CN109191489B - Method and system for detecting and tracking aircraft landing marks - Google Patents
Method and system for detecting and tracking aircraft landing marks Download PDFInfo
- Publication number
- CN109191489B CN109191489B CN201810936437.4A CN201810936437A CN109191489B CN 109191489 B CN109191489 B CN 109191489B CN 201810936437 A CN201810936437 A CN 201810936437A CN 109191489 B CN109191489 B CN 109191489B
- Authority
- CN
- China
- Prior art keywords
- image frame
- initial image
- current image
- tracking
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/207—Analysis of motion for motion estimation over a hierarchy of resolutions
-
- 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 invention relates to the field of image processing, and discloses a method and a system for detecting and tracking an aircraft landing mark, which are used for realizing accurate and rapid detection and tracking of the aircraft landing mark and assisting in finishing accurate landing of the aircraft; the method of the invention comprises the following steps: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, determining a target area corresponding to a landing sign in an initial image frame, then determining a minimum circumscribed rectangle of the target area, and determining four-corner point coordinates of the minimum circumscribed rectangle through a library function; taking the next frame of the initial image frame as the current image frame, calculating an affine transformation matrix between the initial image frame and the current image frame, predicting the position of coordinates of four corner points in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame diagram in the current image frame; and taking the current image frame as a new initial image frame, re-determining the current frame, and repeating the steps to complete the tracking of the landing mark.
Description
Technical Field
The invention relates to the field of image processing, in particular to a method and a system for detecting and tracking an aircraft landing sign.
Background
With the rapid development of the computer industry, the target detection and tracking technology based on computer vision is also greatly improved, and at present, the target detection method mainly comprises an angular point feature detection method based on edges, a detection method based on circle features, a detection method based on invariant moment and a target detection method based on deep learning; the target tracking method mainly comprises a template matching-based method, a region-based tracking method, a dynamic contour tracking algorithm and a particle filter tracking algorithm; although the edge-based corner point detection method is high in precision and good in stability, the dependence of corner points on an edge extraction algorithm is large, and broken edge lines have great influence on corner point detection; the fitting of the circular curve in the detection method based on the circular characteristics depends on the accuracy of the figure contour data, and the calculated amount is large; the invariant moment based target detection method can well detect the target, but the invariant moment has high calculation complexity and is only suitable for processing simple two-dimensional images. The tracking method based on model matching can be applied to the situations of complex environment and target attitude change, but the method needs to calculate complex models, and the calculation amount is very large; the area matching algorithm is large in calculation amount, and in the case of change of occlusion and illumination intensity, the method may fail to track; the dynamic contour tracking method has the disadvantages of sensitivity to the initial value of the target, and once the initial target contour parameter is detected incorrectly, the subsequent tracking process is completely ineffective. The particle filter algorithm is complex, the calculation amount is large, and the required storage space is large; the target detection method based on deep learning has very high accuracy, but higher and higher dimensional feature description brings burden to calculation, and the pre-training of a classifier model requires huge data volume, and one model only corresponds to one target condition.
Disclosure of Invention
The invention aims to provide a method and a system for detecting and tracking an aircraft landing sign, so as to realize accurate and rapid detection and tracking of the aircraft landing sign and assist in finishing accurate landing of the aircraft.
In order to achieve the above object, the present invention provides a method for detecting and tracking an aircraft landing sign, which comprises the following steps:
step S1: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, determining a target area corresponding to a landing sign in an initial image frame, then determining a minimum circumscribed rectangle of the target area, and determining coordinates of four corner points of the minimum circumscribed rectangle through a library function;
step S2: taking the next frame of the initial image frame as a current image frame, calculating an affine transformation matrix between the initial image frame and the current image frame, predicting the positions of coordinates of four corner points in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame in the current image frame;
step S3: and taking the current image frame as a new initial image frame, and repeating the step S2 to complete the tracking of the landing mark.
Preferably, the step S1 specifically includes the following steps:
step S11: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, selecting all contours in an initial image frame, and judging all the contours by adopting a plane figure invariant moment sequence to obtain a target contour;
step S12: selecting a first target contour to perform rotation and stretching treatment to obtain a circular contour, selecting one N of the circular radius of the circular contour as a step length, and performing gray sampling on the circular contour by N concentric circles;
step S13: comparing the sampling result of each concentric circle with a corresponding set calibration value, determining whether the characters in the disc are consistent with a set landing mark, if so, locking a target area in the initial image frame, and if not, returning to the step S12, selecting a next target contour for judging again until the target area in the initial image frame is successfully locked;
step S14: and determining the minimum circumscribed rectangle of the target area, and determining the coordinates of four corner points of the minimum circumscribed rectangle through a library function.
