CN109961065A - A kind of surface vessel object detection method - Google Patents
A kind of surface vessel object detection method Download PDFInfo
- Publication number
- CN109961065A CN109961065A CN201711417359.9A CN201711417359A CN109961065A CN 109961065 A CN109961065 A CN 109961065A CN 201711417359 A CN201711417359 A CN 201711417359A CN 109961065 A CN109961065 A CN 109961065A
- Authority
- CN
- China
- Prior art keywords
- vertical edge
- interest
- area
- image
- sobel
- 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
- 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
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
-
- 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)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of surface vessel object detection method, by calculating the quantity of vertical edge points on Sobel vertical edge figure and come left and right, the upper and lower boundary that obtains area-of-interest, so that it is determined that area-of-interest；Image segmentation is carried out to area-of-interest with the method that profile growth combines based on the segmentation of ranks mean value；The feature of suspected target is calculated on gray level image and bianry image corresponding to area-of-interest, and effectively Ship Target is determined by feature decision.The present invention is wide with adaptation scene, verification and measurement ratio is quasi-, has a very important significance and is worth.
Description
Technical field
The present invention relates to image object detection field, specifically a kind of surface vessel object detection method.
Background technique
Image object detection is an important branch of pattern-recognition, the application in communications and transportation and national defense safety field
It is very extensive.Marine vessel has very high economy and strategic value as important navigation transport facility, especially extra large
Foreign resource growing tension and cause under the situation of local conflicts, control Ship Target is the weight of effective, the rapid control hazard state of affairs
Want means.Mainly extra large day juncture area is detected current surface vessel target detection in the picture, and is had a common boundary to extra large day
Ship Target is extracted only with dividing method in region, to define the regional scope of target detection, the Ship Target of extraction is not
Completely, cause to can't detect Ship Target or the Ship Target of detection is imperfect.Therefore, based on the segmentation of ranks mean value and profile
Growing the surface vessel object detection method combined has important application value, and method relates generally to image procossing and mould
Formula identification.
Summary of the invention
In view of the deficiencies of the prior art, the present invention provides a kind of essence combined based on the segmentation of ranks mean value with profile growth
Really effective surface vessel object detection method.
Present invention technical solution used for the above purpose is:
It is a kind of that the surface vessel object detection method combined, including following step are grown with profile based on the segmentation of ranks mean value
It is rapid:
Step 1: extracting area-of-interest by calculating the quantity of vertical edge points on Sobel vertical edge figure；
Step 2: area-of-interest being carried out with the dividing method that profile growth combines using based on the segmentation of ranks mean value
Segmentation；
The area-of-interest exacting method are as follows:
Step 1: calculating the quantity of each column vertical edge points and on Sobel vertical edge figure to determine area-of-interest
Left and right boundary；
Step 2: calculating the number of every row vertical edge points and on Sobel vertical edge figure to determine area-of-interest
Upper and lower boundary；
Step 3: as feeling emerging by left margin, right margin, coboundary and lower boundary area encompassed in gray level image
Interesting region.
The size of infrared sea image is M × N, and the number of each column vertical edge points is calculated on Sobel vertical edge figure
It is vertical on Sobel vertical edge figure from image left end each column is counted to the right the left and right boundary that determines area-of-interest
Marginal point number and Tleft(j), j=1,2 ..., N, works as Tleft(j) > T0And NumFirstleftWhen=0, Tleft(j) corresponding
Column position j be region left margin BL；It is vertical from the right end of image each column to be counted to the left on Sobel vertical edge figure
Marginal point number and Tright(j), j=N, N-1 ..., 1, work as Tright(j) > T0And NumFirstrightWhen=0, Tright(j) institute
Corresponding column position j is the right margin B in regionR；Wherein, NumFirstleftFor Tleft(j) meet constraint condition Tleft(j) >
T0Number, initial value 0；NumFirstrightFor Tright(j) meet constraint condition Tright(j) > T0Number, initial value is
0；T0=5.If BL> BR, then BL=N/4, BR=N/2.
