmodulin binding domain CaM growth side
The method of method labelling water area, realizes, by binaryzation and Morphological scale-space, the method that land, sea template is split automatically, logical
The method etc. of overweight sampling processing interpolation, has obtained sea an equal amount of with former remote sensing images land segmentation mould rapidly and accurately
Plate.2, the present invention is by using the extra large land that obtains of the inventive method to divide mating of partiting template and former remote sensing images, it is possible to more
Carry out well sea-surface target separation, and then achieve the goal of the invention of surface vessel target detection of the present invention rapidly and accurately.
The rapid extraction of the high definition remote sensing images under 3, the inventive method is not only applicable to Complex Sea land background, and evaded and having defended
Star chart picture corrects the problem bringing inactive pixels.Present invention can be widely used to high definition harbour, various civilian and military field
During surface vessel target detection in remote sensing images.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet that the present invention detects Ship Target
Fig. 2 is present invention High Resolution Visible Light to be detected harbour remote sensing images schematic diagram
Fig. 3 is the schematic flow sheet of present invention water area based on scan line seed points detection method
Fig. 4 is that the present invention obtains sea an equal amount of with former remote sensing images land segmentation template schematic diagram
Detailed description of the invention
With embodiment, the present invention is described in detail below in conjunction with the accompanying drawings.
As it is shown in figure 1, the surface vessel object detection method of the present invention comprises the following steps (as shown in Figure 1):
1) use extra large land template automatic division method based on scan line detection, generate an equal amount of with former remote sensing images
Land, sea segmentation template, comprising:
1. input High Resolution Visible Light harbour remote sensing images (as shown in Figure 2) to be detected, according to prior art based on
The resampling processing method that elemental area is relevant, carries out resampling process；
2. the harbour remote sensing images after processing for resampling, use seed points detection side, water area based on scan line
Method, detection water area seed points, its process is following (as shown in Figure 3):
Harbour remote sensing images after a, the process of progressive scan resampling from top to bottom, first scan the first row；
B, judge to scan inactive pixels point in this row number (inactive pixels point refer to by satellite photo correction cause distant
Some completely black area pixel point in sense image):
If the number of inactive pixels point is more than or equal to N/10, then this row is set to inactive line, scans next line, return to
Step b；Wherein N is this row pixel count；
If the number of inactive pixels point is less than N/10, then to this row gray scale by doing difference after forward direction, i.e. current pixel point ash
Angle value deducts the gray value of later pixel, and the absolute value of its difference is difference value, will often go last pixel
Difference value be set to 0 (this is because for last pixel, there is no thereafter pixel, the most just without subtracting
Number, the difference value the most directly last put is set to 0)；
C, investigation difference result, in the result of difference, if difference value is less than 2, be set to 0；
D, judge that whether contiguous pixels number that difference value is 0 is more than N/5:
If it exceeds, then it is assumed that continuous print flat site occurs, is water area, take pixel now as sea
Region initial seed point, enters next step；
Otherwise, scan next line, return to step b, until completing the scanning of last column.
3. the water area initial seed point 2. obtained according to step, utilizes region growing algorithm to search institute in adjacent domain
Have gray value to meet the point of following condition: this pixel grey scale with adjoin seed points gray scale difference value within 2 and with just
Beginning, seed points gray scale difference value was within 8, and it is water area that labelling meets the pixel of this condition, and other region is set to land
Region, ground；
4. the region, extra large land 3. obtained step carries out binaryzation, will be set to 255, by sea by land area gray value
Area grayscale value is set to 0；
5. binary conversion treatment result is carried out routine Morphological scale-space, i.e. one time etching operation and an expansive working,
Obtain the extra large land segmentation template after using resampling to process；
6. to the extra large land segmentation template 5. obtained by step, according to prior art based on the relevant resampling of elemental area at
Reason method interpolation, obtains sea land segmentation template (as shown in Figure 4) an equal amount of with former remote sensing images.
2) step 1 is utilized) former harbour remote sensing images mate by the extra large land that obtains segmentation template, obtain each connection
Region minimum enclosed rectangle, comprising:
1. by being labeled as the part on land in land, sea segmentation template, in the land area gray scale that former harbour remote sensing images are corresponding
Value is set to 0；And the part on sea will be labeled as in land, sea segmentation template, in the sea district that former harbour remote sensing images are corresponding
Territory gray value keeps constant, thus obtains water area image；
2., on the image of water area, use the region growing algorithm of known technology that sea water part is marked；
3. by binaryzation, the sea water region gray value of labelling being set to 0, other parts gray value is set to 255, obtains
The image of sea surface drag；
4. utilize the connected region processing method of known technology, obtain each candidate's sea-surface target connected region；
5. utilize the minimum enclosed rectangle method of the lookup connected region of known technology, obtain outside each connected region minimum
Connect rectangle；
3) to through step 2) obtain each connected region minimum enclosed rectangle, screen, to determine surface vessel
Target, comprising:
1. using width, length and the length-width ratio parameter of minimum enclosed rectangle as the parameter of connected region；
2. according to priori or the alternate manner of Ship Target, the width of connected region minimum enclosed rectangle, length are set
The wide threshold restriction than three characteristic parameters of shape with connected region, meets three above screening parameter threshold restriction
Separation target connected region in sea is surface vessel target.