Preferably, the step S2 specifically includes the following steps:
step S21: extracting feature points of an initial image frame by using an SURF feature point detector to serve as a first feature point set;
step S22: taking the next frame of the initial image frame as a current image frame, and predicting the position of the first feature point set in the current image frame by adopting a sparse optical flow to be used as a second feature point set;
step S23: removing feature points in the second feature point set which are failed to track compared with the first feature point set to obtain a third feature point set, and calculating an affine transformation matrix between the initial image frame and the current image frame according to the first feature point set and the third feature point set;
step S24: and calculating the position of four corner point coordinates of a target area in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame diagram in the current image frame.
In order to achieve the above object, the present invention further provides a system for detecting and tracking an aircraft landing indicator, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.
The invention has the following beneficial effects:
the invention provides a method and a system for detecting and tracking an aircraft landing sign, which comprise the following steps: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, determining a target area corresponding to a landing sign in an initial image frame, then determining a minimum circumscribed rectangle of the target area, and determining four-corner point coordinates of the minimum circumscribed rectangle through a library function; then, taking the next frame of the initial image frame as a current image frame, calculating an affine transformation matrix between the initial image frame and the current image frame, predicting the position of coordinates of four corner points in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular block diagram in the current image frame; taking the current image frame as a new initial image frame, repeating the steps, re-determining the current frame, and completing the tracking of the landing mark by iteration; the invention adopts the characteristic of plane graph invariant moment to enable the circular deformation in the angular view range to be more accurate, and simultaneously adopts a Hough circle transformation detection method to increase the detectability and accuracy of a deformation disc area, further improve the stability and robustness of the system, quickly realize the accurate and quick detection and tracking of the aircraft landing mark and assist in finishing the accurate landing of the aircraft.
The present invention will be described in further detail below with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of a detection tracking method in accordance with a preferred embodiment of the present invention;
fig. 2 is a flow chart of an aircraft landing sign detection and tracking method with a character H according to a preferred embodiment of the invention.
Detailed Description
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
Example 1
Referring to fig. 1, the present embodiment provides a method for detecting and tracking an aircraft landing indicator, including the following steps:
step S1: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, determining a target area corresponding to a landing sign in an initial image frame, then determining a minimum circumscribed rectangle of the target area, and determining four-corner point coordinates of the minimum circumscribed rectangle through a library function;
step S2: taking the next frame of the initial image frame as the current image frame, calculating an affine transformation matrix between the initial image frame and the current image frame, predicting the position of coordinates of four corner points in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame diagram in the current image frame;
step S3: the current image frame is used as a new initial image frame, and the step S2 is repeated to complete the tracking of the landing flag.
Specifically, the following description will be given by taking the example of detecting and tracking a landing sign of an aircraft with a character H. As shown in fig. 2.
The specific steps are as follows, step S11: the method comprises the steps of obtaining more than two image frame information collected by a video detection device on the aircraft, selecting all contours in an initial image frame, and judging all the contours by adopting a plane figure invariant moment sequence to obtain a target contour.
Specifically, preprocessing such as graying is performed on the image frames acquired by the video detection device to obtain a binary image with obvious area division. And (3) carrying out contour search in the binary image, sequentially calculating the center distances and affine invariant moments of all searched contours, comparing the affine invariant moments of the contours with those of standard circular contours, and if the affine invariant moments of a certain contour are close to the theoretical values of the affine invariant moments of the circles, regarding the contour as an elliptical contour (target contour), and finally obtaining an elliptical contour sequence. It should be noted that, after the elliptical contour sequence is obtained, the number of the pixel points of the elliptical contour is determined, and the elliptical contours with the number of the pixel points being less than 100 are deleted, so that some interference points which are particularly small and mistaken for the elliptical contours in the binary image are removed. It should be noted that the present invention is not limited to this, and the limited range of the number of the pixels can be adjusted correspondingly under the condition of achieving the same purpose.
Step S12: selecting a first target contour to perform rotation and stretching treatment to obtain a circular contour, selecting one N times of the circular radius of the circular contour as a step length, and performing gray sampling on the circular contour by N concentric circles. Preferably, in the practical operation process, in this embodiment, one fifth of the circle radius of the circular profile is selected as the step size to achieve the best sampling effect, and the gray sampling of five concentric circles is performed on the circular profile. It should be noted that the concentric circle binary image sampling method can simplify the difficulty of detecting the characters without increasing the geometric features, can quickly detect the characters, and improves the robustness and accuracy of the system.
