CN110838127B - Feature image edge detection method for intelligent automobile - Google Patents
Feature image edge detection method for intelligent automobile Download PDFInfo
- Publication number
- CN110838127B CN110838127B CN201911049836.XA CN201911049836A CN110838127B CN 110838127 B CN110838127 B CN 110838127B CN 201911049836 A CN201911049836 A CN 201911049836A CN 110838127 B CN110838127 B CN 110838127B
- Authority
- CN
- China
- Prior art keywords
- image
- edge
- detection
- canny
- edge detection
- 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
- 238000003708 edge detection Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000001514 detection method Methods 0.000 claims abstract description 21
- 230000009466 transformation Effects 0.000 claims abstract description 20
- 238000009499 grossing Methods 0.000 claims abstract description 8
- 238000001914 filtration Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims abstract description 3
- 238000011156 evaluation Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 238000007781 pre-processing Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a special for an intelligent automobileThe edge detection method of the sign image comprises the following steps: A. acquiring an image to be detected by using an on-vehicle industrial camera, converting the image into a gray image, and selecting a sigma value of 1.0 for Gaussian smoothing filtering; B. pre-extracting a Canny edge with a threshold value of 0.5 from the smoothed image; C. inputting the Canny detected edge image matrix into a two-dimensional digital filter, detecting the mode maximum point of the two-dimensional wavelet transformation through binary wavelet transformation, and determining the edge point of the image; connecting the maximum points of the mould under the scale along the boundary direction to form an edge curve, thereby obtaining an edge detection image; D. selecting proper E value according to different types of pictures in different environments, and calculating E k Judgment e k And C, outputting an edge detection image if the E is not less than or equal to the E, and repeating the step C if the E is not less than the E. The method can obtain the characteristic image edge contour line which is easy to identify and judge by the intelligent automobile, and the detection edge is accurate and clear.
Description
Technical Field
The invention relates to the field of image edge detection, in particular to a feature image edge detection method for an intelligent automobile.
Background
In recent years, intelligent automobiles become research hotspots in the automobile industry, and represent the development direction of the automobile industry in the future. The environment sensing system is used as an 'eye' of the intelligent automobile, and environment information sensing is mainly carried out by means of various sensors to make further judgment, wherein an image recognition technology based on machine vision is widely adopted in the intelligent automobile environment sensing system due to strong adaptability and low cost.
The image recognition technology based on machine vision mainly comprises the following stages: image preprocessing stage (such as denoising, image enhancement, morphological processing), image segmentation or object separation stage, feature extraction stage, and decision classification stage. In intelligent automobile image recognition, various obstacles, marks and the like such as automobiles, lane lines, pedestrians, traffic signs and the like need to be detected to make corresponding judgment, and preprocessing and feature extraction of images acquired by the intelligent automobiles are the basis for image recognition judgment classification.
The edge detection of the image is the basis and the core of the whole image preprocessing and the feature extraction, and only the feature image outline acquired by the intelligent automobile is accurately and clearly extracted, the further feature recognition and other works can be carried out on the feature image outline. The edge of the image is an area with the gray value changed drastically in the image, and mainly exists between targets, between targets and backgrounds, and between areas, and the more well-known edge detection operators mainly include a Sobel operator, a Laplacian operator, a Robert operator, a Canny operator and the like. The Canny operator has a function remarkably superior to other operators and is regarded as an operator with most classical edge detection, but the Canny operator is particularly sensitive to image noise in the process of non-maximum suppression of gradient amplitude, false edges can be generated, and particularly, the probability of generating the false edges is higher due to the fact that a plurality of uncertain factors exist in an image acquired by an intelligent automobile, and the requirement on the next step of feature extraction is higher.
In view of this, the edge detection method in the prior art is not fully suitable for detecting the edge of the feature image of the intelligent automobile, so that it is required to develop a feature image edge detection method for the intelligent automobile, which has low sensitivity to noise and accurate and clear detection edge.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a characteristic image edge detection method for an intelligent automobile, which solves the problems that the edge detection method in the prior art has high sensitivity to noise and is easy to generate false edges.
In order to achieve the above purpose, the technical scheme provided by the invention is as follows:
the edge detection method for the feature image of the intelligent automobile comprises the following steps:
A. acquiring an image to be detected by using an on-vehicle industrial camera, converting the image into a gray image, and selecting a sigma value of 1.0 for Gaussian smoothing filtering;
B. pre-extracting a Canny edge with a threshold value of 0.5 from the smoothed image;
C. inputting the Canny detected edge image matrix into a two-dimensional digital filter, detecting the mode maximum point of the two-dimensional wavelet transformation through binary wavelet transformation, and determining the edge point of the image; connecting the maximum points of the mould under the scale along the boundary direction to form an edge curve, thereby obtaining an edge detection image;
D. selecting proper E value according to different types of pictures in different environments, and calculating E k Judgment e k C, outputting an edge detection image if the E is not less than or equal to the E, and repeating the step C if the E is not less than the E;
further, the step A specifically includes:
a1, carrying out gray processing on an image acquired by an industrial camera to obtain an image to be processed;
a2, taking the horizontal and vertical coordinate values of each position of the original digital image o (x, y) into a two-dimensional Gaussian function formula for filtering and smoothing, and taking 1.0 from the standard deviation sigma of the Gaussian function to obtain a smoothed image h (x, y).