The dimensional information of the minimum enclosed rectangle width of above-mentioned connected region, length-width ratio parameter reflection candidate target, connection
The polymerism of the form parameter reflection candidate target region in region, form parameter F is defined as follows:
F=| | B | |2/ 4 π A,
Wherein, B is the girth of connected region, and A is the area of connected region, and form parameter F reflects to a certain extent
The compactedness in region, it does not has dimension, to yardstick, rotationally-varying insensitive, and the value model that neither one is fixing
Enclosing, numerical value is the biggest, and shape is typically got over the compactest regular, rationally selects this parameter, it becomes possible to remove jagged doubtful
Vessel area.
Above-described embodiment is merely to illustrate the present invention, every equivalents carried out on the basis of technical solution of the present invention
And improvement, the most should not get rid of outside protection scope of the present invention.
Claims (3)
1. a surface vessel object detection method, it comprises the following steps:
1) use extra large land template automatic division method based on scan line detection, generate an equal amount of with former remote sensing images
Land, sea segmentation template, comprising:
1. High Resolution Visible Light harbour remote sensing images to be detected are inputted, according to the resampling relevant based on elemental area
Processing method, carries out resampling process；
2. the harbour remote sensing images after processing for resampling, use seed points detection side, water area based on scan line
Method, detection water area seed points, comprise the following steps:
Harbour remote sensing images after a, the process of progressive scan resampling from top to bottom, first scan the first row；
B, judgement scan the number of inactive pixels point in this row, and wherein N is this row pixel count:
If the number of inactive pixels point is more than or equal to N/10, then this row is set to inactive line, scans next line, return to
Step b；
If the number of inactive pixels point is less than N/10, then to this row gray scale by doing difference after forward direction, i.e. current pixel point ash
Angle value deducts the gray value of later pixel, and the absolute value of its difference is difference value, will often go last pixel
Difference value be set to 0；
C, investigation difference result, in the result of difference, if difference value is less than 2, be set to 0；
D, judge that whether contiguous pixels number that difference value is 0 is more than N/5:
If it exceeds, then it is assumed that continuous print flat site occurs, is water area, take pixel now as sea
Region initial seed point, enter step 1) step 3.；
Otherwise, scan next line, return to step b, until completing the scanning of last column；
3. the water area initial seed point 2. obtained according to step, utilizes region growing algorithm to search institute in connected region
Have gray value to meet the point of following condition: this pixel grey scale with adjoin seed points gray scale difference value within 2 and with just
Beginning, seed points gray scale difference value was within 8, and it is water area that labelling meets the pixel of this condition, and other region is set to land
Region, ground；
4. to step 3) region, extra large land that obtains carries out binaryzation, will be set to 255, by sea by land area gray value
Area grayscale codomain is set to 0；
5. binary conversion treatment result is carried out Morphological scale-space, i.e. one time etching operation and an expansive working, obtains port
Template is split in extra large land after mouth remote sensing images use resampling to process；
6. the extra large land segmentation template after processing the resampling 5. obtained by step, according to the weight relevant based on elemental area
Sampling processing method interpolation, obtains sea an equal amount of with former remote sensing images land segmentation template；
2) step 1 is utilized) former harbour remote sensing images mate by the extra large land that obtains segmentation template, obtain each connection
Region minimum enclosed rectangle, comprising:
1. by being labeled as the part on land in land, sea segmentation template, in the land area gray scale that former harbour remote sensing images are corresponding
Value is set to 0；And the part on sea will be labeled as in land, sea segmentation template, in the sea district that former harbour remote sensing images are corresponding
Territory gray value keeps constant, thus obtains water area image；
2., on the image of water area, use region growing algorithm that sea water part is marked；
3. by binaryzation, the sea water region gray value of labelling being set to 0, other parts gray value is set to 255, obtains
The image of sea surface drag；
4. utilize connected region processing method, obtain each candidate's sea-surface target connected region；
5. utilize the minimum enclosed rectangle method searching connected region, obtain each connected region minimum enclosed rectangle；
3) to through step 2) obtain each connected region minimum enclosed rectangle, screen, determine surface vessel mesh
Mark, comprising:
1. using width, length and the length-width ratio parameter of minimum enclosed rectangle as the parameter of connected region；
2. three characteristic parameters of shape of the width of connected region minimum enclosed rectangle, length-width ratio and connected region are set
Threshold restriction, the sea separation target connected region meeting three above screening parameter threshold restriction is surface vessel mesh
Mark.
