CN110717396B - Target identification method in UUV cross-column type recovery - Google Patents
Target identification method in UUV cross-column type recovery Download PDFInfo
- Publication number
- CN110717396B CN110717396B CN201910855232.8A CN201910855232A CN110717396B CN 110717396 B CN110717396 B CN 110717396B CN 201910855232 A CN201910855232 A CN 201910855232A CN 110717396 B CN110717396 B CN 110717396B
- Authority
- CN
- China
- Prior art keywords
- light source
- target
- uuv
- underwater image
- source array
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention provides a target identification method in UUV cross-post recovery. Acquiring an underwater image through an underwater camera on a UUV; secondly, preprocessing an image through median filtering and threshold segmentation to obtain an underwater image containing an L-shaped light source array; step three, in the UUV cross-column type recovery process, identifying the number of target light sources in the L-shaped light source array by a sub-pixel outline target number identification method; otherwise, adjusting the UUV position and the heading to continue collecting the underwater image; and step four, identifying the target light source of the L-shaped light source array by a target identification method of slope and relative distance in the UUV cross-column type recovery process. The method and the device realize the target identification of the L-shaped light source array in the UUV recovery process, improve the accuracy of the target identification in the UUV recovery process, and further reduce the positioning error of the UUV relative to the L-shaped light source array.
Description
Technical Field
The invention relates to a target identification method, in particular to a target identification method based on a slope and a relative distance and a target number identification method based on a sub-pixel contour.
Background
Due to the limited volume of an Unmanned Underwater Vehicle (UUV), the energy carried by the UUV is limited, so that the UUV needs to be recovered periodically to supplement the energy.
Due to the complexity and the variability of the underwater environment, the underwater image is affected by factors such as noise, impurities and the like, so that the target recognition in the underwater image is more difficult than that of a general image. These effects are detrimental to the identification of the target luminaire by the camera on the UUV.
Disclosure of Invention
The invention aims to provide a target identification method in UUV cross-column type recovery, which can ensure the accuracy of visual target identification in UUV cross-column type recovery and can well identify an L-shaped light source array in an image.
The purpose of the invention is realized as follows:
acquiring an underwater image through an underwater camera on a UUV;
secondly, preprocessing an image through median filtering and threshold segmentation to obtain an underwater image containing an L-shaped light source array;
step three, in the UUV cross-column type recovery process, identifying the number of target light sources in the L-shaped light source array by a sub-pixel outline target number identification method; otherwise, adjusting the UUV position and the heading to continue collecting the underwater image;
and step four, identifying the target light source of the L-shaped light source array by a target identification method of slope and relative distance in the UUV cross-column type recovery process.
The present invention may further comprise:
1. the adopted target light source is an L-shaped light source array, and the L-shaped light source array is arranged on the recovery device and consists of 4 light sources; three light sources are longitudinally recorded as A, B, C; only one light source D is disposed in the lateral direction centering on C.
2. The identifying the target light source of the L-shaped light source array by the target identification method of the slope and the relative distance specifically includes:
(1) judging a target light source ABC according to the slopes of four target light sources in the L-shaped light source array, and calculating the slopes of AB, AC, AD, BC, BD and CD, wherein the slopes areIs represented by k, wherein k AB 、k AC 、k BC Are the same as each other, k AD 、k BD 、k CD The values of the ABC are different, so that the ABC is judged;
(2) knowing that the number of the target light sources is 4, and after judging ABC, the remaining target light source is D;
(3) according to the distances from the ABC three target light sources to the target light source D, the target light source A, B, C is judged, and by calculating AD |, | BD |, and | CD |, and comparing the magnitudes of | AD |, | BD |, and | CD |, wherein | AD | > | BD | > | CD |, the farthest distance from D is A, the second distance is B, and the nearest distance is C.