Further, the step B performs an edge detection on the smoothed image h (x, y) based on the Canny operator to obtain a pre-extracted edge image f (x, y).
Further, the step C specifically includes:
c1, inputting an edge image f (x, y) detected based on a Canny operator into a two-dimensional digital filter;
c2, f (x, y) at the scale s=2 j The upper two-dimensional wavelet transform contains two components W x f(2 j ，x，y)、W y f(2 j Calculating the modulus ratio value Mf (2) of each pixel point (x, y) j X, y), calculating the gradient vector and horizontal phase angle Af (2) at each pixel point (x, y) j Determining a threshold T (T > 0), obtaining boundary points (m, n), and connecting the boundary points along the boundary directionForming an edge curve.
Further, the step D specifically includes:
d1, selecting an optimal edge detection image under 1-3-level wavelet transformation by taking detection definition as an index, and calculating e through an evaluation function k The value is used as an evaluation index, the whole image is divided into a grid of w×h, and the ith Canny is used for detecting edge points (x i ，y i ) For the reference point, the coordinates are plotted as (x i ±a，y i Rectangular frame of + -a), solving for point (x) i ，y i ) The distance to the jth level wavelet detection edge curve within the rectangular box,
in the method, in the process of the invention,
(x i ，y i ) Representing the coordinates of the Canny detection edge points selected in the ith square;
d2, selecting proper E value as judgment basis according to different types of pictures in different environments, and judging E k And if not, outputting the level wavelet detection edge curve, and if not, performing level one wavelet transformation edge detection, so that the contour is clearer.
The invention provides a characteristic image edge detection method for an intelligent automobile, which is suitable for the intelligent automobile to detect images acquired under different objects, different illuminance and other conditions, has good edge detection result, low noise sensitivity, accurate and clear detection edge, can obtain an optimal image in 1-3-level wavelet transformation for further judgment and identification, can be used for extracting the edge contour of the acquired image, and has an important role in the development of image identification of intelligent automobile machine vision.
Drawings
FIG. 1 is an overall flow chart of a feature image edge detection method for an intelligent automobile of the present invention;
fig. 2 (a) is a front image of a vehicle to be detected;
FIG. 2 (b) is an image with a threshold of 0.3 using Canny edge detection only;
FIG. 2 (c) is an image with a threshold of 0.5 using Canny edge detection only;
FIG. 2 (d) is an image with a threshold of 0.7 using Canny edge detection only;
FIG. 3 is an edge detection image of a vehicle in different orientations in accordance with the method of the present invention;
fig. 4 is an edge detection image of a traffic sign.
Detailed Description
The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.
Example 1
Step 1), acquiring front, side and tail images of an automobile at the same time and with the same illuminance and in different directions, converting the front, side and tail images into gray images, taking the horizontal and vertical coordinate values of each position of an original digital image o (x, y) into a two-dimensional Gaussian function formula, filtering and smoothing, and taking a Gaussian function standard deviation sigma to 1.0 to obtain a smoothed image h (x, y);
and 2) carrying out primary edge detection on the smoothed image h (x, y) based on a Canny operator to obtain a pre-extraction edge image f (x, y), and selecting a threshold value to be 0.5 when specific edge pre-extraction is carried out based on the Canny operator. Acquiring front, side and tail Canny edge detection images shown in fig. 3;
and 3) inputting the edge image f (x, y) detected based on the Canny operator into a two-dimensional digital filter.
f (x, y) is at the scale s=2 j The upper two-dimensional wavelet transform contains two components W x f(2 j ，x，y)、W y f(2 j Calculating the modulus ratio value Mf (2) of each pixel point (x, y) j X, y), calculating the gradient vector and horizontal phase angle Af (2) at each pixel point (x, y) j X, y). Determining a threshold T (T > 0), if Mf (2) at pixel point (m, n) j ，m，n)...T，Mf(2 j M, n) takes a local maximum, the point (m, n) is the modulo maximum point, i.e. a boundary point. And connecting two adjacent boundary points in each discrete sampling grid point to form a maximum curve along the boundary, thereby obtaining an edge detection image under the scale.
Step 4), calculating e, wherein a large number of defects exist in the first-stage wavelet transformation edge detection image contour k And comparing the value with the selected E value, wherein the E value is selected according to the environment where the intelligent automobile is positioned and the output of the illumination sensor, and the selection standard is shown in the following table.