Calculate the number of every row vertical edge points and on Sobel vertical edge figure to determine the upper and lower of area-of-interest
Two kinds of situations are divided on boundary: there are sea horizon is not deposited in sea horizon and image in image, there are sea horizon, that is, sea horizon inspections in image
It surveys effectively, there is no sea horizon, that is, sea horizon detections in image in vain, the first situation: there are Hai Tian in hot outer sea image
When line, by area-of-interest left margin BL, right margin BRDefined by Sobel vertical edge figure, by the vertical edge of extraction
Make horizontal direction projection, that is, calculate every row marginal point number and；Firstly, calculating every row marginal point from the bottom to top from sea horizon position
Number and Ttop(i), i=K, K-1 ..., 1,1≤K≤M, work as Ttop(i) > 0 and Ttop(i-1)=0 when, line position corresponding to i
It sets as coboundary BT, last line calculates every row marginal point number and T from the bottom to top from imagedown(i), i=M, M-
1 ..., K, works as TdownAnd T (i)=0down(i-1) when > 0, line position corresponding to i is set as lower boundary BB；Second case: when
When sea horizon being not present in image, in the Sobel vertical edge figure as defined by area-of-interest left and right boundary, from image
The first row from top to bottom calculates every row marginal point number and Ttop(i), i=1,2 ..., M, works as Ttop(i) > THAnd
NumFirsttopWhen=0, line position corresponding to i is set as coboundary BT, last line calculates every row from the bottom to top from image
Marginal point number and Tdown(i), i=M, M-1 ..., 1, work as Tdown(i) > THAnd NumFirstdownWhen=0, row corresponding to i
Position is lower boundary BB, wherein NumFirsttopFor Ttop(i) meet constraint condition Ttop(i) > THNumber, initial value is
0, NumFirstdownFor Tdown(i) meet constraint condition Tdown(i) > THNumber, initial value 0；TH=3.If BT> BB, then
BT=M/4, BB=M/2.
By left margin B in gray level imageL, right margin BR, coboundary BTWith lower boundary BBArea encompassed is to feel emerging
Interesting region.
It is described that the dividing method combined is grown with profile based on the segmentation of ranks mean value are as follows:
Area-of-interest having a size of M ' × N ', using the dividing method based on ranks mean value to area-of-interest into
Row binaryzation, R (i, j) are the gray values at any point on image after area-of-interest binaryzation, wherein i=1,
2 ..., M ', j=1,2 ..., N '；If (i, j) is target point, R (i, j)=1；If (i, j) is non-target point, R (i, j)
=0.
To non-targeted point R (i each in image after area-of-interest binaryzation0, j0) judged, wherein R (i0, j0)=
0, if (the i in the image after area-of-interest binaryzation0, j0) left and right and lower direction at least there is a target point: work as j=
1,2 ..., j0When -1, there are R (j0, j)=1, work as j=j0When+1 ..., N ', there are R (i0, j)=1, work as i=i0+ 1 ...,
When M ', there are R (i, j0)-1；And left and right, lower three directions of corresponding position search and pass through wheel in two-value contour images
Current non-targeted point is then set to target point by the same objective contour that exterior feature growth obtains.
The invention has the following beneficial effects and advantage:
1. the present invention is based on the segmentations of ranks mean value can obtain accurate naval vessel region with the method that profile growth combines,
Avoid the naval vessel region obtained more excessive than true naval vessel or too small, differentiate for the calculating of naval vessel feature and naval vessel provide effectively according to
According to the accuracy rate of raising naval vessel detection；
2., mainly complex background under infrared surface vessel of less demanding to the infrared sea picture quality of acquisition of the invention
Target detection；
3. the present invention has, adaptation scene is wide, verification and measurement ratio is quasi-, has a very important significance and is worth.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention；
Fig. 2 is sea horizon detection process figure of the invention；
Fig. 3 is sea horizon testing result figure of the invention；
Fig. 4 is region of interesting extraction procedure chart of the invention；
Fig. 5 is area-of-interest cutting procedure figure of the invention；
Fig. 6 is neighborhood background area schematic of the invention；
Fig. 7 is infrared surface vessel object detection results figure of the invention.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and embodiments.
