CN108681691A - A kind of marine ships and light boats rapid detection method based on unmanned water surface ship - Google Patents
A kind of marine ships and light boats rapid detection method based on unmanned water surface ship Download PDFInfo
- Publication number
- CN108681691A CN108681691A CN201810309174.4A CN201810309174A CN108681691A CN 108681691 A CN108681691 A CN 108681691A CN 201810309174 A CN201810309174 A CN 201810309174A CN 108681691 A CN108681691 A CN 108681691A
- Authority
- CN
- China
- Prior art keywords
- ships
- light boats
- water surface
- candidate frame
- target candidate
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of marine ships and light boats rapid detection method based on unmanned water surface ship.Marine ships and light boats detection is one of unmanned water surface ship (unmanned surface vehicle, USV) most important task of vision system.The present invention extracts the marginal information of image first, and establishes " Objective " score function and obtain target candidate frame.Then, the sea horizon in extra large day environment hypograph is detected, target candidate frame is screened based on sea horizon anticipation again.Subsequently, histograms of oriented gradients (Histograms of Oriented Gradient, HOG) feature modeling is carried out to ships and light boats target, using support vector machines, using " boot strap " repetitive exercise grader.Finally, the Feature Descriptor of target candidate frame is input in grader, carries out ships and light boats detection.Compared with traditional detection method, detection method provided by the invention can more quickly and accurately detect marine ships and light boats target, and have higher verification and measurement ratio, also have stronger robustness to the variation of scale and illumination condition.
Description
Technical field
The present invention relates to unmanned water surface ship target detection techniques, and in particular to a kind of marine ships and light boats based on unmanned water surface ship
Rapid detection method.
Background technology
Unmanned water surface ship (unmanned surface vehicle, USV) is a kind of novel marine intelligent body, Ke Yiyong
To execute the civilian task such as the military missions such as scouting, antisubmarine, patrol and search and rescue, navigation, hydro_geography prospecting.Wherein, nobody
The effect of water surface ship vision system is that human eye is replaced to be detected, track and measure marine target and barrier, is gone forward side by side
The understanding of row scene and behavior.View-based access control model marine ships and light boats detection be the most important task of unmanned water surface ship vision system it
One, it is the basis realized unmanned water surface ship and marine ships and light boats are identified and are tracked.Therefore the characteristic model of marine ships and light boats is studied
And object detection method, it is of great significance to the development of unmanned water surface ship.
There are one common ground for traditional algorithm of target detection, have been all made of " sliding window formula " search strategy.This strategy is logical
It crosses and grader is slided to traversal on each the window's position of image, to detect the position of target in the picture.Sliding window
Quantity and the detection scale of grader be linearly related.Under single scale, every Image Classifier is probably needed
Test 104-105A window, under multiple dimensioned, the quantity of test window can be increased with several orders of magnitude.In addition, detection now
Device also requires to predict the ratio of width to height of target, then test window quantity will reach 106-107It is a.Obviously, this " poor
The detection method of formula to the greatest extent " can generate many redundancy windows, cause computationally intensive and take very much.So much in this way
Object detection system, can all select some better simply graders.These simple graders often use weaker feature
Model obtains faster calculating speed to make up the drawbacks of sliding window formula search strategy is brought.But although using weak characteristic model
The calculating speed for improving grader, but has lost verification and measurement ratio and accuracy of detection.
By analyzing above, regarded if the detection algorithm of these sliding window formula search strategies is grafted directly to unmanned water surface ship
Feel and carry out ships and light boats detection in detecting system, and computational efficiency is made up using simple grader, then the verification and measurement ratio of ships and light boats
It will be substantially reduced with accuracy of detection, while will produce more wrong reports.If using more complicated grader, although can be promoted
Verification and measurement ratio reduces wrong report, but " exhaustive " feature of sliding window formula search strategy can be such that entire detection process takes very much.This
Sample, the ships and light boats location information that unmanned water surface ship detects will lose promptness, certain to be generated to the execution of follow-up work
Influence.
Invention content
Goal of the invention is for the defects in the prior art and water surface feature, to provide a kind of based on unmanned water surface ship
Marine ships and light boats rapid detection method can more quickly and accurately detect marine ships and light boats target, and have higher detection
Rate also has stronger robustness to the variation of scale and illumination condition.