Table 1E value selection table (a < b < c < d < e)
Judgment e k Not less than E? If not, repeating the step 3) to obtain a secondary wavelet transformation edge detection image, wherein the image contour is clear but still contains partial deletion, and the contour shape is finely calculated as e k Again judge e k Not less than E? If not, repeating the step 3) to obtain three-level wavelet transformationEdge detection image e k And the detection result meets the requirement of vehicle identification and is easier to judge.
The specific calculation method of the step 4) is as follows:
selecting an optimal edge detection image under 1-3-level wavelet transformation by taking detection definition as an index, and calculating e through an evaluation function k The value is an evaluation index. The whole image is divided into a grid of w x h.
Calculating the average distance between the j-th level wavelet transformation image and the Canny detection edge:
in the method, in the process of the invention,
(x i ，y i ) Representing the coordinates of the Canny detected edge points selected in the ith square.
Order the
The rewriting is as follows:
e is as above k The calculation amount of the traversal algorithm for detecting the edge points of all wavelets for a certain Canny edge point is quite huge, and for reducing the calculation, the ith Canny edge point is used for detecting the edge points (x i ，y i ) For the reference point, the coordinates are plotted as (x i ±a，y i Rectangular frame of + -a), solving for point (x) i ，y i ) The distance to the jth level wavelet detection edge curve within the rectangular box.
Selecting proper E value as judgment basis according to different types of pictures in different environments, and judging E k Not less than E? If yes, outputting the wavelet detection edge curve of the stage, and if not, performing the first-stage edge detection, so that the contour is clearer.
Example 2
Step 1), obtaining a speed-limiting traffic sign image, converting the speed-limiting traffic sign image into a gray image, and taking the horizontal and vertical coordinate values of each position of an original digital image o (x, y) into a two-dimensional Gaussian function formula for filtering and smoothing, wherein the standard deviation sigma of the Gaussian function is 1.0, and obtaining a smoothed image h (x, y).
And 2) carrying out edge detection on the smoothed image h (x, y) once based on a Canny operator to obtain a pre-extracted edge image f (x, y). The threshold is chosen to be 0.5 when the particular edge pre-extraction is based on the Canny operator.
And 3) inputting the edge image f (x, y) detected based on the Canny operator into a two-dimensional digital filter.
f (x, y) is at the scale s=2 j The upper two-dimensional wavelet transform contains two components W x f(2 j ，x，y)、W y f(2 j Calculating the modulus ratio value Mf (2) of each pixel point (x, y) j X, y), each pixel is calculatedGradient vector and horizontal phase angle Af (2) at point (x, y) j ，x，y)。
Determining a threshold T (T > 0), if Mf (2) at pixel point (m, n) j ，m，n)...，Mf(2 j M, n) takes a local maximum, the point (m, n) is the modulo maximum point, i.e. a boundary point. And connecting two adjacent boundary points in each discrete sampling grid point to form a maximum curve along the boundary, thereby obtaining an edge detection image under the scale.
Step 4), detecting partial missing of the image contour by the first-stage wavelet transformation edge, and calculating e, wherein the partial missing is not easy to further detect k Comparing the value with the selected E value, wherein the E value is selected according to the environment where the intelligent automobile is positioned and the output of the illumination sensor, the selection standard is shown in table 1, and the E is judged k Not less than E? If not, repeating the step 3) to obtain a secondary wavelet transformation edge detection image, at the moment, highlighting the edge of the image, smoothing the fine texture structure of the image, being the most readable, having smaller possibility of misjudgment, and calculating e k ，e k And (5) outputting the edge detection image under the scale.
The specific calculation method of the step 4) is as follows:
selecting an optimal edge detection image under 1-3-level wavelet transformation by taking detection definition as an index, and calculating e through an evaluation function k The value is an evaluation index. The whole image is divided into a grid of w x h.
Calculating the average distance between the j-th level wavelet transformation image and the Canny detection edge:
in the method, in the process of the invention,
(x i ，y i ) Representing the coordinates of the Canny detected edge points selected in the ith square.
Order the
The rewriting is as follows:
e is as above k The calculation amount of the traversal algorithm for detecting the edge points of all wavelets for a certain Canny edge point is quite huge, and for reducing the calculation, the ith Canny edge point is used for detecting the edge points (x i ，y i ) For the reference point, the coordinates are plotted as (x i ±a，y i Rectangular frame of + -a), solving for point (x) i ，y i ) The distance to the jth level wavelet detection edge curve within the rectangular box.
Selecting proper E value as judgment basis according to different types of pictures in different environments, and judging E k Not less than E? If yes, outputting the wavelet detection edge curve of the stage, and if not, performing the first-stage edge detection, so that the contour is clearer.