It is as shown in Figure 1 flow chart of the method for the present invention.
It is a kind of that the surface vessel object detection method combined is grown with profile based on the segmentation of ranks mean value, firstly, using
Method based on Hough transform detects sea horizon；Then, on Sobel vertical edge figure calculate area-of-interest it is left and right,
Upper and lower boundary, so that it is determined that area-of-interest；Further, the side combined is grown with profile using based on the segmentation of ranks mean value
Method carries out Target Segmentation to area-of-interest；Next, being counted on gray level image and bianry image corresponding to area-of-interest
The feature of suspected target is calculated, and effectively Ship Target is determined by feature decision.
1. detecting sea horizon using the method based on Hough transform, specific step is as follows；
Step 1: infrared sea image and vertical direction Sobel operator are done convolution, and two are carried out to filtered image
Value, to obtain Sobel horizontal edge image；
Vertical direction Sobel operator:
Step 2: progress Hough transform is become on Sobel horizontal edge image in dry 10 ° and 0.5 ° of precision
Matrix is changed, obtaining length should be greater than the straight line of 160 pixels, choose longest 3 as candidate straight line；If being chosen less than 3
All straight lines that length is greater than 160 pixels are candidate straight line；If there is no sea horizon, sea horizon testing results in image without if
In vain.
Step 3: calculating the effective length of straight line, method are as follows:
SdotFor number of effective points, LCiFor the continuous points of i-th of line segment, LIjFor j-th of intermittent continuous points.
Step 4: according to 15 row gray variance of region under the straight line above and below equal value difference absolute value of each 15 row area grayscale, straight line
Judge whether all candidate straight lines meet constraint condition
And
If meeting constraint condition, otherwise it is not sea horizon which, which is sea horizon, wherein
NOn=15, NUnder=15, TOn iFor the gray average of 15 the i-th rows of row region on straight line, TUnder iFor 15 row regions under straight line
The gray average of i row,For 15 area grayscale mean value under straight line.
It is illustrated in figure 2 the procedure chart of the method detection sea horizon of the invention based on Hough transform, wherein (a) is red
Outer sea ship images (b) are vertical direction Sobel operator filtering image, (c) are Sobel horizontal edge image, are (d) sea
Antenna detection result.
It is illustrated in figure 3 sea horizon testing result figure of the invention；Wherein (a) is that sea hierarchical scenario is plunged into the commercial sea antenna detection
Effect (b) is plunged into the commercial sea antenna detection effect for backlight situation, (c) is plunged into the commercial sea antenna detection effect for cloudy background.
2. region of interesting extraction
Step 1: infrared sea image and horizontal direction Sobel operator are done convolution, and two are carried out to filtered image
Value, to obtain Sobel vertical edge image；
Step 2: the left and right boundary in zoning, counts to the right each column from image left end on Sobel vertical edge figure
Vertical edge points number and Tleft(j), j=1,2 ..., N, works as Tleft(j) > T0And NumFirstleftWhen=0, Tleft(j) institute
Corresponding column position j is the left margin B in regionL；Each column is counted to the left from the right end of image on Sobel vertical edge figure
Vertical edge points number and Tright(j), j=N, N-1 ..., 1, work as Tright(j) > T0And NumFirstrightWhen=0, Tright
(j) the column position j corresponding to is the right margin B in regionR；Wherein, NumFirstleftFor Tleft(j) meet constraint condition Tleft
(j) > T0Number, initial value 0；NumFirstrightFor Tright(j) meet constraint condition Tright(j) > T0Number, just
Initial value is 0；T0=5.If BL> BR, then BL=N/4, BR=N/2.
