CN114842354A - Rapid and accurate detection method for high-resolution remote sensing image edge line - Google Patents
Rapid and accurate detection method for high-resolution remote sensing image edge line Download PDFInfo
- Publication number
- CN114842354A CN114842354A CN202210484485.0A CN202210484485A CN114842354A CN 114842354 A CN114842354 A CN 114842354A CN 202210484485 A CN202210484485 A CN 202210484485A CN 114842354 A CN114842354 A CN 114842354A
- Authority
- CN
- China
- Prior art keywords
- image
- edge line
- pixel
- detection result
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000000605 extraction Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000003706 image smoothing Methods 0.000 claims abstract description 8
- 238000009499 grossing Methods 0.000 claims abstract description 6
- 238000001228 spectrum Methods 0.000 claims abstract description 6
- 238000003708 edge detection Methods 0.000 claims description 6
- 230000001629 suppression Effects 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 235000020061 kirsch Nutrition 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
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
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
Abstract
The invention relates to a method for quickly and accurately detecting a high-resolution remote sensing image edge line, and belongs to the technical field of remote sensing image processing and information extraction. The invention comprises the following steps: s1: image smoothing treatment: smoothing the high-resolution remote sensing image by using a simple linear iterative clustering algorithm, and eliminating the influence of image noise and the phenomenon of same-object different-spectrum on image edge line detection; s2: detecting an initial edge line of an image: detecting the smooth image edge line by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result; s3: optimizing the detection result of the image edge line: and optimizing the initial image edge line detection result by using a skeleton line extraction algorithm to obtain an image edge line detection result which is smooth, continuous and good in single-pixel and image ground object boundary adherence. The method can effectively overcome the influence of image noise and image isomorphism spectrum phenomenon on the image edge line detection, and obtain an accurate image edge line detection result.
Description
Technical Field
The invention relates to a method for quickly and accurately detecting a high-resolution remote sensing image edge line, and belongs to the technical field of remote sensing image processing and information extraction.
Background
With the development of ground-to-ground observation technology, the data acquisition amount of high-resolution remote sensing images is increasing, and the high-resolution remote sensing images are widely used in the aspects of urban planning, disaster monitoring, agricultural management, military reconnaissance and the like. The high-resolution remote sensing image edge line detection is used as a key technology for automatic extraction, analysis and understanding of image information, and plays an important role in the aspects of image target detection, image segmentation, surface feature extraction and the like. At present, the traditional image edge line detection methods such as Sobel, Robert, Prewitt, Kirsch, Gauss-Laplace and Canny are generally adopted to process the image so as to obtain the edge line detection result of the image. However, the traditional image edge line detection method is easily affected by image noise and isomorphism phenomenon in the high-resolution remote sensing image, and the extracted image edge line has the phenomena of fracture, burr, edge line redundancy and the like. Therefore, it is necessary to research a fast and accurate detection method for high-resolution remote sensing image lines, so as to overcome the influence of image noise and isomorphic and multispectral phenomena on image edge line detection and obtain smooth and continuous image edge detection results with good single-pixel and image ground object boundary adherence.
Disclosure of Invention
The invention provides a method for quickly and accurately detecting a high-resolution remote sensing image edge line, aiming at the problem of detecting the image edge line by high-resolution remote sensing image noise and the phenomenon of same object and different spectrum.