3. The identification of the number of the target light sources in the L-shaped light source array by the sub-pixel outline target number identification method specifically comprises the following steps:
(1) extraction of sub-pixel edges of underwater images by Canny operator
The Canny operator is specifically processed as follows:
an ideal gaussian function H (x, y) smoothes the underwater image, with the following formula:
G(x,y)＝f(x,y)*H(x,y)
wherein, in the Gaussian function H (x, y), σ 2 Determining the amplitude of Gaussian graduation for the scale parameter; f (x, y) is an underwater image; g (x, y) is an underwater image after Gaussian smoothing;
secondly, the finite difference of the underwater image is calculated by the first-order difference convolution template, the gradient amplitude and the direction of the image are solved by utilizing the finite difference,
the finite difference of the underwater image is:
wherein the content of the first and second substances,
the amplitude and direction of the underwater image are:
the gradient amplitude of the underwater image is inhibited by the non-maximum value;
detecting the edges of the images by using a dual-threshold algorithm and connecting the edges, and when the amplitude of the underwater image is greater than a set low threshold and smaller than a set high threshold, considering the amplitude as an edge point and connecting the edge point with the known edge point of the underwater image;
(2) dividing the outline of the target light source by adopting a straight line and circle dividing mode;
(3) connecting the sub-pixel outlines with similar end points to complete the combination of the target light source outlines;
(4) and fitting a circle with a known contour by using an algebraic distance method, and acquiring relevant parameters of the circle and a sub-pixel contour of a target light source in the underwater image.
4. In the L-shaped light source array, the distance between the three longitudinal light sources A, B, C is 0.3m, the distance between the light source C and the transverse light source D is 0.4m, and the target light sources in the L-shaped light source array are all of the same type.
In order to realize the target identification of the L-shaped light source array in the UUV recovery process, the technical method provided by the invention is a target identification method based on the slope and the relative distance, so that the accuracy of the target identification in the UUV recovery process is improved, and the positioning error of the UUV relative to the L-shaped light source array is further reduced. The invention also solves the problem that the number of the target light sources in the underwater image cannot be accurately identified due to the fact that the target light source areas are connected together because the area of the target light sources is too large or the distance between the target light sources is too small, and provides a target number identification method based on the sub-pixel outline in the UUV recovery process.
Effects of the invention
The invention adopts an L-shaped light source array as a UUV cross-column type recycled target light source. In an actual marine environment, a common target recognition algorithm hardly meets the recognition precision requirement; and the target identification algorithm based on the slope and the relative distance and the target number identification method based on the sub-pixel outline have small calculated amount and are suitable for real-time identification occasions. According to the method, firstly, an L-shaped light source array is used as a target light source, then, an underwater image is preprocessed in a median filtering and threshold segmentation mode, a target identification method based on a slope and a relative distance and a target number identification method based on a sub-pixel outline are designed, so that the accuracy of visual target identification in UUV cross-column type recovery is guaranteed, and the L-shaped light source array in the image can be well identified.
As a result of the object recognition of the present invention, as shown in fig. 5, the pixel coordinates of ABCD in the L-type light source array are (417.70, 221.76), (256.89, 222.60), (127.88, 224.15), (132.92, 415.40); as shown in fig. 6, the pixel coordinates of ABCD in the L-type light source array are (307.46, 32.17), (216.38, 154.13), (137.25, 265.54), (288.55, 384.27), respectively; as shown in fig. 7, the pixel coordinates of ABCD in the L-shaped light source array are (138.30, 154.27), (133.76,301.06), (137.13,430.84), (322.45,441.32), respectively; as shown in fig. 8, the pixel coordinates of ABCD in the L-shaped light source array are (170.36,196.07), (271.82,304.09), (369.07,407.38), and (516.68,267.51), respectively. In the whole target identification process, the target identification method based on the slope and the relative distance can identify the target cursor A, B, C, D and calculate the pixel coordinates of the target light sources no matter the L-shaped light source array appears at any position of the underwater image, thereby verifying the correctness and the rationality of the proposed target identification algorithm.