2. a kind of surface vessel object detection method as claimed in claim 1, it is characterised in that: wherein step 3)
Step 1. or step 2. in, form parameter F of described connected region is defined as follows:
F||B||2/ 4 π A,
Wherein, B is the girth of connected region, and A is the area of connected region.
3. a kind of surface vessel object detection method as claimed in claim 1 or 2, it is characterised in that: wherein step
3) step 2. in, the shape three of width, length-width ratio and the connected region of described setting connected region minimum enclosed rectangle
The threshold restriction of individual characteristic parameter, is the priori according to Ship Target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410317291.7A CN104077777B (en) | 2014-07-04 | 2014-07-04 | Sea surface vessel target detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410317291.7A CN104077777B (en) | 2014-07-04 | 2014-07-04 | Sea surface vessel target detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104077777A CN104077777A (en) | 2014-10-01 |
| CN104077777B true CN104077777B (en) | 2017-01-11 |
Family
ID=51599019
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410317291.7A Active CN104077777B (en) | 2014-07-04 | 2014-07-04 | Sea surface vessel target detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104077777B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108229433A (en) * | 2018-02-01 | 2018-06-29 | 中国电子科技集团公司第十五研究所 | A kind of Inshore ship detection method based on line segment detection and shape feature |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095846B (en) * | 2014-09-28 | 2019-07-30 | 航天恒星科技有限公司 | Region growing seed point extracting method and system towards the segmentation of remote sensing images sea land |
| CN104331886B (en) * | 2014-10-29 | 2017-05-03 | 中国人民解放军国防科学技术大学 | Port region ship and warship detection method based on high resolution SAR image |
| CN104408750B (en) * | 2014-10-30 | 2017-12-05 | 中国科学院深圳先进技术研究院 | A kind of method and system for calculating seawater land ratio |
| CN104966295B (en) * | 2015-06-16 | 2017-10-24 | 武汉大学 | A kind of naval vessel extracting method based on wire-frame model |
| CN105225236B (en) * | 2015-09-21 | 2016-12-28 | 中国科学院半导体研究所 | A kind of bianry image connected region paralleled detection method and system |
| CN105243360A (en) * | 2015-09-21 | 2016-01-13 | 西安空间无线电技术研究所 | Ship object self-organizing cluster method based on distance search |
| CN106056084B (en) * | 2016-06-01 | 2019-05-24 | 北方工业大学 | Remote sensing image port ship detection method based on multi-resolution hierarchical screening |
| CN106780525B (en) * | 2016-11-18 | 2020-08-14 | 北方工业大学 | Optical remote sensing image ship direction feature extraction method based on coordinate rotation minimum circumscribed rectangle |
| CN106709426B (en) * | 2016-11-29 | 2020-04-14 | 上海航天测控通信研究所 | Ship target detection method based on infrared remote sensing image |
| CN107045126A (en) * | 2017-03-06 | 2017-08-15 | 中国人民解放军国防科学技术大学 | A kind of synthetic aperture radar movement overseas Ship Target Detection method |
| CN107301646B (en) * | 2017-06-27 | 2019-09-17 | 深圳市云洲创新科技有限公司 | Unmanned boat intelligent barrier avoiding method and apparatus based on monocular vision |
| CN110334706B (en) * | 2017-06-30 | 2021-06-01 | 清华大学深圳研究生院 | Image target identification method and device |
| CN107862271B (en) * | 2017-10-31 | 2021-05-25 | 西安电子科技大学 | Detection method of ship target |
| CN108647648A (en) * | 2018-05-14 | 2018-10-12 | 电子科技大学 | A kind of Ship Recognition system and method under visible light conditions based on convolutional neural networks |
| CN108921066B (en) * | 2018-06-22 | 2021-12-17 | 西安电子科技大学 | Optical remote sensing image ship detection method based on feature fusion convolutional network |
| CN111368599B (en) * | 2018-12-26 | 2023-11-07 | 北京眼神智能科技有限公司 | Remote sensing image sea surface ship detection method and device, readable storage medium and equipment |
| CN110414343B (en) * | 2019-06-24 | 2022-12-02 | 西安电子科技大学 | On-orbit satellite remote sensing image ship detection method |
| CN111105419B (en) * | 2019-10-09 | 2023-04-18 | 中国船舶重工集团公司第七0九研究所 | Vehicle and ship detection method and device based on polarized SAR image |
| CN110807424B (en) * | 2019-11-01 | 2024-02-02 | 深圳市科卫泰实业发展有限公司 | Port ship comparison method based on aerial image |
| CN111680565B (en) * | 2020-05-08 | 2022-06-07 | 湖北航天技术研究院总体设计所 | Port area ship target detection method based on SAR image |