The invention relates to a method for quickly and accurately detecting edge lines of high-resolution remote sensing images, which comprises the following steps:
s1: image smoothing treatment: the method comprises the following steps of smoothing the high-resolution remote sensing image by using a simple linear iterative clustering algorithm, eliminating the influence of image noise and the phenomenon of same-object different-spectrum on image edge line detection, and adopting the following formula:
wherein: gray ij The gray value of a pixel with a label of j in the super pixel with the label of i is obtained; n is i The number of pixels included in the super pixel with the label of i;
s2: detecting an initial edge line of the image: detecting the smooth image edge line by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result, wherein the method comprises the following steps:
s21: calculating image pixel gradient: calculating the pixel gradient amplitude and direction after image smoothing, wherein:
calculating the gradient amplitude of the image pixel in the X coordinate direction and the Y coordinate direction, namely:
G x (i,j)＝S(i+1,j)-S(i-1,j) (2)
G y (i,j)＝S(i,j+1)-S(i,j-1) (3)
calculating the gradient amplitude and direction of the image pixel, namely:
wherein: (i, j) is the position of the pixel in the XY coordinate axis; g x (i, j) and G y (i, j) are the gradient amplitudes of the pixel in the X-axis and Y-axis directions at the (i, j) position, respectively; g (i, j) and θ (i, j) are the gradient magnitude and direction of the pixel at the (i, j) position, respectively;
s22: obtaining candidate edge line points: processing the gradient image of the image by using a non-maximum suppression algorithm to obtain a local maximum on the edge point, wherein the local maximum is used as an image candidate edge line point;
s23: tracking an image edge line: based on the candidate edge line points, tracking the edge line points and connecting the edge line points by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result;
s3: optimizing the detection result of the image edge line: and optimizing the initial image edge line detection result by using a skeleton line extraction algorithm to obtain an image edge line detection result which is smooth, continuous and good in single-pixel and image ground object boundary attachment.
Preferably, in step S1, the simple linear iterative clustering algorithm converts the image into a 5-dimensional feature vector in CIELAB color space and XY coordinates, and then constructs a metric standard for the 5-dimensional feature vector, and performs local clustering on the image pixels to generate superpixels;
compared with superpixels generated by an image over-segmentation algorithm of a watershed algorithm, a region growing algorithm and a graph-based image segmentation algorithm, the superpixels generated by the simple linear iterative clustering algorithm have better ground object boundary attachment and a more regular and compact shape, and the number of the superpixels generated by the simple linear iterative clustering algorithm through artificial control and good noise resistance.
Preferably, in step S22, the non-maximum suppression algorithm includes the following steps:
and performing interpolation calculation in a 3 × 3 neighborhood taking the pixel (i, j) as the center, wherein if the gradient amplitude of the pixel (i, j) is larger than the gradient amplitude of two adjacent interpolation values in the gradient direction theta (i, j), the pixel (i, j) is taken as a candidate edge pixel, and otherwise, the pixel (i, j) is taken as a non-edge pixel.
Preferably, in step S23, in the dual-threshold edge line tracking algorithm:
s231: setting a low gradient amplitude threshold T l And a high gradient magnitude threshold T h ；
S232: the gradient amplitude of the pixel in the gradient image of the image is larger than T h The gradient amplitude is less than T l The gradient amplitude is larger than T l And is less than T h The pixels of (2) are taken as suspected edge line points;
s233: for the suspected edge line point, if the pixel adjacent to the suspected edge line point is the edge line point, the suspected edge line point is the edge line point; otherwise, it is a non-edge line point.
Preferably, in the step S3, if the image edge line detection result obtained by using the dual-threshold edge line tracking algorithm has a fracture, a burr, and an edge line redundancy, the initial image edge line detection result needs to be optimized, and the skeleton line extraction algorithm is used to optimize the initial image edge line detection result, so as to obtain a smooth and continuous image edge detection result with good single-pixel and image ground object boundary adhesion.
Preferably, in step S3, the skeleton line extraction algorithm includes two steps, that is, firstly, performing expansion processing on edge line points of the initial edge line detection result, specifically, taking all pixels in the eight neighborhood range of the edge line points as the edge line points; and then thinning the edge line points after the expansion processing, thereby obtaining an image edge detection result which is smooth, continuous and good in single-pixel and image ground object boundary adhesion.
The invention has the beneficial effects that: the method for quickly and accurately detecting the edge line of the high-resolution remote sensing image can effectively overcome the influence of image noise and image homomorphism and heteromorphism phenomena on the detection of the edge line of the image and obtain an accurate image edge line detection result.
Drawings
FIG. 1 is a flow diagram of the present invention.
Fig. 2(a) is a high-resolution remote sensing image of the present invention.
FIG. 2(b) is a diagram showing the result of super-pixel segmentation according to the present invention.
FIG. 2(c) is a graph showing the image smoothing process result of the present invention.
FIG. 3 is a diagram illustrating an initial detection result of an image edge line according to the present invention.
FIG. 4 is a diagram illustrating the result of the optimization process of the initial image edge line detection according to the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The embodiment is as follows:
the technical scheme of the invention can adopt a computer software mode to support the automatic operation process. The technical scheme of the invention is explained in detail by combining the drawings and the embodiment.
Fig. 1 shows a flow chart of a technique for rapidly and accurately detecting an edge line of a high-resolution remote sensing image, and the method of the present invention is described in further detail below with respect to each step in the flow of an embodiment.
(1) Image smoothing treatment: and carrying out smoothing treatment on the high-resolution remote sensing image by using a simple linear iterative clustering algorithm, and eliminating the influence of image noise and the phenomenon of same-object different spectrum on image edge line detection.
In the invention, a simple linear iterative clustering algorithm is used for segmenting the image to generate superpixels, the image is subjected to smoothing treatment (formula (1)) according to the generated superpixels, and the flow of the smoothing treatment on the image is shown in FIGS. 2(a) -2 (c). The simple linear iterative clustering algorithm generally converts an image into a 5-dimensional feature vector under a CIELAB color space and XY coordinates, then constructs a measurement standard for the 5-dimensional feature vector, and performs local clustering on image pixels to generate superpixels. Compared with superpixels generated by traditional image over-segmentation algorithms such as a watershed algorithm, a region growing algorithm, a graph-based image segmentation algorithm and the like, the superpixels generated by the simple linear iterative clustering algorithm have better ground object boundary attachment and a more regular and compact shape, and the number of the generated superpixels can be artificially controlled by the simple linear iterative clustering algorithm and have good noise immunity.
Wherein: gray ij The gray value of a pixel with a label of j in the super pixel with the label of i is obtained; n is i The number of pixels included in the super pixel labeled i.
(2) Detecting an initial edge line of an image: and detecting the smooth image edge line by using a double-threshold edge line tracking algorithm to obtain an initial image edge line detection result.
In the invention, a double-threshold edge line tracking algorithm is used to obtain the initial edge line detection result of the image. The dual-threshold edge line tracking algorithm is implemented as follows:
1) calculating image pixel gradients
The pixel gradient magnitude and direction after image smoothing are calculated using the following formulas. And (3) respectively calculating the gradient amplitudes of the image pixels in the X coordinate direction and the Y coordinate direction by using the formula (2) and the formula (3). And (4) respectively calculating the gradient amplitude and the gradient direction of the image pixel by using a formula (4) and a formula (5).
G x (i,j)＝S(i+1,j)-S(i-1,j) (2)
G y (i,j)＝S(i,j+1)-S(i,j-1) (3)
Wherein: (i, j) is the position of the pixel in the XY coordinate axis; g x (i, j) and G y (i, j) are the gradient amplitudes of the pixel in the X-axis and Y-axis directions at the (i, j) position, respectively; g (i, j) and θ (i, j) are the gradient magnitude and direction of the pixel at the (i, j) location, respectively.
2) Obtaining candidate edge line points
The pixels on the edge points usually have larger amplitude, the gradient image of the image is processed by using a non-maximum suppression algorithm to obtain local maximum values on the edge points, and the local maximum values are used as candidate edge line points of the image. The non-maximum suppression algorithm is implemented by performing interpolation calculation in a 3 × 3 neighborhood centered on the pixel (i, j), and if the gradient magnitude of the pixel (i, j) is greater than the gradient magnitudes of two adjacent interpolations in the gradient direction θ (i, j), the pixel (i, j) is used as a candidate edge pixel, otherwise, the pixel is used as a non-edge pixel.
3) Tracking image edge line
And based on the candidate edge line points, tracking the edge line points and connecting the edge line points by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result. Wherein, a low gradient amplitude threshold value T is set in the dual-threshold edge line tracking algorithm l And a high gradientAmplitude threshold T h . The gradient amplitude of the pixel in the gradient image of the image is larger than T h The gradient amplitude is less than T l The gradient amplitude is larger than T l And is less than T h The pixel of (2) is taken as a pseudo edge line point. For the suspected edge line point, if the pixel adjacent to the suspected edge line point is the edge line point, the suspected edge line point is the edge line point; otherwise, it is a non-edge line point.
(3) Optimizing the detection result of the image edge line: and optimizing the initial image edge line detection result by using a skeleton line extraction algorithm to obtain an image edge line detection result which is smooth, continuous and good in single-pixel and image ground object boundary attachment.
The image edge line detection result obtained by using the dual-threshold edge line tracking algorithm has the phenomena of fracture, burr, edge line redundancy and the like, as shown in fig. 3; the initial edge line detection result of the image needs to be optimized, and the method uses a skeleton line extraction algorithm to optimize the initial edge line detection result of the image to obtain an image edge detection result which is smooth, continuous, and good in single-pixel and image ground object boundary attachment, as shown in fig. 4.
The skeleton line extraction algorithm mainly comprises two steps, firstly, carrying out expansion processing on edge line points of an initial edge line detection result, and specifically, taking pixels in eight neighborhood regions of the edge line points as the edge line points; and then thinning the edge line points after the expansion processing, thereby obtaining an image edge detection result which is smooth, continuous and has good single-pixel and image ground object boundary adherence.
The invention can be widely applied to remote sensing image processing and information extraction occasions.
The specific examples described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made or substituted in a similar manner to the specific embodiments described herein by those skilled in the art without departing from the spirit of the invention or exceeding the scope thereof as defined in the appended claims.
Claims (6)
1. A method for quickly and accurately detecting the edge line of a high-resolution remote sensing image is characterized by comprising the following steps:
s1: image smoothing treatment: the method comprises the following steps of smoothing the high-resolution remote sensing image by using a simple linear iterative clustering algorithm, eliminating the influence of image noise and the phenomenon of same-object different-spectrum on image edge line detection, and adopting the following formula:
wherein: gray ij The gray value of a pixel with a label of j in the super pixel with the label of i is obtained; n is i The number of pixels included in the super pixel with the label of i;
s2: detecting an initial edge line of an image: detecting the smooth image edge line by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result, wherein the method comprises the following steps:
s21: calculating image pixel gradient: calculating the pixel gradient amplitude and direction after image smoothing, wherein:
calculating the gradient amplitude of the image pixel in the X coordinate direction and the Y coordinate direction, namely:
G x (i,j)＝S(i+1,j)-S(i-1,j) (2)
G y (i,j)＝S(i,j+1)-S(i,j-1) (3)
calculating the gradient amplitude and direction of the image pixel, namely:
wherein: (i, j) is the position of the pixel in the XY coordinate axis; g x (i, j) and G y (i, j) are the gradient amplitudes of the pixel in the X-axis and Y-axis directions at the (i, j) position, respectively; g (i, j)And θ (i, j) is the gradient magnitude and direction of the pixel at the (i, j) location, respectively;
s22: obtaining candidate edge line points: processing the gradient image of the image by using a non-maximum suppression algorithm to obtain a local maximum on the edge point, wherein the local maximum is used as an image candidate edge line point;
s23: tracking an image edge line: based on the candidate edge line points, tracking the edge line points and connecting the edge line points by using a dual-threshold edge line tracking algorithm to obtain an initial image edge line detection result;
s3: optimizing the detection result of the image edge line: and optimizing the initial image edge line detection result by using a skeleton line extraction algorithm to obtain an image edge line detection result which is smooth, continuous and good in single-pixel and image ground object boundary attachment.