Step 3: the upper and lower boundary in zoning, is divided to two kinds of situations: there are do not deposit Hai Tian in sea horizon and image in image
Line, there are sea horizon, that is, sea horizon detections in image effectively, and there is no sea horizon, that is, sea horizon detection is invalid in image.By feeling
Interest region left margin BL, right margin BRDefined by Sobel vertical edge figure, the vertical edge of extraction is made into horizontal direction
Projection, that is, calculate every row marginal point number and；The first situation: there are when sea horizon in hot outer sea image, from sea horizon
Position calculates every row marginal point number and T from the bottom to toptop(i), i=K, K-1 ..., 1,1≤K≤M, work as Ttop(i) > 0 and
Ttop(i-1)=0 when, line position corresponding to i is set as coboundary BT, last line calculates every row side from the bottom to top from image
Edge point number and Tdown(i), i=M, M-1 ..., K work as TdownAnd T (i)=0down(i-1) when > 0, line position corresponding to i is set
As lower boundary BB；Second case: when sea horizon is not present in image, every row side is from top to bottom calculated from image the first row
Edge point number and Ttop(i), i=1,2 ..., M, works as Ttop(i) > THAnd NumFirsttopWhen=0, line position corresponding to i is set i.e.
For coboundary BT, last line calculates every row marginal point number and T from the bottom to top from imagedown(i), i=M, M-1 ..., 1,
Work as Tdown(i) > THAnd NumFirstdownWhen=0, line position corresponding to i is set as lower boundary BB, wherein NumFirsttopFor
Ttop(i) meet constraint condition Ttop(i) > THNumber, initial value 0, NumFirstdownFor Tdown(i) meet constraint condition
Tdown(i) > THNumber, initial value 0；TH=3.If BT> BB, then BT=M/4, BB=M/2.
By left margin B in gray level imageL, right margin BR, coboundary BTWith lower boundary BBArea encompassed is to feel emerging
Interesting region.
It is illustrated in figure 4 region of interesting extraction procedure chart of the invention；Wherein (a) is infrared surface vessel image, (b)
(c) it is Sobel vertical edge image for horizontal direction Sobel operator filtering image, (d) is region of interesting extraction result.
3. region of interest regional partition
Step 1: carrying out binaryzation, area-of-interest ruler to area-of-interest using the dividing method based on ranks mean value
Very little is M ' × N ', and R (i, j) is the gray value at any point on image after area-of-interest binaryzation, wherein i=1,2 ...,
M ', j=1,2 ..., N '；If (i, j) is target point, R (i, j)=1；If (i, j) is non-target point, R (i, j)=0.
Step 2: to non-targeted point R (i each in image after area-of-interest binaryzation0, j0) judged, wherein R
(i0, j0)=0 the, if (i in the image after area-of-interest binaryzation0, j0) left and right and lower direction at least there is a mesh
Punctuate: work as j=1,2 ..., j0When -1, there are R (i0, j)=1, work as j=j0When+1 ..., N ', there are R (i0, j)=1, work as i
=i0When+1 ..., M ', there are R (i, j0)=1；And left and right, lower three directions of corresponding position are equal in two-value contour images
The same objective contour grown by profile is searched, then current non-targeted point is set to target point.
Note: in order to keep segmentation result more acurrate, when realization, retains certain background area, respectively expands in area-of-interest surrounding
Open up 20 pixels.
It is area-of-interest cutting procedure figure of the invention as shown in Figure 5；Wherein (a) is region of interest area image, (b) is
Ranks mean value segmentation result image (c) is contour images, (d) combines segmentation to grow based on the segmentation of ranks mean value with profile
The image result of method.
4. suspected target calculates
Suspected target geometrical characteristic is calculated in area-of-interest segmented image.
Step 1: the method using connected component labeling carries out suspected target label to area-of-interest bianry image, together
When record it is each label target edge pixel coordinate and quantity；
Step 2: passing through the centroid, area, width and the altitude feature that can be calculated suspected target；Then, according to doubtful
The area of target carries out descending arrangement to suspected target all in bianry image.
The average gray of suspected target and neighborhood background is calculated in region of interest area image, wherein neighborhood background is mesh
It marks minimum circumscribed rectangle surrounding 3 and encloses area encompassed, be illustrated in figure 6 neighborhood background area schematic of the invention, solid line
Frame is target minimum circumscribed rectangle.