In order to achieve the above objectives, the present invention uses following technical proposals：
A kind of marine ships and light boats rapid detection method based on unmanned water surface ship, is grasped using unmanned water surface ship vision system
Make, which includes camera, image pick-up card, industrial personal computer；Wherein, camera be mounted on unmanned water surface ship just on
Side, industrial personal computer are fixed in the ship cabin of unmanned water surface ship, and camera is connected to industrial personal computer by 1394 interfaces of IEEE
On, image pick-up card is connected to by pci card slot on industrial personal computer, and this method operating procedure is as follows：
(1) down-sampled to the image progress of camera acquisition, obtain down-sampled image；
(2) edge detector is utilized, the skirt response of each pixel in original image is obtained, by these skirt response groups
It is combined to obtain the edge graph of original image；
(3) " Objective " score function is established, target candidate frame is screened in edge graph；
(4) sea horizon is detected, target candidate frame is screened based on sea horizon anticipation again；
(5) histograms of oriented gradients carried out to ships and light boats target, HOG feature modelings, obtaining one complicated has 5796 dimensions
Feature vector；
(6) support vector machines is utilized, using " boot strap " repetitive exercise grader；
(7) Feature Descriptor of the target candidate frame after screening is input in grader, ships and light boats detection is carried out, if deposited
In ships and light boats, then the location information of ships and light boats in the picture is exported.
Compared with prior art, the beneficial effects of the invention are as follows：
The detection speed of method provided by the invention can reach the promotion of several orders of magnitude, thus can more quickly,
Marine ships and light boats target is accurately detected, while there is higher verification and measurement ratio.In addition, the variation to scale and illumination condition also has
There is stronger robustness.
Description of the drawings
By reading with reference to the following drawings and to being described in detail made by non-limiting embodiment, other spies of the invention
Sign, objects and advantages will become more apparent upon：
Fig. 1 is the system structure diagram of the present invention；
Fig. 2 is the principle of the present invention block diagram；
Fig. 3 is that the present invention obtains target candidate frame schematic diagram；
Fig. 4 is that the present invention is based on sea horizons to screen target candidate frame schematic diagram；
Fig. 5 is classifier training flow chart of the present invention.
Specific implementation mode
Technical solution of the present invention is described in detail below, but protection scope of the present invention is not limited to the implementation
Example.
As shown in Figure 1, the present invention is operated using unmanned water surface ship vision system, the vision system include camera,
The equipment such as image pick-up card, industrial personal computer.Wherein, camera is mounted on the surface of unmanned water surface ship apart from about 1.5 meters of deflection arch portion
Position, industrial personal computer is fixed in the ship cabin of unmanned water surface ship.Camera is connected to industry control by 1394 interfaces of IEEE
On machine, image pick-up card is connected to by pci card slot on industrial personal computer.
As shown in Fig. 2, the present invention provides a kind of marine ships and light boats rapid detection method based on unmanned water surface ship.First,
The marginal information of image is extracted, and establishes " Objective " score function and obtains target candidate frame.Then, to extra large day environment hypograph
In sea horizon be detected, based on sea horizon anticipation again screen target candidate frame.Subsequently, to ships and light boats target into line direction
Histogram of gradients (Histograms of Oriented Gradient, HOG) feature modeling, utilizes support vector machines
(Support Vector Machine, SVM), using " boot strap " repetitive exercise grader.Finally, by the spy of target candidate frame
Sign description is input in grader, carries out ships and light boats detection.Compared with traditional detection method, detection method provided by the invention
It can more quickly and accurately detect marine ships and light boats target, and there is higher verification and measurement ratio, to scale and illumination condition
Variation also has stronger robustness.
The above-mentioned marine ships and light boats rapid detection method based on unmanned water surface ship, specifically includes following steps:
(1) down-sampled to the image progress of camera acquisition, the image drop of 1440 × 1080 pixels collected is adopted
Sample is handled to 640 × 480 pixels；
(2) as shown in figure 3, using edge detector, the skirt response of each pixel in original image is obtained, by these
Skirt response is combined to obtain the edge graph of original image, and the edge graph directly obtained in this way is relatively close, passes through
The local maximum that a non-maxima suppression (Non-Maximal Suppression, NMS) obtains skirt response is executed, from
And a relatively sparse edge graph is obtained, as shown in Fig. 3 (b).