According to the target number recognition result of the invention, as shown in fig. 10, the areas of the target light source A and the target light source B are overlapped, the sub-pixel outline is from the outline of the primary rough extraction target to the outline of the precise extraction target, the outline of A, B light sources in the overlapped area as shown in fig. 11 is accurately distinguished, and the accurate recognition of the target number in the underwater image is completed. Therefore, the method based on the sub-pixel outline can solve the problem of identification of the number of the target light sources when the target light sources are connected.
Drawings
FIG. 1 is a pictorial view of a camera;
FIG. 2 is a schematic structural diagram of an L-shaped light source array system;
FIG. 3 is a schematic diagram of an L-shaped light source array;
FIG. 4 is a flow chart of a method of target identification based on slope and relative distance;
FIG. 5 is a schematic diagram of target identification when the UUV heading deviation angle is 0 ° in the method of the present invention;
FIG. 6 is a schematic diagram of target identification when the UUV heading deviation angle is 60 degrees;
FIG. 7 is a schematic diagram of target identification when the UUV heading deviation angle is 90 degrees according to the method of the present invention;
FIG. 8 is a schematic diagram of target identification when the UUV heading deviation angle is 120 degrees according to the method of the present invention;
FIG. 9 is a flow chart of target number identification based on sub-pixel contours;
FIG. 10 is a schematic view of the target light sources connected together;
fig. 11 is a schematic diagram of identifying the number of target light sources when the target light sources are connected together according to the method of the present invention.
Detailed Description
The invention is described in more detail below by way of example.
Step one, acquiring an underwater image through an underwater camera on a UUV.
And step two, obtaining an underwater image containing the L-shaped light source array through image preprocessing such as median filtering, threshold segmentation and the like.
Step three, in the UUV cross-column type recovery process, identifying the number of target light sources in the L-shaped light source array by a sub-pixel outline target number identification method; otherwise, the UUV position and the heading are adjusted to continue collecting the underwater image.
And step four, identifying the target light source of the L-shaped light source array by a target identification method of slope and relative distance in the UUV cross-column type recovery process.
The invention also includes:
1. slope and relative distance based object identification
(1) And judging the target light source ABC according to the slopes of the four target light sources in the L-shaped light source array. The slopes of AB, AC, AD, BC, BD, CD are calculated. The slope is represented by k, where k AB 、k AC 、k BC Are almost the same as each other, k AD 、k BD 、k CD Are different from each other and are k AB 、k AC 、k BC The difference is large. Thus, ABC can be determined, but the specific area is not separated A, B, C from the light source.
(2) Knowing that the number of the target light sources is 4, ABC can be judged through the second step, and the remaining target light source is D.
(3) And judging the target light source A, B, C according to the distances from the three target light sources ABC to the target light source D. By calculating | AD |, | BD |, | CD |, and comparing the magnitudes of | AD |, | BD |, | CD |, where | AD | > | BD | > | CD |. The farthest distance D is a, next B, and closest C.
2. And identifying the number of targets based on the sub-pixel outline.
The area of the target light source is too large or the distance between the target light sources is too small, so that the target light source areas are connected together, and the number of the target light sources in the underwater image cannot be accurately identified. The invention adopts a sub-pixel outline method to distinguish the connected target light sources, thereby accurately identifying the number of the target light sources. The sub-pixel profile of the target light source refers to a set of sequential control points in the underwater image, and the sequence of the control points indicates the association between pixels. The sub-pixel outline is formed by key pixel points in the image instead of connected pixel points, so that the target feature can be accurately extracted. The method for identifying the number of target light sources by the sub-pixel outline comprises the following steps:
(1) extraction of sub-pixel edges of underwater images by Canny operator
The Canny operator is an operator for detecting the edge of the underwater image, and is mainly used for eliminating underwater noise and extracting the edge of the underwater image. The Canny operator is specifically processed as follows:
an ideal gaussian function H (x, y) smoothes the underwater image, with the following formula:
G(x,y)＝f(x,y)*H(x,y)
wherein, in the Gaussian function H (x, y), σ 2 Determining the Gaussian graduation amplitude for the scale parameter; f (x, y) is an underwater image; g (x, y) is the underwater image after Gaussian smoothing.