5. suspected target differentiates
Suspected target needs just be determined as effective Ship Target by feature decision, and feature constraint is as follows:
A. area: doubtful Ship Target pixel sum.
The area of effective target > 100 pixels.
B. the ratio of width to height: the width pixel value of doubtful Ship Target minimum circumscribed rectangle is divided by height pixel value.
The ratio of width to height of effective target is between 0.5~10.
C. contrast: the difference of doubtful Ship Target average brightness and neighborhood background average brightness, neighborhood background therein
The gray average for enclosing region is enclosed by target minimum circumscribed rectangle surrounding 3.
The contrast > 5 (gray scale difference) of effective target.
D. position: the position coordinates of doubtful Ship Target minimum circumscribed rectangle.
Target-recognition is carried out to suspected target and determines true Ship Target, obtains object detection results.
It is illustrated in figure 7 infrared surface vessel object detection results figure of the invention.
Claims (5)
1. a kind of surface vessel object detection method, which comprises the following steps:
Step 1: extracting area-of-interest by calculating the quantity of vertical edge points on Sobel vertical edge figure；
Step 2: area-of-interest being split with the dividing method that profile growth combines using based on the segmentation of ranks mean value.
2. surface vessel object detection method according to claim 1, it is characterised in that: the extraction area-of-interest packet
Include following procedure:
Step 1: on Sobel vertical edge figure calculate each column vertical edge points quantity and come determine area-of-interest a left side,
Right margin；
Step 2: calculated on Sobel vertical edge figure every row vertical edge points number and come determine area-of-interest it is upper,
Lower boundary；
Step 3: by left margin, right margin, coboundary and lower boundary area encompassed being region of interest in gray level image
Domain.
3. surface vessel object detection method according to claim 2, it is characterised in that: described in Sobel vertical edge
The quantity of each column vertical edge points is calculated on figure and includes: come the left and right boundary for determining area-of-interest
The size of infrared sea image is M × N；
Each column vertical edge points number and T are counted to the right from image left end on Sobel vertical edge figureleft(j), j=1,
2 ..., N, works as Tleft(j) < T0And NumFirstleftWhen=0, Tleft(j) the column position j corresponding to is the left margin in region
BL；
Each column vertical edge points number and T are counted to the left from the right end of image on Sobel vertical edge figureright(j), j=
N, N-1 ..., 1, work as Tright(j) > T0 and NumFirstrightWhen=0, Tright(j) the column position j corresponding to is region
Right margin BR；
Wherein, NumFirstleftFor Tleft(j) meet constraint condition Tleft(j) number of > T0, initial value 0；
NumFirstrightFor Tright(j) meet constraint condition Tright(j) > T0Number, initial value 0, T0=5；If BL> BR, then
BL=N/4, BR=N/2.