| CN113408615B (en) * | 2021-06-16 | 2022-04-12 | 中国石油大学(华东) | Automatic ship matching method based on optical satellite remote sensing image |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1727912A (en) * | 2004-07-30 | 2006-02-01 | 中国科学院电子学研究所 | Method for detecting disturbance of ship through biplane of reference |
| CN102855622A (en) * | 2012-07-18 | 2013-01-02 | 中国科学院自动化研究所 | Infrared remote sensing image sea ship detecting method based on significance analysis |
| CN103134476A (en) * | 2013-01-28 | 2013-06-05 | 中国科学院研究生院 | Sea and land boundary detection method based on level set algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7764817B2 (en) * | 2005-08-15 | 2010-07-27 | Siemens Medical Solutions Usa, Inc. | Method for database guided simultaneous multi slice object detection in three dimensional volumetric data |
-
2014
- 2014-07-04 CN CN201410317291.7A patent/CN104077777B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1727912A (en) * | 2004-07-30 | 2006-02-01 | 中国科学院电子学研究所 | Method for detecting disturbance of ship through biplane of reference |
| CN102855622A (en) * | 2012-07-18 | 2013-01-02 | 中国科学院自动化研究所 | Infrared remote sensing image sea ship detecting method based on significance analysis |
| CN103134476A (en) * | 2013-01-28 | 2013-06-05 | 中国科学院研究生院 | Sea and land boundary detection method based on level set algorithm |
Non-Patent Citations (2)
| Title |
|---|
| 基于特征融合的可见光图像舰船检测新方法;尤晓建等;《计算机工程与应用》;20050701(第19期);第199-202页 * |
| 海天背景下红外舰船自动目标识别算法;郭小威等;《激光与红外》;20121231;第42卷(第12期);第1398-1402页 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108229433A (en) * | 2018-02-01 | 2018-06-29 | 中国电子科技集团公司第十五研究所 | A kind of Inshore ship detection method based on line segment detection and shape feature |
| CN108229433B (en) * | 2018-02-01 | 2021-10-26 | 中国电子科技集团公司第十五研究所 | Method for detecting ship landing on shore based on straight-line segment detection and shape characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104077777A (en) | 2014-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104077777B (en) | Sea surface vessel target detection method | |
| CN107527352B (en) | Remote sensing ship target contour segmentation and detection method based on deep learning FCN network | |
| CN109711295B (en) | Optical remote sensing image offshore ship detection method | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| CN103942803B (en) | SAR (Synthetic Aperture Radar) image based automatic water area detection method | |
| CN101515325B (en) | Character extracting method in digital video based on character segmentation and color cluster | |
| CN102279973B (en) | Sea-sky-line detection method based on high gradient key points | |
| CN106845372A (en) | The ship target detection recognition method and system of a kind of space remote sensing optical imagery | |
| CN105741276B (en) | A kind of ship waterline extracting method | |
| CN103198476B (en) | Image detection method of thick line type cross ring mark | |
| CN103020975A (en) | Wharf and ship segmentation method combining multi-source remote sensing image characteristics | |
| CN111476120B (en) | Unmanned aerial vehicle intelligent ship water gauge identification method and device | |
| CN105654091A (en) | Detection method and apparatus for sea-surface target | |
| CN105761286A (en) | Water color exception object extraction method and system based on multi-spectral remote sensing image | |
| CN104063870A (en) | Automatic land and sea template segmentation method based on scanning line detection and application thereof | |
| CN104992172B (en) | Port remote sensing image shore ship detection method based on coastal outburst and different area scanning | |
| CN103530629A (en) | Method and system for detecting extended range of small infrared signal target | |
| Cai et al. | Study on shadow detection method on high resolution remote sensing image based on HIS space transformation and NDVI index | |
| CN103745221A (en) | Two-dimensional code image correction method | |
| CN102054275B (en) | Real-time detection method for sea sky line of gray level image | |
| CN103914829B (en) | Method for detecting edge of noisy image | |
| CN103871072A (en) | Method for automatic extraction of orthoimage embedding line based on projection digital elevation models | |
| CN109961065B (en) | Sea surface ship target detection method | |
| CN110221312A (en) | A method of quickly detecting ground point cloud based on laser radar | |
| CN106709952A (en) | Automatic calibration method of display screen |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |