CN104537663B - A kind of method for quickly correcting of flating - Google Patents
A kind of method for quickly correcting of flating Download PDFInfo
- Publication number
- CN104537663B CN104537663B CN201410824200.9A CN201410824200A CN104537663B CN 104537663 B CN104537663 B CN 104537663B CN 201410824200 A CN201410824200 A CN 201410824200A CN 104537663 B CN104537663 B CN 104537663B
- Authority
- CN
- China
- Prior art keywords
- line
- image
- matching
- flating
- characteristic curve
- 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.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Abstract
The present invention disclose a kind of method for quickly correcting of flating, for the quick correction of flating in video hydrological observation, the characteristic curve of extraction reference picture first, determines the reference line of reference picture;The extraction of characteristic curve is carried out to realtime graphic again;Then the characteristic curve of realtime graphic is matched with reference line, calculates correction parameter;The correction of flating is finally carried out using correction parameter.The present invention is 360 degree of angles for being divided into several fixations camera rotation, for each angle, there is a corresponding masterplate line-segment sets.In actual match, the angle of camera is prejudged first, it is determined that corresponding application scenarios, so that it is determined that the subsequent match masterplate line-segment sets to be used.So each rotationally-varying corresponding application scenarios of camera can be converted into fixed camera scene by an anticipation first, so as to reduce the time of feature lines matching operation, quick obtaining correction parameter, realize quick image dithering correcting.
Description
Technical field
The present invention relates to field of video image processing, more particularly, to a kind of figure being applied in video hydrological observation
As the method for quickly correcting of shake.
Background technology
Because traditional hydrologic monitoring means system installs complexity, installation difficulty is big, and later maintenance cost is big, it is necessary to regularly
The shortcomings of manual maintenance, carry out real-time hydrological observation using video monitoring and have become a kind of trend.Video hydrologic monitoring uses
Be static image analysis technical limit spacing hydrographic information, i.e., corresponding hydrological environment information is extracted by Computer Vision.Phase
It is easy to install than in the hydrologic monitoring equipment of traditional contact, it is hardly damaged, and follow-up maintenance cost is low.
And hydrological observation is carried out using Computer Vision, easily by various wild environment factors(Such as wind, rain)Influence.Regarding
Before frequency image analysis hydrographic information, necessary image dithering correcting --- Image Feature Matching must be carried out.Traditionally use
It is that the matching process based on region is corrected, is easily influenceed by the change of gray scale light and shade, it is impossible to adapts to the change of environment occasion very well
In practical application, Changes in weather is run into(Rain, mist etc.)Occasion the match is successful rate reduces, and under dusk and night-time scene, base
It can not be used completely in Character Area Matching method.Matching and correlation is carried out compared to Character Area Matching using characteristic curve, be not easy by
Half-tone information change interference, line characteristic information has higher robustness than simple half-tone information, because line feature is to figure
As radiation distortion is insensitive.
In general image dithering correcting flow below figure, reads in image first, determines whether still image, then carry out reality
When image feature line extraction and matching, calculate correction parameter, finally utilize correction parameter carry out flating displacement correction.
The important step of two of which is feature line extraction and feature lines matching.And in actual applications, above-mentioned image dithering correcting side
Method existence time complexity is higher, the larger deficiency of amount of calculation, and its operation time reaches several seconds even tens of seconds.According to national water
Position observation standard, video monitoring is higher to algorithm requirement of real time, in order to propose a kind of performance better image method of compensating for hand shake
It is necessary.
Known features line drawing includes detection and the lines detection of image edge information.Edge-Detection Algorithm is domestic and international
Existing many achievements in research, Canny algorithms therein are more ripe one kind.Canny operators have advantages below:(1) believe
Make an uproar than good, i.e., to the false retrieval Loss Rate of marginal point than relatively low;(2) positioning performance is good, that is, the marginal point detected exists as far as possible
The center of actual edge;(3) there is unique response to single edge, i.e., to suppress false edge as far as possible.But Canny algorithms
It is not a kind of quick image processing algorithm.
In lines detection, the noiseproof feature of classical Hough transform is good, and can connect conllinear short straight line, but shortcoming
It is that parameter is difficult to select and calculates complexity, the straight line resolution ratio of extraction is relatively low and lacks local characteristicses.
Based on LocalSearch algorithm characteristics lines matching bearing calibrations, can overcome traditional based on gradation of image information
The shortcomings that with algorithm, in matching by the affine transformation of image, the interference such as image quality change and image deformation is smaller, has stronger
Robustness, can still provide preferable correction in the case of other algorithm cisco unity malfunctions, draw the correct displacement of image
Information.
LocalSearch was put forward by J.Ross Beveridge in 1993, and Local Search are exactly local
The meaning of search.The core of Local Search matching algorithms is depth-first search, and depth optimum search is exactly from present bit
String sets out, and searches for initial bit string of the optimal result as next iterative search in all spectra that its Hamming distances is 1, directly
It is optimal to current bit string, other field combination is not optimal, and this optimal result is exactly local optimum result.In order to ensure
Global optimum is obtained, must repeatedly be searched for, each most initial bit string all randomly generates.Now time complexity is decreased
, but operation time do not reach the speed of project actual requirement still(Below 500ms).Due to Local Search matching algorithms
Run time is long, is greatly differed from each other with the requirement of real-time of actual items, therefore existing LocalSearch matching algorithms are also only
It is to rest on theoretical research stage.
The content of the invention
In order to overcome the above-mentioned deficiencies of the prior art, the present invention proposes a kind of method for quickly correcting of flating, uses
The bearing calibration can improve the robustness of the jitter correction of video hydrological observation and meet the requirement of its real-time monitored.
To achieve these goals, the technical scheme is that:
A kind of method for quickly correcting of flating, for the quick correction of flating in video hydrological observation, including
Following steps:
S1. the characteristic curve of reference picture is extracted, determines the reference line of reference picture;
S2. the extraction of characteristic curve is carried out to realtime graphic;
S3. the characteristic curve of realtime graphic is matched with reference line, calculates correction parameter;
S4. the correction of flating is carried out using correction parameter;
The concrete mode of step S1, S2 is:
S11. all line segments in reference picture are extracted:Set using edge detection algorithm and be adapted to adaptive threshold, extraction ginseng
Examine the edge of image;The vector form of straight line is obtained using Line feature algorithm again, therefrom selected straightway, which is used as, refers to mould
Plate;
S12. a mask is made to each selected straightway, the mask of all selected straightways is added to one
Rise, make the mask plate of image procossing;
S13. based on mask plate to the scene image that gathers in real time, using with step S11 identicals edge detection algorithm, line
Feature detection algorithm carries out the extraction of edge line segment to image, when handling real-time scene image, only in mask plate
Region is handled.
The mask plate that this method makes can reduce the amount of calculation of image procossing in specific image processing process, so as to
The time-consuming of program operation is reduced, basis is provided for fast image processing.When the scene image to gathering in real time is handled, it is only necessary to
Processing calculating is carried out to the information in mask plate region, the amount of calculation of image procossing can be significantly reduced, in mask plate region
Interior calculating processing, equally can accurately extract characteristic curve.
Preferably, it is to make the image masks version of different angle to make image masks version, is specially:By camera rotation
360 degree of angles for being divided into several fixations, for each angle, select out a corresponding masterplate line-segment sets, so as to formulate
Image masks version corresponding to going out.
Several image masks versions are produced according to the anglec of rotation of camera, that is, reduce template line segment aggregate, so as to
Enough spaces for reducing matching, are finally reached the time for reducing the operation of feature lines matching.
Preferably, the step S3 is the matching that image is carried out using LocalSearch characteristic curves matching process, is calculated real
When image and reference picture relative displacement size --- amount of jitter size, unit is pixel;By realtime graphic in step S3
The concrete mode that is matched with reference line of characteristic curve be:
S31. to the anticipation of camera, according to the anglec of rotation of camera, it is determined that the reference template picture for matching, so as to really
The fixed image masks version for matching;
S32. line segments extraction is carried out to collection scene image in real time, using Depth Priority Algorithm in image masks version area
Domain lining scans for, and finds optimum combination, and output matching result, obtains correction parameter.
By the anticipation to camera, each rotationally-varying corresponding application scenarios of camera can be converted into stationary phase
Airport scape, so as to reduce the time of feature lines matching operation.
Preferably, after in extraction, collection scene image carries out line segment in real time, always according to default priori exclusive segment
The line segment pair that can not possibly be matched.
Preferably, the default priori includes the angle between line segment pair, the length ratio and line segment of line segment pair
The distance between to.To the data line-segment sets of input, by some prioris, such as:Angle between two line segments is scope 2 ~ 7
Degree, the length ratio of two line segments be 1/7 ~ 9/2 or two the distance between line segment for 6 ~ 8 pixels exclude data line-segment sets in advance
In can not possibly can reduce package space, with the line segment to match in masterplate line-segment sets so as to reduce amount of calculation.Between line segment pair
Angle, line segment pair length ratio and line segment to the distance between this three parameter, be Practical Project in empirical value, do not determine
Amount analysis and derivation, running into specific environment can adjust.In addition, these parameter setting restrictive conditions are too wide to directly result in matching bar
Part is loose, and result is that matching possibility is more, and match time complexity is bigger;Too strictly the match is successful for possible influence for restrictive condition
Rate.
Compared with prior art, beneficial effects of the present invention are:The method of the present invention is applied in video hydrological observation
Jitter correction, overcome feature based Region Matching easily by half-tone information change etc. factor influenceed the shortcomings that, using mask plate
Method reduce computation complexity, experimental result, show the whole process for calculating picture displacement(Including line drawing and lines matching)
It can control within 1s, meet the requirement of scan picture.
Brief description of the drawings
Fig. 1 is conventional images jitter correction flow chart.
Fig. 2 is the flow chart of existing LocalSearch algorithms.
Fig. 3 is the flow chart of improved LocalSearch algorithms in the present invention.
Fig. 4 is mask design drawings;Wherein d is that mask buffers bandwidth, and AB is with reference to line segment.
Fig. 5 is mask plate administrative division map;White portion is mask, and characteristic curve image procossing is carried out in this region.
Fig. 6 is realtime graphic feature line extraction design sketch.
Fig. 7 is emulation schematic diagram.
Embodiment
The present invention will be further described below in conjunction with the accompanying drawings, but embodiments of the present invention are not limited to this.
Embodiment 1
The present embodiment improves traditional images match bearing calibration, is matched using characteristic curve.By improving feature
Lines matching algorithm, reduce the time needed for calculating(Control is within 1s), it had both been improved the stability of images match correction,
Meet the requirement of real-time of video surveillance applications.
The run time key of characteristic curve matching algorithm is the size of package space;Therefore when reducing the operation of feature lines matching
Between main method be reduce package space size.Package space is by masterplate line-segment sets M and real-time diagram data line-segment sets N
Determine.Before not optimized to package space, the size of package space is, i.e. complete or collected works.Certainly, it is to expire
The requirement of sufficient real-time system.
Reducing characteristic matching space can consider in terms of two:First, the absolute quantity of line segment, another is matching
The relative populations of line segment pair.Absolute quantity refers to only needing building principal outline line segment in image, as long as being built because possessing
Matching can just be completed by building the main profile line segment of thing.The absolute number value of line segment is reduced, can directly reduce M and N absolute size.
Another method for reducing package space size is exactly that the quantity of matching line segment pair is reduced using priori.Although
In actual various projects, when gathering picture, it the various change such as can be related to translating, rotate and scale, but the present embodiment is first only
Consider the situation of translation.Can thus utilize as line segment to the distance between, the length of the angle between line segment pair and line segment pair
The prioris such as short ratio exclude some line segments pair that can not possibly be matched, and thus reduce package space, and then reduce characteristic curve
Time with operation.
In hydrologic monitoring, although the position of camera and focal length are fixed, therefore the scaling of picture is put aside,
In addition to translation, it can also be related to the rotation of camera.Add and rotate this change, package space will be much larger, matching
Time is necessarily elongated.It is 360 degree of angles for being divided into several fixations camera rotation in the method that embodiment uses, for every
One angle, there are a corresponding masterplate line-segment sets.
In actual match, the angle of camera is prejudged first, it is determined that corresponding application scenarios, so that it is determined that subsequent match institute
The masterplate line-segment sets to be used.So first by an anticipation, can be by each rotationally-varying corresponding application of camera
Scene conversion is into fixed camera scene, so as to reduce the time of feature lines matching operation.
The present embodiment to improve the robustness of the jitter correction of video hydrological observation and meeting the requirement of its real-time monitored, its
It is mainly characterized in that:(1)Feature line extraction process is that the line segment of every matching makes mask, greatly reduces the calculating of view data
Amount, reduce and take, reach requirement of real-time;(2)The application scenarios that characteristic curve matching rotation changes are converted into fixed camera
Application scenarios, reduce characteristic matching space, meet the requirement of real-time of Practical Project, realize the quick correction of flating.
The total implementing procedure of set forth below fast jitter bearing calibration:
1st, characteristic curve Detection and Extraction flow:
(1)All line segments in extraction known reference image in advance:Set using the preferable Canny edge detection operators of robustness
Surely it is adapted to adaptive threshold, extracts the edge of image.Then, the vector form of straight line is obtained using Line feature algorithm;From
In select suitable straightway as reference template.
Use Canny edge detection operators realize the process of the extraction of image border line for:Detected using canny algorithms
Band phase in image(Angle)Edge, recycle the Phase Tracking edge at edge(It is assumed that marginal point on the same line its
Phase is roughly the same), obtain one group of group by group of edge points into the point for having linear feature set(Referred to herein as point set), reject
Point set number is less than a certain threshold value(It is typically set to 10-20), finally obtain the vector of straight line using least square fitting
Form.
When selecting suitable straightway as reference template, it is one partially subjective with reference to " standard " that line segment is chosen
Concept.Standard in the present embodiment is:1)Selected line segment effect of trying one's best is good(It is good that effect is bonded with the object of reference in image),
2)The straight line with horizontal both direction vertically is included as far as possible, can form the profile of object of reference as far as possible.3)The reference typically chosen is straight
Number of lines is 4-10.
(2)A mask, the calculation window as scan picture are made for each selected line segment.It is all selected
The mask of line segment is superimposed together, and makes the mask plate of an image procossing, it is therefore an objective to the amount of calculation of image procossing is reduced, from
And the time-consuming of program operation is reduced, provide possibility for fast image processing.
(3)According to mask plate, to the scene image gathered in real time, edge line is still carried out to image using Canny algorithms
The extraction of section, only only the region in mask plate is calculated during image procossing.So image procossing amount of calculation substantially drops
It is low.
Canny edge detection operators are the extractions for characteristic curve;This image dithering correcting method includes characteristics of image
With matching, characteristics of image includes the features such as point, edge, profile, straightway for extraction.
2nd, feature lines matching flow
In flating fast method, LocalSearch algorithms are the matchings for characteristic curve, and images match is then
It is the feature of the same name for finding image, calculates two images(Reference picture and realtime graphic)Shake displacement.It is right below
The principle of LocalSearch algorithms makees a brief description with application:
The flow chart of existing LocalSearch algorithms is as shown in Figure 2:
1)The line segment information collection M corresponding to template image is inputted first(With TXT representation of file)With real time data image
Corresponding line segment information collection N, two data are all the line segments extracted in 1.
2)Then package space is determined according to M and N, because the changes such as rotation, scaling and skew be present in image, thus it is corresponding
Package space be M × N complete or collected works;
3)Followed by Depth Priority Algorithm it is determined that package space in find optimum combination;
4)Output matching result.
This LocalSearch matching algorithms are only merely resting on theoretical research stage, and reason is its long fortune
The row time, greatly differed from each other with the requirement of real-time of actual items.
In order to which LocalSearch matching algorithms are applied in Practical Project, the present embodiment has been made following excellent to it
Change to reduce the run time of algorithm:
(1)By the anticipation to camera, it is determined that the masterplate picture matched will be used for, can for the masterplate picture of determination
To reduce line segment quantity that masterplate picture included as far as possible to reduce the amount of calculation of algorithm;
(2)To the data line-segment sets of input, by some prioris, such as:Angle, the length of two line segments between two line segments
The distance between short ratio and two line segments to exclude in advance data line segment concentrates can not possibly be with the line that matches in masterplate line-segment sets
Section, to reduce package space, so as to reduce amount of calculation.
Above three parameter setting has a certain standard, the scope that several parameters presented below can be set, and 1, angular range
2 ~ 7 degree(Experiment is set as 5 degree);2nd, two line segment length ratio:1/7 ~ 4.5 times;3rd, 6 ~ 8 pixels of line segment distance.Explanation:These
Parameter, it is the empirical value in Practical Project, does not make quantitative analysis with deriving, running into specific environment can adjust.In addition, these are joined
The number setting too wide matching condition that directly results in of restrictive condition is loose, and result is that matching possibility is more, and match time complexity is more
Greatly;Restrictive condition may strictly influence very much the match is successful rate.
By two above measure, the amount of calculation of LocalSearch algorithms can be greatly reduced, so as to can meet to regard
The requirement of real-time of frequency hydrologic monitoring system, LocalSearch matching algorithms are applied among Practical Project for the first time.
The mode of feature lines matching in hydrologic monitoring is as shown in figure 3, key step is as follows:
1) camera is prejudged, determines the angle of camera rotation, the masterplate line that should be used by the angle-determining of camera
Section collection;
2) real-time line-segment sets are inputted, parameter corresponding to priori is set, excludes some line segments pair that can not possibly be matched,
Reduce characteristic matching space;
3) scanned for using Depth Priority Algorithm in package space, find optimum combination, and output matching knot
Fruit.
The method used in the present embodiment is the jitter correction being applied in video hydrological observation, overcomes feature based area
The shortcomings that domain matching is easily influenceed by factors such as half-tone information changes, computation complexity, experiment knot are reduced using the method for mask plate
Fruit, show the whole process for calculating picture displacement(Including line drawing and lines matching)It can control within 1s, meet realtime graphic
The requirement of processing.
As shown in Table 1, under different image-forming conditions, when its feature line extraction success rate is identical, mask extractions are not used
Time used in characteristic curve significantly larger than extracts the time used in characteristic curve using mask, i.e., using method energy disclosed in the present embodiment
The time of characteristic matching is enough significantly reduced, accelerates the correction of flating.
The present embodiment is based primarily upon two algorithms --- feature line extraction and matching algorithm.Two Algorithms T-cbmplexity sheets
Body is very high, and Canny algorithms belong to image processing algorithm, if without Optimal improvements, its time complexity isO(xy),x、yFor
Width, the height of image.Characteristic curve matching algorithm uses LocalSearch algorithms, and time complexity is alsoO(mn),m、nFor with reference to straight
Line and straight line number to be matched.After mask technologies,x、yCan substantially it reduce.In addition, using mask technologies, can be to extraction
To line segment be grouped in advance, i.e., with reference to line segment can only be in corresponding mask the line segment extracted be combined, match
In reference straight line and straight line to be matched between the mapping relations of multi-to-multi be changed into one-to-many, now Algorithms T-cbmplexity isO (n), n is straight line number to be matched.Add priori(For example the anglec of rotation is no more than △ θ, the length ratio of line segment can not be less than
Constant K etc.)Afterwards, some impossible combinations are excluded so that Algorithms T-cbmplexity is less thanO(n).Algorithm in the present embodiment leads to
Various improvement are crossed, it is really achieved fast jitter correcting algorithm.Put into practice in project application, algorithm is in office desktop machine operation
Time control is no more than 1.0s in below 0.5s, maximum.And before algorithm improvement, run time needs more than ten seconds even tens seconds.
The embodiment of invention described above, is not intended to limit the scope of the present invention..It is any in this hair
Made modifications, equivalent substitutions and improvements etc. within bright spiritual principles, it should be included in the claim protection of the present invention
Within the scope of.
Claims (2)
1. a kind of method for quickly correcting of flating, for the quick correction of flating in video hydrological observation, including with
Lower step:
S1. the characteristic curve of reference picture is extracted, determines the reference line of reference picture;
S2. the extraction of characteristic curve is carried out to realtime graphic;
S3. the characteristic curve of realtime graphic is matched with reference line, calculates correction parameter;
S4. the correction of flating is carried out using correction parameter;
Characterized in that, the concrete mode of step S1, S2 is:
S11. all line segments in reference picture are extracted:Set using edge detection algorithm and be adapted to adaptive threshold, extract reference chart
The edge of picture;The vector form of straight line is obtained using Line feature algorithm again, therefrom selectes straightway as reference template;
S12. a mask is made to each selected straightway, the mask of all selected straightways is superimposed together, made
It is used as the mask plate of image procossing;It is the image masks version for making different angle to make pattern mask plate, is specially:Camera is revolved
360 degree turned are divided into the angle of several fixations, for each angle, select out a corresponding masterplate line-segment sets, so as to
Image masks version corresponding to making;
S13. based on mask plate to the scene image that gathers in real time, using with step S11 identicals edge detection algorithm, line feature
Detection algorithm carries out the extraction of edge line segment to image, when handling real-time scene image, only to the region in mask plate
Handled;
The step S3 is the matching that image is carried out using LocalSearch characteristic curves matching process, calculates realtime graphic and ginseng
Examine the size of the relative displacement of image --- amount of jitter size, unit are pixel;In step S3 by the characteristic curve of realtime graphic with
The concrete mode that reference line is matched is:
S31. to the anticipation of camera, according to the anglec of rotation of camera, it is determined that the reference template picture for matching, so that it is determined that with
In the image masks version of matching;
S32. line segments extraction is carried out to collection scene image in real time, using Depth Priority Algorithm in image masks version region
In scan for, find optimum combination, and output matching result, obtain correction parameter;
After extraction in real time collection scene image line segment, the line segment that can not possibly be matched always according to default priori exclusive segment
It is right.
2. the method for quickly correcting of flating according to claim 1, it is characterised in that the default priori
Including the angle between line segment pair, the length ratio of line segment pair and line segment to the distance between.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410824200.9A CN104537663B (en) | 2014-12-26 | 2014-12-26 | A kind of method for quickly correcting of flating |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410824200.9A CN104537663B (en) | 2014-12-26 | 2014-12-26 | A kind of method for quickly correcting of flating |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104537663A CN104537663A (en) | 2015-04-22 |
CN104537663B true CN104537663B (en) | 2018-01-02 |
Family
ID=52853182
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410824200.9A Expired - Fee Related CN104537663B (en) | 2014-12-26 | 2014-12-26 | A kind of method for quickly correcting of flating |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104537663B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108366243B (en) * | 2018-01-23 | 2019-10-29 | 微幻科技(北京)有限公司 | A kind of video jitter removing method and device |
CN109493391A (en) * | 2018-11-30 | 2019-03-19 | Oppo广东移动通信有限公司 | Camera calibration method and device, electronic equipment, computer readable storage medium |