4. surface vessel object detection method according to claim 2, it is characterised in that: described in Sobel vertical edge
The number of every row vertical edge points is calculated on figure and includes: come the upper and lower boundary for determining area-of-interest
If sea horizon detection is effective, by area-of-interest left margin B there are when sea horizon in infrared sea imageL, the right
Boundary BRDefined by Sobel vertical edge figure, the vertical edge of extraction is made into horizontal direction projection, that is, calculates every row marginal point
Number and；Firstly, calculating every row marginal point number and T from the bottom to top from sea horizon positiontop(i), i=K, K-1 ..., 1,1≤
K≤M works as Ttop(i) > 0 and Ttop(i-1)=0 when, line position corresponding to i is set as coboundary BT, the last line from image
Every row marginal point number and T are calculated from the bottom to topdown(i), i=M, M-1 ..., K work as TdownAnd T (i)=0down(i-1) 0 >
When, line position corresponding to i is set as lower boundary BB；
If sea horizon is not present in infrared sea image, sea horizon detection is invalid, by the left and right boundary institute of area-of-interest
In the Sobel vertical edge figure of restriction, every row marginal point number and T are from top to bottom calculated from image the first rowtop(i), i=1,
2 ..., M, works as Ttop(i) > THAnd NumFirsttopWhen=0, line position corresponding to i is set as coboundary BT, last from image
A line calculates every row marginal point number and T from the bottom to topdown(i), i=M, M-1 ..., 1, work as Tdown(i) > THAnd
BumFirstdownWhen=0, line position corresponding to i is set as lower boundary BB；
Wherein, NumFirsttopFor Ttop(i) meet constraint condition Ttop(i) > THNumber, initial value 0, NumFirstdown
For Tdown(i) meet constraint condition Tdown(i) > THNumber, initial value 0；TH=3；If BT> BB, then BT=M/4, BB=M/
2。
5. surface vessel object detection method according to claim 1, it is characterised in that: described to be carried out to area-of-interest
Segmentation includes following procedure:
Step 1: binaryzation being carried out to area-of-interest, obtains binary image；
Step 2: point non-targeted in binary image being judged, if the non-targeted point is left and right in binary image
At least there is a target point with lower direction, and left and right, lower three directions of corresponding position are searched in two-value contour images
To the same objective contour grown by profile, then current non-targeted point is set to target point；
Step 3: the non-targeted point of each of traversal binary image executes step 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711417359.9A CN109961065B (en) | 2017-12-25 | 2017-12-25 | Sea surface ship target detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711417359.9A CN109961065B (en) | 2017-12-25 | 2017-12-25 | Sea surface ship target detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109961065A true CN109961065A (en) | 2019-07-02 |
| CN109961065B CN109961065B (en) | 2021-05-25 |
Family
ID=67020608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711417359.9A Active CN109961065B (en) | 2017-12-25 | 2017-12-25 | Sea surface ship target detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109961065B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222511A (en) * | 2020-04-13 | 2020-06-02 | 中山大学 | Infrared unmanned aerial vehicle target detection method and system |
| CN112150480A (en) * | 2020-09-17 | 2020-12-29 | 哈尔滨工业大学（威海） | Background segmentation method based on sea-sky-line |
| CN117115082A (en) * | 2023-07-12 | 2023-11-24 | 钛玛科（北京）工业科技有限公司 | Method and equipment for detecting overlap quality of tire |
| CN117557784A (en) * | 2024-01-09 | 2024-02-13 | 腾讯科技（深圳）有限公司 | Target detection method, target detection device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101604383A (en) * | 2009-07-24 | 2009-12-16 | 哈尔滨工业大学 | A kind of method for detecting targets at sea based on infrared image |
| CN103578103A (en) * | 2013-09-24 | 2014-02-12 | 北京环境特性研究所 | Target quick-search method based on navigation attitude correction |
| US20160012606A1 (en) * | 2012-06-14 | 2016-01-14 | International Business Machines Corporation | Multi-cue object detection and analysis |
| CN105303526A (en) * | 2015-09-17 | 2016-02-03 | 哈尔滨工业大学 | Ship target detection method based on coastline data and spectral analysis |
| CN106127728A (en) * | 2016-06-07 | 2016-11-16 | 电子科技大学 | A kind of infrared image sea horizon connected domain detection method under sea and sky background |
| CN108288268A (en) * | 2018-01-23 | 2018-07-17 | 华中科技大学 | Inshore ship detection method in high-resolution remote sensing image based on Projection Analysis |