Step S13: and comparing the sampling result of each concentric circle with a corresponding set calibration value to determine whether the characters in the disc are consistent with a set landing mark, if so, locking the target area in the initial image frame, and if not, returning to the step S12, and selecting the next target contour for judging again until the target area in the initial image frame is successfully locked.
Preferably, concentric circle sampling is performed on the landing sign in advance, a sampled gray value (calibration value) of each concentric circle is obtained, and if the gray value of the sampling result of at least three concentric circles meets the gray value of the set landing sign, the character in the circular outline is considered as the set character H. Otherwise, the character in the circular outline is considered not to be matched with the set character, and the next elliptical outline in the elliptical outline sequence is continuously selected to return to the step S12 for judgment.
Step S14: and determining the minimum circumscribed rectangle of the target area, and determining the coordinates of four corner points of the minimum circumscribed rectangle through a library function. The minimum circumscribed rectangle can realize the visualization of target tracking.
The above steps complete the detection of the target landing sign, and further realize the tracking of the target landing sign, and the specific steps are as follows.
Step S21: and extracting the characteristic points of the initial image frame as a first characteristic point set by adopting an SURF characteristic point detector. The SURF feature point detector can achieve fast tracking of the target.
Step S22: and taking the next frame of the initial image frame as a current image frame, and predicting the position of the first feature point set in the current image frame by adopting sparse optical flow to be used as a second feature point set.
Specifically, whether the number of feature points in the current image frame is greater than 20 is judged, if so, the next step is executed, otherwise, the target area is re-detected to realize more accurate tracking. It should be noted that the number of feature points is not limited, and the number of feature points can be adjusted within a certain range in the case of achieving the same purpose.
Step S23: and removing the feature points in the second feature point set which are failed to track compared with the first feature point set to obtain a third feature point set, and calculating an affine transformation matrix between the initial image frame and the current image frame according to the first feature point set and the third feature point set. It is to be noted that, in calculating the third feature point set, the deleted feature points that have failed tracking include feature points that have not been displaced in the second feature point set as compared with the first feature point set, and points that have failed sparse optical flow tracking. The characteristic points are removed, so that a more accurate affine transformation matrix can be further calculated, and faster and more accurate tracking can be realized.
Step S24: and calculating the position of four corner point coordinates of the target area in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame in the current image frame. Further, the current image frame is used as a new initial image frame, and the above step S2 is repeated to complete the tracking of the landing flag.
Example 2
In accordance with the above method embodiments, the present embodiment provides a system for detecting and tracking an aircraft landing indicator, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.
As described above, the invention provides a method and a system for detecting and tracking an aircraft landing sign, which enable circular deformation in an angular view range to be more accurate by adopting the characteristic of invariant moment of a plane graph, and meanwhile, increase the detectability and accuracy of a deformed disc area by adopting a Hough circle transformation detection method, further improve the stability and robustness of the system, can quickly realize accurate and rapid detection and tracking of the aircraft landing sign, and assist in finishing accurate landing of the aircraft.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (3)
1. A method for detecting and tracking an aircraft landing sign is characterized by comprising the following steps:
step S1: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, determining a target area corresponding to a landing sign in an initial image frame, then determining a minimum circumscribed rectangle of the target area, and determining four-corner point coordinates of the minimum circumscribed rectangle through a library function;
step S2: taking the next frame of the initial image frame as a current image frame, calculating an affine transformation matrix between the initial image frame and the current image frame, predicting the positions of coordinates of four corner points in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame in the current image frame;
step S3: taking the current image frame as a new initial image frame, and repeating the step S2 to complete the tracking of the landing mark;
the step S1 specifically includes the following steps:
step S11: acquiring more than two pieces of image frame information acquired by a video detection device on an aircraft, selecting all contours in an initial image frame, and judging all the contours by adopting a plane figure invariant moment sequence to obtain a target contour;
step S12: selecting a first target contour to perform rotation and stretching treatment to obtain a circular contour, selecting one N of the circular radius of the circular contour as a step length, and performing gray sampling on the circular contour by N concentric circles;
step S13: comparing the sampling result of each concentric circle with a corresponding set calibration value, determining whether the characters in the disc are consistent with a set landing mark, if so, locking the target area in the initial image frame, and if not, returning to the step S12, selecting the next target contour for judging again until the target area in the initial image frame is successfully locked;
step S14: and determining the minimum circumscribed rectangle of the target area, and determining the coordinates of four corner points of the minimum circumscribed rectangle through a library function.