Secondly, the finite difference of the underwater image is calculated by the first-order difference convolution template, and then the gradient amplitude and the direction of the image are obtained by utilizing the finite difference. The finite difference of the underwater image is:
wherein the content of the first and second substances,
the amplitude and direction of the underwater image are:
and thirdly, inhibiting the gradient amplitude of the underwater image by using the non-maximum value.
And detecting and connecting the edges of the images by using a dual-threshold (low threshold and high threshold) algorithm. And when the amplitude of the underwater image is greater than the set low threshold and less than the set high threshold, the amplitude is considered as an edge point. And connecting the edge points with the known edge points of the underwater image.
(2) And dividing the outline of the target light source by adopting a straight line and circle dividing mode.
(3) And connecting the sub-pixel outlines with the similar end points to complete the combination of the target light source outlines.
(4) And fitting a circle with a known contour by using an algebraic distance method, and acquiring relevant parameters of the circle and a sub-pixel contour of a target light source in the underwater image.
The specific implementation mode is as follows: the embodiment is described with reference to fig. 1 and 2, and the target identification method based on the slope and the relative distance in the UUV cross-column type recovery process of the embodiment is specifically prepared according to the following steps:
acquiring an underwater image containing an L-shaped light source array through image preprocessing;
and step two, judging a target light source ABC according to the slopes of the four target light sources in the L-shaped light source array.
The slopes of AB, AC, AD, BC, BD, CD are calculated. The slope is represented by k, where k AB 、k AC 、k BC Are almost the same as each other, k AD 、k BD 、k CD Are different from each other and are k AB 、k AC 、k BC The difference is large. Thus, ABC can be determined, but the specific area is not separated A, B, C from the light source.
Step three, knowing that the number of the target light sources is 4, judging ABC through the step two, and determining the remaining target light source to be D.
And step four, judging the target light source A, B, C according to the distances from the three target light sources ABC to the target light source D. By calculating | AD |, | BD |, | CD |, and comparing the magnitudes of | AD |, | BD |, | CD |, where | AD | > | BD | > | CD |. Then the farthest from D is a, next to B, and closest to C.
With reference to fig. 9, a specific process of the sub-pixel accuracy-based target number identification method in the UUV cross-columnar recovery process of the embodiment is described as follows:
step one, extracting sub-pixel edges of underwater images by Canny operators
The Canny operator is an operator for detecting the edge of the underwater image, and is mainly used for eliminating underwater noise and extracting the edge of the underwater image. The Canny operator is specifically processed as follows:
an ideal gaussian function H (x, y) smoothes the underwater image, with the following formula:
G(x,y)＝f(x,y)*H(x,y)
wherein, in the Gaussian function H (x, y), σ 2 Determining the amplitude of Gaussian graduation for the scale parameter; f (x, y) is an underwater image; g (x, y) is the underwater image after Gaussian smoothing.
Secondly, the finite difference of the underwater image is calculated by the first-order difference convolution template, and then the gradient amplitude and the direction of the image are obtained by utilizing the finite difference. The finite difference of the underwater image is:
wherein the content of the first and second substances,
the amplitude and direction of the underwater image are:
and thirdly, inhibiting the gradient amplitude of the underwater image by using the non-maximum value.
And detecting and connecting the edges of the images by using a dual-threshold (low threshold and high threshold) algorithm. And when the amplitude of the underwater image is greater than the set low threshold and less than the set high threshold, the amplitude is considered as an edge point. And connecting the edge points with the known edge points of the underwater image.
And secondly, segmenting the outline of the target light source by adopting a linear and circular segmentation mode.
And step three, connecting the sub-pixel outlines with similar end points to complete the combination of the target light source outlines.
And step four, fitting a circle with a known contour by using an algebraic distance method, and acquiring related parameters of the circle and the sub-pixel contour of the target light source in the underwater image.