2. The method for rapidly and accurately detecting the edge line of the high-resolution remote sensing image according to claim 1, wherein in the step S1, a simple linear iterative clustering algorithm is used for converting the image into a 5-dimensional feature vector under a CIELAB color space and XY coordinates, then a measurement standard is constructed for the 5-dimensional feature vector, and local clustering is carried out on image pixels to generate superpixels;
compared with superpixels generated by an image over-segmentation algorithm of a watershed algorithm, a region growing algorithm and a graph-based image segmentation algorithm, the superpixels generated by the simple linear iterative clustering algorithm have better ground object boundary attachment and a more regular and compact shape, and the number of the superpixels generated by the simple linear iterative clustering algorithm through artificial control and good noise resistance.
3. The method for rapidly and accurately detecting the edge line of the high-resolution remote sensing image according to claim 1, wherein in the step S22, the non-maximum suppression algorithm comprises the following steps:
and performing interpolation calculation in a 3 × 3 neighborhood taking the pixel (i, j) as the center, wherein if the gradient amplitude of the pixel (i, j) is larger than the gradient amplitude of two adjacent interpolation values in the gradient direction theta (i, j), the pixel (i, j) is taken as a candidate edge pixel, and otherwise, the pixel (i, j) is taken as a non-edge pixel.
4. The method for rapidly and accurately detecting the edge line of the high-resolution remote sensing image according to claim 1, wherein in the step S23, in a dual-threshold edge line tracking algorithm:
s231: setting a low gradient amplitude threshold T l And a high gradient magnitude threshold T h ；
S232: the gradient amplitude of the pixel in the gradient image of the image is larger than T h The gradient amplitude is less than T l The gradient amplitude is larger than T l And is less than T h The pixels of (2) are taken as suspected edge line points;
s233: for the suspected edge line point, if the pixel adjacent to the suspected edge line point is the edge line point, the suspected edge line point is the edge line point; otherwise, it is a non-edge line point.
5. The method for rapidly and accurately detecting the edge line of the high-resolution remote sensing image according to claim 1, wherein in step S3, if the detection result of the edge line of the image obtained by using the dual-threshold edge line tracking algorithm has the phenomena of fracture, burr and edge line redundancy, the detection result of the initial edge line of the image needs to be optimized, and the detection result of the initial edge line of the image needs to be optimized by using the skeleton line extraction algorithm, so that the detection result of the edge of the image, which is smooth, continuous, and has good boundary adherence between single pixel and the ground object of the image, is obtained.
6. The method for quickly and accurately detecting the edge line of the high-resolution remote sensing image according to claim 5, wherein in the step S3, the skeleton line extraction algorithm comprises two steps, firstly, the edge line points of the initial edge line detection result are subjected to expansion processing, and specifically, pixels in eight neighborhood ranges of the edge line points are all used as the edge line points; and then thinning the edge line points after the expansion processing, thereby obtaining an image edge detection result which is smooth, continuous and has good single-pixel and image ground object boundary adherence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210484485.0A CN114842354B (en) | 2022-05-06 | 2022-05-06 | Quick and accurate detection method for edge line of high-resolution remote sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210484485.0A CN114842354B (en) | 2022-05-06 | 2022-05-06 | Quick and accurate detection method for edge line of high-resolution remote sensing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114842354A true CN114842354A (en) | 2022-08-02 |
| CN114842354B CN114842354B (en) | 2024-03-22 |
Family
ID=82568210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210484485.0A Active CN114842354B (en) | 2022-05-06 | 2022-05-06 | Quick and accurate detection method for edge line of high-resolution remote sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114842354B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182754A (en) * | 2014-08-19 | 2014-12-03 | 山东临沂烟草有限公司 | Rural resident point information extraction method based on high-resolution remote-sensing image |