-
2017
- 2017-12-25 CN CN201711417359.9A patent/CN109961065B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101604383A (en) * | 2009-07-24 | 2009-12-16 | 哈尔滨工业大学 | A kind of method for detecting targets at sea based on infrared image |
| US20160012606A1 (en) * | 2012-06-14 | 2016-01-14 | International Business Machines Corporation | Multi-cue object detection and analysis |
| CN103578103A (en) * | 2013-09-24 | 2014-02-12 | 北京环境特性研究所 | Target quick-search method based on navigation attitude correction |
| CN105303526A (en) * | 2015-09-17 | 2016-02-03 | 哈尔滨工业大学 | Ship target detection method based on coastline data and spectral analysis |
| CN106127728A (en) * | 2016-06-07 | 2016-11-16 | 电子科技大学 | A kind of infrared image sea horizon connected domain detection method under sea and sky background |
| CN108288268A (en) * | 2018-01-23 | 2018-07-17 | 华中科技大学 | Inshore ship detection method in high-resolution remote sensing image based on Projection Analysis |
Non-Patent Citations (3)
| Title |
|---|
| XIHUANNIYA: "连通区域计算", 《CSDN》 * |
| 孙健: "红外序列图像中舰船检测与跟踪算法研究", 《中国科学院沈阳自动化研究所》 * |
| 孙健: "红外序列图像中舰船检测与跟踪算法研究", 《中国科学院自动化研究所》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222511A (en) * | 2020-04-13 | 2020-06-02 | 中山大学 | Infrared unmanned aerial vehicle target detection method and system |
| CN111222511B (en) * | 2020-04-13 | 2020-07-24 | 中山大学 | Infrared unmanned aerial vehicle target detection method and system |
| CN112150480A (en) * | 2020-09-17 | 2020-12-29 | 哈尔滨工业大学（威海） | Background segmentation method based on sea-sky-line |
| CN117115082A (en) * | 2023-07-12 | 2023-11-24 | 钛玛科（北京）工业科技有限公司 | Method and equipment for detecting overlap quality of tire |
| CN117115082B (en) * | 2023-07-12 | 2024-04-05 | 钛玛科（北京）工业科技有限公司 | Method and equipment for detecting overlap quality of tire |
| CN117557784A (en) * | 2024-01-09 | 2024-02-13 | 腾讯科技（深圳）有限公司 | Target detection method, target detection device, electronic equipment and storage medium |
| CN117557784B (en) * | 2024-01-09 | 2024-04-26 | 腾讯科技（深圳）有限公司 | Target detection method, target detection device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109961065B (en) | 2021-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109961065A (en) | A kind of surface vessel object detection method | |
| CN105488454B (en) | Front vehicles detection and ranging based on monocular vision | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| CN104915963B (en) | A kind of detection and localization method for PLCC elements | |
| CN111027446B (en) | Coastline automatic extraction method of high-resolution image | |
| CN102663760B (en) | Location and segmentation method for windshield area of vehicle in images | |
| CN105205488B (en) | Word area detection method based on Harris angle points and stroke width | |
| CN105844655B (en) | A kind of laser rays stripe edge extracting method | |
| CN103996198A (en) | Method for detecting region of interest in complicated natural environment | |
| CN104647893B (en) | A kind of printing chromatography error detection method based on cross hairs | |
| CN110298860B (en) | High pole hydrangea detection count system based on machine vision | |
| CN104598907B (en) | Lteral data extracting method in a kind of image based on stroke width figure | |
| CN107705301A (en) | A kind of highway graticule damage testing method based on unmanned plane highway map picture | |
| CN114331986A (en) | Dam crack identification and measurement method based on unmanned aerial vehicle vision | |
| CN110334692A (en) | A kind of blind way recognition methods based on image procossing | |
| CN109145808A (en) | A kind of tuna recognition methods based on adaptive fish body bending die | |
| CN103866551B (en) | Based on the fabric skew quick detecting method of machine vision | |
| CN116704516B (en) | Visual inspection method for water-soluble fertilizer package | |
| CN109766892A (en) | A kind of ray detection image tagged information character dividing method based on edge detection | |
| CN105335688B (en) | A kind of aircraft model recognition methods of view-based access control model image | |
| CN108288268A (en) | Inshore ship detection method in high-resolution remote sensing image based on Projection Analysis | |
| CN108256518A (en) | Detection method and detection device for character region | |
| CN111415348B (en) | Method for extracting bubble characteristics in automobile brake pipeline | |
| CN111161264B (en) | Method for segmenting TFT circuit image with defects | |
| CN109410233A (en) | A kind of accurate extracting method of high-definition picture road of edge feature constraint |
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 |