(3) " Objective " score function is established, target candidate frame is screened in edge graph.The specific steps are：
The set s of given edge groupi∈ S calculate the similarity between each pair of neighboring edge group.For edge group
siAnd sj, the similarity a (s between themi,sj) calculation formula is as follows：
a(si,sj)=| cos (θi-θij)cos(θj-θij)|γ
In formula, θiAnd θjIt is the mean direction of the edges Liang Ge group, θijIt is their mean place xiAnd xjBetween angle
Degree.γ values are the sensibility to similarity for control direction, and γ=2 are taken in this method.
The set S of given edge group, and they are calculated between any two after similarity, it is commented by establishing one
Divide the b scorings of function pair boundary candidate frame.Calculate edge group siThe marginal value summation of middle all pixels p, is denoted as mi.Choose edge
Group siIn any one pixel position, be denoted as
For each edge group si, calculate a successive value wb(si) ∈ [0,1], for weighing siWhether wrap completely
It is contained in bounding box b.wb(si) calculation formula it is as follows：
In formula, T is to start from t1∈Sb, end at t|T|=siOrdered path, a (tj,tj+1) between edge group
Similarity.If there is no such path, w is enabledb(si)=1.
Utilize the w for calculating gainedb(si), it is as follows to the scoring formula of bounding box b：
In formula, bw、bhIt is the width and height of bounding box, w respectivelyb(si) ∈ [0,1] measurements siWhether side is completely contained in
In boundary frame b, miIt is edge group siThe marginal value summation of middle all pixels p.Since the bounding box of bigger can include more sides
Edge, this method take κ=1.5 to offset this deviation.
Since the edge inside bounding box is in the edge near bounding box compared to those, importance will be come low.
To scoring formula be improved, by the marginal value inside bounding box from scoring hbIn cut, improved scoring formula is as follows：
In formula, bw、bhIt is the width and height of bounding box, b respectivelyinWidth and height be respectively bw/ 2 and bh/ 2, mpFor
The marginal value size of each pixel p, similarly takes κ=1.5 in edge graph.Finally, 1000 larger bounding boxes of scoring are chosen to make
For target candidate frame.
(4) as shown in figure 4, sea horizon divides an image into three regions：Sky areas, water area and sea horizon area
Domain.Ships and light boats ride the sea, and can only be in water area and sea horizon region, be not at sky areas.Based on such a
Characteristic, the present invention are further improved target candidate frame generation method, are carried out again to 1000 larger target candidate frames of scoring
Screening, weeds out the target candidate frame for being completely in sea horizon overlying regions, as shown in the red boxes in Fig. 4 (a).At reservation
The target candidate frame intersected below sea horizon and with sea horizon, as shown in the green box in Fig. 4 (a).By at edge
A simple Hough variations detection sea horizon is executed on figure, and target candidate frame is screened based on above-mentioned sea horizon anticipation again.
(5) present invention carries out histograms of oriented gradients (Histograms of Oriented to ships and light boats target
Gradient, HOG) the ratio of width to height of ships and light boats characteristic model is designed as 3 by feature modeling according to the sShape features of ships and light boats:1, feature
Window is sized so as to 192 × 64 pixels.The cell factory lattice of HOG features are designed and sized to 8 × 8 pixels.Each cell list
The histogram number of active lanes of first lattice is set as 9.In this way, the intrinsic dimensionality V calculation formula of the HOG Feature Descriptors of ships and light boats are such as
Under：
(6) as shown in figure 5, the present invention using linear kernel support vector machines (Support Vector Machine,
SVM), using " boot strap " repetitive exercise grader.Specific training step is as follows：First, initial positive sample is by ships and light boats
All true value frame (Ground Truth) compositions, sum are 2000.Then, selection accounts for 20%- with true value frame overlapping area
50% target candidate frame is as initial negative sample.In order to avoid choosing the negative sample approximately repeated, overlapping area is surpassed
70% two negative samples are crossed, one of abandon is selected.Finally, 10000 are randomly selected from all negative samples is used as SVM
Trained negative sample.After obtaining preliminary classification device, the process of a retraining is carried out.By preliminary classification device in negative sample original
The detection of figure (not including ships and light boats target) enterprising ship target of navigating, all rectangle frames detected in this way belong to wrong report (False
Positives).The rectangle frame of these wrong reports is a difficult example (Hard Example) for grader.These difficult examples are preserved
For image, it is added in initial negative sample set, re-starts the training of grader.In this way, point obtained by retraining
Class device just has better classification capacity, that is, the ability of detection ships and light boats target.The process of retraining is can be carried out with iteration
, until the performance of grader is not obviously improved.
(7) Feature Descriptor of the target candidate frame after screening is input in grader, carries out ships and light boats detection.If deposited
In ships and light boats, then the location information of ships and light boats in the picture is exported.
Claims (5)
1. a kind of marine ships and light boats rapid detection method based on unmanned water surface ship, is grasped using unmanned water surface ship vision system
Make, which includes camera, image pick-up card, industrial personal computer；Wherein, camera be mounted on unmanned water surface ship just on
Side, industrial personal computer are fixed in the ship cabin of unmanned water surface ship, and camera is connected to industrial personal computer by 1394 interfaces of IEEE
On, image pick-up card is connected to by pci card slot on industrial personal computer, which is characterized in that this method operating procedure is as follows：
(1) down-sampled to the image progress of camera acquisition, obtain down-sampled image；
(2) edge detector is utilized, the skirt response of each pixel in original image is obtained, these skirt responses combination is existed
The edge graph of original image is obtained together；
(3) " Objective " score function is established, target candidate frame is screened in edge graph；
(4) sea horizon is detected, target candidate frame is screened based on sea horizon anticipation again；
(5) histograms of oriented gradients is carried out to ships and light boats target, HOG feature modelings obtain a complicated spy with 5796 dimensions
Sign vector；
(6) support vector machines is utilized, using " boot strap " repetitive exercise grader；
(7) Feature Descriptor of the target candidate frame after screening is input in grader, ships and light boats detection is carried out, if there is ship
Ship then exports the location information of ships and light boats in the picture.
2. a kind of marine ships and light boats rapid detection method based on unmanned water surface ship according to claim 1, it is characterised in that：
" Objective " score function is established in the step (3), target candidate frame is screened in edge graph is specially：
The set s of given edge groupi∈ S calculate the similarity between each pair of neighboring edge group, for edge group siWith
sj, the similarity a (s between themi,sj) calculation formula is as follows：
a(si,sj)=| cos (θi-θij)cos(θj-θij)|γ
In formula, θiAnd θjIt is the mean direction of the edges Liang Ge group, θijIt is their mean place xiAnd xjBetween angle, γ
Value is the sensibility to similarity for control direction, and γ=2 are taken in this method；
The set S of given edge group, and they are calculated between any two after similarity, by establishing a scoring letter
It is several to score boundary candidate frame b, calculate edge group siThe marginal value summation of middle all pixels p, is denoted as mi；Choose edge group
siIn any one pixel position, be denoted as
For each edge group si, calculate a successive value wb(si) ∈ [0,1], for weighing siWhether side is completely contained in
In boundary frame b；wb(si) calculation formula it is as follows：
In formula, T is to start from t1∈Sb, end at t|T|=siOrdered path, a (tj,tj+1) similar between edge group
Degree；If there is no such path, w is enabledb(si)=1；
Utilize the w for calculating gainedb(si), it is as follows to the scoring formula of bounding box b：
In formula, bw、bhIt is the width and height of bounding box, w respectivelyb(si) ∈ [0,1] measurements siWhether bounding box b is completely contained in
In, miIt is edge group siThe marginal value summation of middle all pixels p；Since the bounding box of bigger can include more edges, we
Method takes κ=1.5 to offset this deviation；
Since the edge inside bounding box is in the edge near bounding box compared to those, importance will be come low；To commenting
Point formula is improved, by the marginal value inside bounding box from scoring hbIn cut, improved scoring formula is as follows：
In formula, bw、bhIt is the width and height of bounding box, b respectivelyinWidth and height be respectively bw/ 2 and bh/ 2, mpFor edge
The marginal value size of each pixel p, similarly takes κ=1.5 in figure；Finally, target candidate frame is screened in edge graph, chooses 1000
The larger bounding box of a scoring is as target candidate frame.
3. a kind of marine ships and light boats rapid detection method based on unmanned water surface ship according to claim 1, it is characterised in that：
Sea horizon is detected in the step (4), screening target candidate frame is specially again based on sea horizon anticipation：
The detection of sea horizon is by executing a Hough variation on edge graph, to obtain the position of sea horizon in the picture
Confidence ceases；Ships and light boats ride the sea, and can only be in water area and sea horizon region, be not at sky areas；Based in this way
One characteristic screens 1000 larger target candidate frames of scoring, weeds out and be completely on sea horizon region again
The target candidate frame of side retains the target candidate frame intersected below sea horizon and with sea horizon.
4. a kind of marine ships and light boats rapid detection method based on unmanned water surface ship according to claim 1, it is characterised in that：
Carrying out histograms of oriented gradients feature modeling to ships and light boats target in the step (5) is specially：
According to the sShape features of ships and light boats, the ratio of width to height of ships and light boats characteristic model is designed as 3:1, characteristic window is sized so as to
The cell factory lattice of 192 × 64 pixels, HOG features are designed and sized to 8 × 8 pixels, the histogram channel of each cell factory lattice
Number is set as 9, in this way, the intrinsic dimensionality V calculation formula of the HOG Feature Descriptors of ships and light boats are as follows：
5. a kind of marine ships and light boats rapid detection method based on unmanned water surface ship according to claim 1, it is characterised in that：
It is specially using support vector machines repetitive exercise grader in the step (6)：
Using the support vector machines of linear kernel, using " boot strap " repetitive exercise grader, specific training step is as follows：First,
Initial positive sample is made of all true value frames of ships and light boats, and sum is 2000；Then, selection is accounted for true value frame overlapping area
The target candidate frame of 20%-50% is as initial negative sample, in order to avoid choosing the negative sample approximately repeated, by faying surface
Product is more than 70% two negative samples, selects one of abandon；Finally, 10000 works are randomly selected from all negative samples
For the negative sample of SVM training, after obtaining preliminary classification device, the process of a retraining is carried out；By preliminary classification device negative
Sample artwork, that is, do not include the detection of the enterprising ship target of navigating of ships and light boats target, and all rectangle frames detected in this way belong to wrong report；
The rectangle frame of these wrong reports is a difficult example for grader, these hardly possible examples are saved as image, are added to initial negative sample
In set, the training of grader is re-started, in this way, just there is better classification capacity by the grader that retraining obtains,
Namely detect the ability of ships and light boats target, the process of retraining can iteration carry out, until the performance of grader is not bright
Until aobvious promotion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810309174.4A CN108681691A (en) | 2018-04-09 | 2018-04-09 | A kind of marine ships and light boats rapid detection method based on unmanned water surface ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810309174.4A CN108681691A (en) | 2018-04-09 | 2018-04-09 | A kind of marine ships and light boats rapid detection method based on unmanned water surface ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108681691A true CN108681691A (en) | 2018-10-19 |
Family
ID=63800769
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810309174.4A Pending CN108681691A (en) | 2018-04-09 | 2018-04-09 | A kind of marine ships and light boats rapid detection method based on unmanned water surface ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108681691A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902692A (en) * | 2019-01-14 | 2019-06-18 | 北京工商大学 | A kind of image classification method based on regional area depth characteristic coding |