| CN105427291A (en) * | 2015-11-12 | 2016-03-23 | 北京建筑大学 | Method for detecting vector edges of multispectral remote sensing images |
| AU2016101504A4 (en) * | 2016-08-15 | 2016-09-29 | Macau University Of Science And Technology | Method for edge detection |
| WO2018107939A1 (en) * | 2016-12-14 | 2018-06-21 | 国家海洋局第二海洋研究所 | Edge completeness-based optimal identification method for image segmentation |
| CN111985329A (en) * | 2020-07-16 | 2020-11-24 | 浙江工业大学 | Remote sensing image information extraction method based on FCN-8s and improved Canny edge detection |
-
2022
- 2022-05-06 CN CN202210484485.0A patent/CN114842354B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182754A (en) * | 2014-08-19 | 2014-12-03 | 山东临沂烟草有限公司 | Rural resident point information extraction method based on high-resolution remote-sensing image |
| CN105427291A (en) * | 2015-11-12 | 2016-03-23 | 北京建筑大学 | Method for detecting vector edges of multispectral remote sensing images |
| AU2016101504A4 (en) * | 2016-08-15 | 2016-09-29 | Macau University Of Science And Technology | Method for edge detection |
| WO2018107939A1 (en) * | 2016-12-14 | 2018-06-21 | 国家海洋局第二海洋研究所 | Edge completeness-based optimal identification method for image segmentation |
| CN111985329A (en) * | 2020-07-16 | 2020-11-24 | 浙江工业大学 | Remote sensing image information extraction method based on FCN-8s and improved Canny edge detection |
Non-Patent Citations (2)
| Title |
|---|
| 吴俐民;於雪琴;黄亮;: "FCM聚类算法协同Canny算子的遥感影像边缘检测方法", 测绘工程, no. 12, 25 December 2014 (2014-12-25) * |
| 张游游;王世东;: "高分辨率遥感影像线状地物提取方法研究", 青海科技, no. 02, 25 April 2020 (2020-04-25) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114842354B (en) | 2024-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106780576B (en) | RGBD data stream-oriented camera pose estimation method | |
| CN107369159B (en) | Threshold segmentation method based on multi-factor two-dimensional gray level histogram | |
| CN110807809B (en) | Light-weight monocular vision positioning method based on point-line characteristics and depth filter | |
| CN109961506A (en) | A kind of fusion improves the local scene three-dimensional reconstruction method of Census figure | |
| CN108663026B (en) | Vibration measuring method | |
| CN110245600B (en) | Unmanned aerial vehicle road detection method for self-adaptive initial quick stroke width | |
| CN108537816B (en) | Salient object segmentation method based on superpixel and background connection prior | |
| CN109708658B (en) | Visual odometer method based on convolutional neural network | |
| CN110288571B (en) | High-speed rail contact net insulator abnormity detection method based on image processing | |
| CN113888461A (en) | Method, system and equipment for detecting defects of hardware parts based on deep learning | |
| CN110210584A (en) | A kind of high speed dynamic two-dimension code recognition methods based on improvement SURF composite algorism | |
| CN109064473B (en) | 2.5D ultrasonic panoramic image segmentation method | |
| CN114972423A (en) | Aerial video moving target detection method and system | |
| CN113205494B (en) | Infrared small target detection method and system based on adaptive scale image block weighting difference measurement | |
| CN113095385A (en) | Multimode image matching method based on global and local feature description | |
| CN110717910B (en) | CT image target detection method based on convolutional neural network and CT scanner | |
| CN114842354B (en) | Quick and accurate detection method for edge line of high-resolution remote sensing image | |
| CN109410227B (en) | GVF model-based land utilization pattern spot contour extraction algorithm | |
| CN115760790A (en) | Airport pavement foreign matter detection method, server and storage medium | |
| CN111784634B (en) | Corner detection method based on Harris-CPDA | |
| CN110298799B (en) | PCB image positioning correction method | |
| CN110717471B (en) | B-ultrasonic image target detection method based on support vector machine model and B-ultrasonic scanner | |
| CN114511803A (en) | Target occlusion detection method for visual tracking task | |
| CN112651936A (en) | Steel plate surface defect image segmentation method and system based on image local entropy | |
| CN110648341B (en) | Target boundary detection method based on scale space and subgraph |
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 |