With reference to fig. 2 and 3, the L-shaped light source array needs to be set as the target light source, and the specific establishment process is as follows:
the core of the recovery strategy is that the long side '┃' of the L shape determines four degrees of freedom of the UUV and the short side '━' of the L shape determines the heading of the UUV, and the vertical deviation, the longitudinal deviation, the transverse deviation and the heading deviation angle of the UUV relative to a recovery device can be calculated by utilizing a target light source in the L-shaped light source array, so that the UUV reaches a set pose, wherein the L-shaped light source array is characterized in that:
(1) the L-shaped light source array is arranged on the recovery device and consists of 4 light sources;
(2) three light sources are longitudinally arranged, the distance between every two light sources is 0.3m, and the light sources are respectively recorded as A, B, C;
(3) only one light source is arranged along the transverse direction by 0.4m with C as the center;
(4) the total length of the L-shaped light source array in the longitudinal direction is 0.6m, and the total length of the L-shaped light source array in the transverse direction is 0.4 m. The target light sources in the L-shaped light source array are all of the same type, so that all the light sources with the same target light source shape, light emitting area and the like are guaranteed to be subjected to waterproof treatment. )
The following examples were used to demonstrate the beneficial effects of the present invention:
the method comprises the steps of carrying out target identification of an L-shaped light source array on a group of underwater image sequences with UUV heading deviation angle changes, testing a target light source identification method based on slope and relative distance, wherein the test results are shown in figures 5, 6, 7 and 8, carrying out target light source number identification on underwater images with connected target light sources, carrying out a water pool test on target number identification based on sub-pixel outlines, and the test results are shown in figure 11. The target identification method based on the slope and the relative distance judges ABC through the slope, the remainder is D, and ABC is specifically distinguished through the distance from ABC to D. In the whole target identification process, no matter the L-shaped light source array appears at any position of the underwater image, the target cursor can be identified and the pixel coordinates of the target light sources can be calculated by the target identification method based on the slope and the relative distance, so that the correctness and the rationality of the proposed target identification algorithm are verified. And based on the identification of the number of the targets of the sub-pixel outline, overlapping the areas of the target light source A and the target light source B, accurately distinguishing the outline of the A, B light source in the overlapping area from the outline of the primary rough extracted target to the outline of the precise extracted target by the sub-pixel outline, and completing the accurate identification of the number of the targets in the underwater image. Therefore, the method based on the sub-pixel outline can solve the problem of identification of the number of the target light sources when the target light sources are connected.
The fork column type recovery device is simple in mechanical structure and convenient to install, and space and cost are saved by fork column type recovery, so that the invention takes UUV fork column type recovery as background, adopts a D550 underwater flashlight as a target light source, designs an L-shaped light source array and an underwater camera Tornado of Tritech company to form a monocular vision system, and researches the problem of identifying the target light source in the recovery process. The L-shaped light source array consists of 4 light sources. Three light sources are longitudinally arranged, the distance between every two light sources is 0.3m, and the light sources are respectively recorded as A, B, C; and only one light source is arranged along the transverse direction of 0.4m by taking the C as the center, so that the total length of the L-shaped light source array in the longitudinal direction is 0.6m, and the total length of the transverse direction is 0.4 m.
Claims (3)
1. A target identification method in UUV cross-column recovery is characterized in that:
acquiring an underwater image through an underwater camera on a UUV;
secondly, preprocessing an image through median filtering and threshold segmentation to obtain an underwater image containing an L-shaped light source array;
the adopted target light source is an L-shaped light source array, and the L-shaped light source array is arranged on the recovery device and consists of 4 light sources; three light sources are longitudinally recorded as A, B, C; only one light source D is arranged along the transverse direction by taking the C as the center;
step three, in the UUV cross-column type recovery process, identifying the number of target light sources in the L-shaped light source array by a sub-pixel outline target number identification method; otherwise, adjusting the UUV position and the heading to continue collecting the underwater image;
identifying a target light source of the L-shaped light source array by a target identification method of slope and relative distance in the UUV cross-column type recovery process;
(1) judging a target light source ABC according to the slopes of four target light sources in the L-shaped light source array, calculating the slopes of AB, AC, AD, BC, BD and CD, wherein the slope is represented by k, and k is AB 、k AC 、k BC Are the same as each other, k AD 、k BD 、k CD The values of the ABC are different, so that the ABC is judged;
(2) knowing that the number of the target light sources is 4, and after judging ABC, the remaining target light source is D;
(3) according to the distances from the ABC three target light sources to the target light source D, the target light source A, B, C is judged, and by calculating | AD |, | BD |, and | CD |, and comparing the magnitudes of | AD |, | BD |, and | CD |, wherein | AD | > | BD | > | CD |, the farthest distance from D is A, the second distance is B, and the nearest distance is C.