| CN110414413A (en) * | 2019-07-25 | 2019-11-05 | 北京麒麟智能科技有限公司 | A kind of logistics trolley pedestrian detection method based on artificial intelligence |
| ES2912040A1 (en) * | 2020-11-24 | 2022-05-24 | Iglesias Rodrigo Garcia | Delivery system of a consumer good (Machine-translation by Google Translate, not legally binding) |
| CN114863373A (en) * | 2022-04-19 | 2022-08-05 | 华南理工大学 | Offshore unmanned platform monitoring method and offshore unmanned platform |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101533466A (en) * | 2009-04-09 | 2009-09-16 | 南京壹进制信息技术有限公司 | Image processing method for positioning eyes |
| CN201853971U (en) * | 2010-11-16 | 2011-06-01 | 中国科学院沈阳自动化研究所 | Mobile robot suitable for inspection for large-span power transmission lines |
| CN102663348A (en) * | 2012-03-21 | 2012-09-12 | 中国人民解放军国防科学技术大学 | Marine ship detection method in optical remote sensing image |
| CN102998001A (en) * | 2012-12-18 | 2013-03-27 | 四川九洲电器集团有限责任公司 | Target detection system |
| CN103198332A (en) * | 2012-12-14 | 2013-07-10 | 华南理工大学 | Real-time robust far infrared vehicle-mounted pedestrian detection method |
| CN103544502A (en) * | 2013-10-29 | 2014-01-29 | 上海市城市建设设计研究总院 | High-resolution remote-sensing image ship extraction method based on SVM |
| CN104239854A (en) * | 2014-08-30 | 2014-12-24 | 电子科技大学 | Pedestrian feature extraction and representing method based on region sparse integration passage |
| US20150213059A1 (en) * | 2014-01-29 | 2015-07-30 | Raytheon Company | Method for detecting and recognizing boats |
| CN105022990A (en) * | 2015-06-29 | 2015-11-04 | 华中科技大学 | Water surface target rapid-detection method based on unmanned vessel application |
| CN105930803A (en) * | 2016-04-22 | 2016-09-07 | 北京智芯原动科技有限公司 | Preceding vehicle detection method based on Edge Boxes and preceding vehicle detection device thereof |
| CN106022307A (en) * | 2016-06-08 | 2016-10-12 | 中国科学院自动化研究所 | Remote sensing image vessel detection method based on vessel rotation rectangular space |
| CN106981071A (en) * | 2017-03-21 | 2017-07-25 | 广东华中科技大学工业技术研究院 | A kind of method for tracking target applied based on unmanned boat |
| CN107016391A (en) * | 2017-04-14 | 2017-08-04 | 中国科学院合肥物质科学研究院 | A kind of complex scene workpiece identification method |
-
2018
- 2018-04-09 CN CN201810309174.4A patent/CN108681691A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101533466A (en) * | 2009-04-09 | 2009-09-16 | 南京壹进制信息技术有限公司 | Image processing method for positioning eyes |
| CN201853971U (en) * | 2010-11-16 | 2011-06-01 | 中国科学院沈阳自动化研究所 | Mobile robot suitable for inspection for large-span power transmission lines |
| CN102663348A (en) * | 2012-03-21 | 2012-09-12 | 中国人民解放军国防科学技术大学 | Marine ship detection method in optical remote sensing image |
| CN103198332A (en) * | 2012-12-14 | 2013-07-10 | 华南理工大学 | Real-time robust far infrared vehicle-mounted pedestrian detection method |
| CN102998001A (en) * | 2012-12-18 | 2013-03-27 | 四川九洲电器集团有限责任公司 | Target detection system |
| CN103544502A (en) * | 2013-10-29 | 2014-01-29 | 上海市城市建设设计研究总院 | High-resolution remote-sensing image ship extraction method based on SVM |
| US20150213059A1 (en) * | 2014-01-29 | 2015-07-30 | Raytheon Company | Method for detecting and recognizing boats |
| CN104239854A (en) * | 2014-08-30 | 2014-12-24 | 电子科技大学 | Pedestrian feature extraction and representing method based on region sparse integration passage |
| CN105022990A (en) * | 2015-06-29 | 2015-11-04 | 华中科技大学 | Water surface target rapid-detection method based on unmanned vessel application |
| CN105930803A (en) * | 2016-04-22 | 2016-09-07 | 北京智芯原动科技有限公司 | Preceding vehicle detection method based on Edge Boxes and preceding vehicle detection device thereof |
| CN106022307A (en) * | 2016-06-08 | 2016-10-12 | 中国科学院自动化研究所 | Remote sensing image vessel detection method based on vessel rotation rectangular space |
| CN106981071A (en) * | 2017-03-21 | 2017-07-25 | 广东华中科技大学工业技术研究院 | A kind of method for tracking target applied based on unmanned boat |
| CN107016391A (en) * | 2017-04-14 | 2017-08-04 | 中国科学院合肥物质科学研究院 | A kind of complex scene workpiece identification method |