2. The method for detecting and tracking an aircraft landing flag according to claim 1, wherein said step S2 specifically comprises the following steps:
step S21: extracting feature points of an initial image frame by using an SURF feature point detector to serve as a first feature point set;
step S22: taking the next frame of the initial image frame as a current image frame, and predicting the position of the first feature point set in the current image frame by adopting a sparse optical flow to be used as a second feature point set;
step S23: removing the feature points which are failed to track in the second feature point set compared with the first feature point set to obtain a third feature point set, and calculating an affine transformation matrix between the initial image frame and the current image frame according to the first feature point set and the third feature point set;
step S24: and calculating the position of four corner point coordinates of a target area in the initial image frame in the current image frame according to the affine transformation matrix, and marking a predicted rectangular frame diagram in the current image frame.
3. A system for detecting and tracking aircraft landing markers, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 2 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810936437.4A CN109191489B (en) | 2018-08-16 | 2018-08-16 | Method and system for detecting and tracking aircraft landing marks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810936437.4A CN109191489B (en) | 2018-08-16 | 2018-08-16 | Method and system for detecting and tracking aircraft landing marks |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109191489A CN109191489A (en) | 2019-01-11 |
| CN109191489B true CN109191489B (en) | 2022-05-20 |
Family
ID=64918457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810936437.4A Active CN109191489B (en) | 2018-08-16 | 2018-08-16 | Method and system for detecting and tracking aircraft landing marks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109191489B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110111364B (en) * | 2019-04-30 | 2022-12-27 | 腾讯科技（深圳）有限公司 | Motion detection method and device, electronic equipment and storage medium |
| CN110969601B (en) * | 2019-11-25 | 2023-11-24 | 同济大学 | Structure rotation response non-contact identification method based on visual characteristic tracking algorithm |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110135206A1 (en) * | 2008-06-20 | 2011-06-09 | National University Corporation Shizuoka Universit | Motion Extraction Device and Program, Image Correction Device and Program, and Recording Medium |
| CN103455797B (en) * | 2013-09-07 | 2017-01-11 | 西安电子科技大学 | Detection and tracking method of moving small target in aerial shot video |
| CN107423702B (en) * | 2017-07-20 | 2020-06-23 | 西安电子科技大学 | Video target tracking method based on TLD tracking system |
| CN107563396B (en) * | 2017-08-10 | 2019-08-30 | 南京大学 | The construction method of protection screen intelligent identifying system in a kind of electric inspection process |
| CN107798691B (en) * | 2017-08-30 | 2019-02-19 | 西北工业大学 | A kind of unmanned plane independent landing terrestrial reference real-time detection tracking of view-based access control model |
-
2018
- 2018-08-16 CN CN201810936437.4A patent/CN109191489B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109191489A (en) | 2019-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hughes et al. | A deep learning framework for matching of SAR and optical imagery | |
| Wang et al. | Lidar iris for loop-closure detection | |
| CN106530347B (en) | Stable high-performance circle feature detection method | |
| Dibeklioglu et al. | 3D facial landmarking under expression, pose, and occlusion variations | |
| CN110334762B (en) | Feature matching method based on quad tree combined with ORB and SIFT | |
| Liu et al. | Stochastic lane shape estimation using local image descriptors | |
| CN110288612B (en) | Nameplate positioning and correcting method and device | |
| CN108447094B (en) | Method and system for estimating attitude of monocular color camera | |
| Tang et al. | Robust pattern decoding in shape-coded structured light | |
| CN111046856A (en) | Parallel pose tracking and map creating method based on dynamic and static feature extraction | |
| CN108550165A (en) | A kind of image matching method based on local invariant feature | |
| CN111126393A (en) | Vehicle appearance refitting judgment method and device, computer equipment and storage medium | |
| CN109191489B (en) | Method and system for detecting and tracking aircraft landing marks | |
| CN110288040B (en) | Image similarity judging method and device based on topology verification | |
| CN111222507A (en) | Automatic identification method of digital meter reading and computer readable storage medium | |
| Lopez-Rincon et al. | Binary large object-based approach for QR code detection in uncontrolled environments | |
| CN113989604A (en) | Tire DOT information identification method based on end-to-end deep learning | |
| Malinas et al. | Vehicle track detection in CCD imagery via conditional random field | |
| CN115760898A (en) | World coordinate positioning method for road sprinklers in mixed Gaussian domain | |
| Zhang et al. | A YOLOv3-Based Industrial Instrument Classification and Reading Recognition Method | |
| CN116309817A (en) | Tray detection and positioning method based on RGB-D camera | |
| CN114332814A (en) | Parking frame identification method and device, electronic equipment and storage medium | |
| EP3469517A1 (en) | Curvature-based face detector | |
| CN112184723A (en) | Image processing method and device, electronic device and storage medium | |
| Sun et al. | Small sample parts recognition and localization from unfocused images in precision assembly systems using relative entropy |
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 |