2. The method of claim 1 for identifying an object in UUV forked recovery, wherein the method comprises the following steps: the identification of the number of the target light sources in the L-shaped light source array by the sub-pixel outline target number identification method specifically comprises the following steps:
(1) extraction of sub-pixel edges of underwater images by Canny operator
The Canny operator is specifically processed as follows:
an ideal gaussian function H (x, y) smoothes the underwater image, with the following formula:
G(x,y)＝f(x,y)*H(x,y)
wherein, in the Gaussian function H (x, y), σ 2 Determining the amplitude of Gaussian graduation for the scale parameter; f (x, y) is an underwater image; g (x, y) is an underwater image after Gaussian smoothing;
secondly, the finite difference of the underwater image is calculated by the first-order difference convolution template, the gradient amplitude and the direction of the image are solved by utilizing the finite difference,
the finite difference of the underwater image is:
wherein the content of the first and second substances,
the amplitude and direction of the underwater image are:
the gradient amplitude of the underwater image is inhibited by the non-maximum value;
detecting the edges of the images by using a dual-threshold algorithm and connecting the edges, and when the amplitude of the underwater image is greater than a set low threshold and smaller than a set high threshold, considering the amplitude as an edge point and connecting the edge point with the known edge point of the underwater image;
(2) dividing the outline of the target light source by adopting a straight line and circle dividing mode;
(3) connecting the sub-pixel outlines with similar end points to complete the combination of the target light source outlines;
(4) and fitting a circle with a known contour by using an algebraic distance method, and acquiring relevant parameters of the circle and a sub-pixel contour of a target light source in the underwater image.
3. The method of claim 2 for target identification in UUV cross-columnar recovery, wherein: in the L-shaped light source array, the distance between the three longitudinal light sources A, B, C is 0.3m, the distance between the light source C and the transverse light source D is 0.4m, and the target light sources in the L-shaped light source array are all of the same type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910855232.8A CN110717396B (en) | 2019-09-11 | 2019-09-11 | Target identification method in UUV cross-column type recovery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910855232.8A CN110717396B (en) | 2019-09-11 | 2019-09-11 | Target identification method in UUV cross-column type recovery |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110717396A CN110717396A (en) | 2020-01-21 |
| CN110717396B true CN110717396B (en) | 2022-08-02 |
Family
ID=69209804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910855232.8A Active CN110717396B (en) | 2019-09-11 | 2019-09-11 | Target identification method in UUV cross-column type recovery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110717396B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101900811A (en) * | 2010-07-23 | 2010-12-01 | 哈尔滨工程大学 | Passive distance measurement method based on single hydrophone |
| CN104298996A (en) * | 2014-08-08 | 2015-01-21 | 中国科学院自动化研究所 | Underwater active vision tracking method applied to bionic robot fish |
| CN109949361A (en) * | 2018-12-16 | 2019-06-28 | 内蒙古工业大学 | A kind of rotor wing unmanned aerial vehicle Attitude estimation method based on monocular vision positioning |
| CN109961012A (en) * | 2019-02-20 | 2019-07-02 | 博雅工道（北京）机器人科技有限公司 | A kind of underwater target tracking recognition methods |
| CN110186655A (en) * | 2019-06-25 | 2019-08-30 | 西安工业大学 | Imaging detection distance test system based on simulated target and optical energy attenuation device |
-
2019
- 2019-09-11 CN CN201910855232.8A patent/CN110717396B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101900811A (en) * | 2010-07-23 | 2010-12-01 | 哈尔滨工程大学 | Passive distance measurement method based on single hydrophone |
| CN104298996A (en) * | 2014-08-08 | 2015-01-21 | 中国科学院自动化研究所 | Underwater active vision tracking method applied to bionic robot fish |
| CN109949361A (en) * | 2018-12-16 | 2019-06-28 | 内蒙古工业大学 | A kind of rotor wing unmanned aerial vehicle Attitude estimation method based on monocular vision positioning |
| CN109961012A (en) * | 2019-02-20 | 2019-07-02 | 博雅工道（北京）机器人科技有限公司 | A kind of underwater target tracking recognition methods |
| CN110186655A (en) * | 2019-06-25 | 2019-08-30 | 西安工业大学 | Imaging detection distance test system based on simulated target and optical energy attenuation device |
Non-Patent Citations (5)
| Title |
|---|
| AUV Docking System for Sustainable Science Missions;Tarun Podder等;《Proceedings ofthe 2004 IEEE internatlonai Conference on Robotics &Automation New Orleans》;20040430;第4478-4485页 * |
| 一种实用的图像滤波算法;杨小冈等;《计算机应用》;20090630;第216-230页 * |
| 基于前视声纳的成像与多目标特征提取;吴丽媛等;《计算机工程与应用》;20120521(第02期);第222-225页 * |
| 水下目标亮点拓扑特征提取及自动识别方法;朱兆彤等;《声学学报》;20180315(第02期);第28-36页 * |
| 水下高速运动目标的瞬态被动探测方法研究;容慧等;《自动化应用》;20180725(第07期);第50-53页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110717396A (en) | 2020-01-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107462223B (en) | Automatic measuring device and method for sight distance of vehicle before turning on highway | |
| CN105865344A (en) | Workpiece dimension measuring method and device based on machine vision | |
| CN108597009B (en) | Method for detecting three-dimensional target based on direction angle information | |
| CN104751151B (en) | A kind of identification of multilane in real time and tracking | |
| WO2020253010A1 (en) | Method and apparatus for positioning parking entrance in parking positioning, and vehicle-mounted terminal | |
| CN111598952B (en) | Multi-scale cooperative target design and online detection identification method and system | |
| Li et al. | Nighttime lane markings recognition based on Canny detection and Hough transform | |
| CN101509782B (en) | Small-sized ground marker capturing and positioning method | |
| CN110569861B (en) | Image matching positioning method based on point feature and contour feature fusion | |
| CN105373135A (en) | Method and system for guiding airplane docking and identifying airplane type based on machine vision | |
| CN104535066A (en) | Marine target and electronic chart superposition method and system in on-board infrared video image | |
| CN111311679A (en) | Free floating target pose estimation method based on depth camera | |
| CN110675449B (en) | Binocular camera-based offshore flow detection method | |
| CN110378957B (en) | Torpedo tank car visual identification and positioning method and system for metallurgical operation | |
| CN106815553B (en) | A kind of infrared forward sight image Ship Detection based on edge matching | |
| CN103954280A (en) | Rapid, high-robustness and autonomous fixed star identification method | |
| CN113624225B (en) | Pose resolving method for mounting engine positioning pins | |
| CN110675341A (en) | Monocular-vision-guided underwater robot and seabed platform butt joint method | |
| CN111415376A (en) | Automobile glass sub-pixel contour extraction method and automobile glass detection method | |
| CN110472451B (en) | Monocular camera-based artificial landmark oriented to AGV positioning and calculating method | |
| CN114463425A (en) | Workpiece surface featureless point positioning method based on probability Hough linear detection | |
| CN117710458A (en) | Binocular vision-based carrier aircraft landing process relative position measurement method and system | |
| CN116883446B (en) | Real-time monitoring system for grinding degree of vehicle-mounted camera lens | |
| CN103679713A (en) | Two-dimensional image registration method for partially matched images | |
| CN110717396B (en) | Target identification method in UUV cross-column type recovery |
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 |