DE102019201702A1 (en) * | 2019-02-11 | 2020-08-13 | Conti Temic Microelectronic Gmbh | Modular inpainting process |
CN111340888B (en) * | 2019-12-23 | 2020-10-23 | 首都师范大学 | Light field camera calibration method and system without white image |
CN113382177B (en) * | 2021-05-31 | 2023-03-28 | 上海东方传媒技术有限公司 | Multi-view-angle surrounding shooting method and system |
CN115760595B (en) * | 2022-10-26 | 2023-07-21 | 中国电子科技集团公司第五十四研究所 | Super-wide-angle lens photo distortion correction method based on line segment characteristics |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103177441A (en) * | 2013-01-24 | 2013-06-26 | 中国科学院对地观测与数字地球科学中心 | Image geometric correction method based on straight line segments |
CN103345642A (en) * | 2013-06-28 | 2013-10-09 | 华中科技大学 | Image matching method based on dotted line dual |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7815275B2 (en) * | 2007-07-27 | 2010-10-19 | Shilin Guo | Interactive visual card-selection process for mitigating light-area banding in a pagewide array |
-
2014
- 2014-12-26 CN CN201410824200.9A patent/CN104537663B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103177441A (en) * | 2013-01-24 | 2013-06-26 | 中国科学院对地观测与数字地球科学中心 | Image geometric correction method based on straight line segments |
CN103345642A (en) * | 2013-06-28 | 2013-10-09 | 华中科技大学 | Image matching method based on dotted line dual |
Non-Patent Citations (1)
Title |
---|
基于特征线匹配的城市建筑物识别方法研究;李松霖等;《遥感技术与应用》;20120430;第27卷(第2期);参见摘要、第191页左栏第15行-第195页右栏第20行,图2-5 * |
Also Published As
Publication number | Publication date |
---|---|
CN104537663A (en) | 2015-04-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104537663B (en) | A kind of method for quickly correcting of flating | |
CN104794421B (en) | A kind of positioning of QR codes and recognition methods | |
CN116664557B (en) | Visual detection method for surface defects of fan blade | |
WO2020042866A1 (en) | Line segment detection method and apparatus, device, and computer-readable storage medium | |
CN106157323B (en) | A kind of insulator division and extracting method of dynamic division threshold value and block search combination | |
US9390330B2 (en) | Apparatus and method for extracting correspondences between aerial images | |
CN104732578B (en) | A kind of building texture optimization method based on oblique photograph technology | |
WO2009026857A1 (en) | Video image motion processing method introducing global feature classification and implementation device thereof | |
CN106408025B (en) | Aerial Images insulator classifying identification method based on image procossing | |
CN109325498B (en) | Vein extraction method for improving Canny operator based on window dynamic threshold | |
CN107564017A (en) | A kind of city high score remote sensing image shadow Detection and dividing method | |
CN111738211B (en) | PTZ camera moving object detection and recognition method based on dynamic background compensation and deep learning | |
CN108133216A (en) | The charactron Recognition of Reading method that achievable decimal point based on machine vision is read | |
CN109961399A (en) | Optimal stitching line method for searching based on Image distance transform | |
CN106600580A (en) | Hough transform-based abnormal recognition method and system of power line | |
CN105279745B (en) | The normalized multi-characteristic points of remote sensing image color constrain the regular method of histogram | |
CN107480678A (en) | A kind of chessboard recognition methods and identifying system | |
CN108492263A (en) | A kind of camera lens Lens Distortion Correction method | |
US10097805B2 (en) | Multi-image color refinement with application to disparity estimation | |
CN103400343A (en) | Method for compensating uneven brightness of bottom view image under nighttime infrared | |
TWI383690B (en) | Method for image processing | |
CN108257153A (en) | A kind of method for tracking target based on direction gradient statistical nature | |
CN113095385B (en) | Multimode image matching method based on global and local feature description | |
CN104408432B (en) | Infrared image target detection method based on histogram modification | |
CN106485264B (en) | It is divided based on gradient sequence and the curve of mapping policy describes and matching process |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180102 |
|
CF01 | Termination of patent right due to non-payment of annual fee |