Non-Patent Citations (6)
| Title |
|---|
| C. LAWRENCE ZITNICK 等: "Edge Boxes: Locating Object Proposals from Edges", 《ECCV 2014》 * |
| MASIKKK: "用初次训练的SVM+HOG分类器在负样本原图上检测HardExample", 《HTTPS://BLOG.CSDN.NET/MASIBUAA/ARTICLE/DETAILS/16113373》 * |
| MATEJ KRISTAN 等: "Fast Image-Based Obstacle Detection From Unmanned Surface Vehicles", 《IEEE TRANSACTIONS ON CYBERNETICS》 * |
| SHENGXIANG QI 等: "Unsupervised Ship Detection Based on Saliency and S-HOG Descriptor From Optical Satellite Images", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 * |
| 安博文 等: "基于 Hough 变换的海天线检测算法研究", 《红外技术》 * |
| 李小毛 等: "基于 3D 激光雷达的无人水面艇海上目标检测", 《上海大学学报（自然科学版）》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902692A (en) * | 2019-01-14 | 2019-06-18 | 北京工商大学 | A kind of image classification method based on regional area depth characteristic coding |
| CN110414413A (en) * | 2019-07-25 | 2019-11-05 | 北京麒麟智能科技有限公司 | A kind of logistics trolley pedestrian detection method based on artificial intelligence |
| ES2912040A1 (en) * | 2020-11-24 | 2022-05-24 | Iglesias Rodrigo Garcia | Delivery system of a consumer good (Machine-translation by Google Translate, not legally binding) |
| CN114863373A (en) * | 2022-04-19 | 2022-08-05 | 华南理工大学 | Offshore unmanned platform monitoring method and offshore unmanned platform |
| CN114863373B (en) * | 2022-04-19 | 2024-06-04 | 华南理工大学 | Marine unmanned platform monitoring method and marine unmanned platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107016357B (en) | Video pedestrian detection method based on time domain convolutional neural network | |
| CN101981582B (en) | Method and apparatus for detecting object | |
| CN103886325B (en) | Cyclic matrix video tracking method with partition | |
| CN109711288A (en) | Remote sensing ship detecting method based on feature pyramid and distance restraint FCN | |
| CN108681691A (en) | A kind of marine ships and light boats rapid detection method based on unmanned water surface ship | |
| Zhang et al. | Multi-scale adversarial network for vehicle detection in UAV imagery | |
| CN105022990A (en) | Water surface target rapid-detection method based on unmanned vessel application | |
| CN109255317A (en) | A kind of Aerial Images difference detecting method based on dual network | |
| CN104392228A (en) | Unmanned aerial vehicle image target class detection method based on conditional random field model | |
| CN104537689B (en) | Method for tracking target based on local contrast conspicuousness union feature | |
| CN104408482A (en) | Detecting method for high-resolution SAR (Synthetic Aperture Radar) image object | |
| CN110569782A (en) | Target detection method based on deep learning | |
| CN110263712A (en) | A kind of coarse-fine pedestrian detection method based on region candidate | |
| CN104123529A (en) | Human hand detection method and system thereof | |
| CN105930794A (en) | Indoor scene identification method based on cloud computing | |
| CN114926747A (en) | Remote sensing image directional target detection method based on multi-feature aggregation and interaction | |
| CN112200163B (en) | Underwater benthos detection method and system | |
| CN109902576B (en) | Training method and application of head and shoulder image classifier | |
| CN109165603B (en) | Ship detection method and device | |
| CN104268574A (en) | SAR image change detecting method based on genetic kernel fuzzy clustering | |
| CN112613565B (en) | Anti-occlusion tracking method based on multi-feature fusion and adaptive learning rate updating | |
| Zhou et al. | A fusion algorithm of object detection and tracking for unmanned surface vehicles | |
| Shi et al. | Obstacle type recognition in visual images via dilated convolutional neural network for unmanned surface vehicles | |
| Zhu et al. | DiamondNet: Ship detection in remote sensing images by extracting and clustering keypoints in a diamond | |
| Jiang et al. | A robust end-to-end deep learning framework for detecting Martian landforms with arbitrary orientations |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181019 |
|
| RJ01 | Rejection of invention patent application after publication |