The invention provides a characteristic image edge detection method for an intelligent automobile, which is suitable for edge detection of characteristic images acquired by the intelligent automobile through pretreatment, wavelet transformation and other methods.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather to enable any modification, equivalent replacement, improvement or the like to be made within the spirit and principles of the invention.
Claims (4)
1. The characteristic image edge detection method for the intelligent automobile is characterized by comprising the following steps of:
A. acquiring an image to be detected by using an on-vehicle industrial camera, converting the image into a gray image, and selecting a sigma value of 1.0 for Gaussian smoothing filtering;
B. pre-extracting a Canny edge with a threshold value of 0.5 from the smoothed image;
C. inputting the Canny detected edge image matrix into a two-dimensional digital filter, detecting the mode maximum point of the two-dimensional wavelet transformation through binary wavelet transformation, and determining the edge point of the image; dimension s=2 along the boundary direction j The maximum points of the lower die are connected to form an edge curve, so that an edge detection image is obtained;
D. selecting proper E value according to different types of pictures in different environments, and calculating E k Judgment e k C, outputting an edge detection image if the E is not less than or equal to the E, and repeating the step C if the E is not less than the E;
the step D specifically comprises the following steps:
d1, selecting an optimal edge detection image under 1-3-level wavelet transformation by taking detection definition as an index, and calculating e through an evaluation function k The value is used as an evaluation index, the whole image is divided into a grid of w×h, and the ith Canny is used for detecting edge points (x i ,y i ) For the reference point, the coordinates are plotted as (x i ±a,y i Rectangular frame of + -a), solving for point (x) i ,y i ) The distance to the jth level wavelet detection edge curve within the rectangular box,
in the method, in the process of the invention,
(x i ,y i ) Representing the coordinates of the Canny detection edge points selected in the ith square;
d2, selecting proper E value as judgment basis according to different types of pictures in different environments, and judging E k If yes, outputting the wavelet detection edge curve of the stage, and if not, performing primary wavelet transformation edge detection, so that the contour is clearer.
2. The method for detecting the edges of the feature images of the intelligent automobile according to claim 1, wherein the step a specifically comprises:
a1, carrying out gray processing on an image acquired by an industrial camera to obtain an image to be processed;
a2, taking the horizontal and vertical coordinate values of each position of the original digital image o (x, y) into a two-dimensional Gaussian function formula for filtering and smoothing, and taking 1.0 from the standard deviation sigma of the Gaussian function to obtain a smoothed image h (x, y).
3. The method for detecting the edges of the feature images of the intelligent automobile according to claim 2, wherein the step B performs edge detection on the smoothed image h (x, y) once based on a Canny operator to obtain a pre-extracted edge image f (x, y).
4. The method for detecting the edges of the feature images of the intelligent automobile according to claim 3, wherein the step C specifically comprises:
c1, inputting an edge image f (x, y) detected based on a Canny operator into a two-dimensional digital filter;
c2, f (x, y) at the scale s=2 j The upper two-dimensional wavelet transform contains two components W x f(2 j ,x,y)、W y f(2 j Calculating the modulus ratio value Mf (2) of each pixel point (x, y) j X, y), calculating the gradient vector and horizontal phase angle Af (2) at each pixel point (x, y) j X, y), determining a threshold T, where T>0, obtaining boundary points (m, n), and connecting the boundary points along the boundary direction to form an edge curve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911049836.XA CN110838127B (en) | 2019-10-30 | 2019-10-30 | Feature image edge detection method for intelligent automobile |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911049836.XA CN110838127B (en) | 2019-10-30 | 2019-10-30 | Feature image edge detection method for intelligent automobile |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110838127A CN110838127A (en) | 2020-02-25 |
| CN110838127B true CN110838127B (en) | 2023-06-23 |
Family
ID=69575867
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911049836.XA Active CN110838127B (en) | 2019-10-30 | 2019-10-30 | Feature image edge detection method for intelligent automobile |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110838127B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101001691B1 (en) * | 2006-03-13 | 2010-12-15 | 노틸러스효성 주식회사 | Recognizing the Denomination of a Note Using Wavelet transform |
| WO2019041590A1 (en) * | 2017-08-31 | 2019-03-07 | 中国科学院微电子研究所 | Edge detection method using arbitrary angle |
| CN108596931A (en) * | 2018-05-11 | 2018-09-28 | 浙江大学 | A kind of noise robustness hub edge detection algorithm based on Canny operators |
-
2019
- 2019-10-30 CN CN201911049836.XA patent/CN110838127B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110838127A (en) | 2020-02-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898610B (en) | Object contour extraction method based on mask-RCNN | |
| CN109101924B (en) | Machine learning-based road traffic sign identification method | |
| CN115082419B (en) | Blow-molded luggage production defect detection method | |
| KR101403876B1 (en) | Method and Apparatus for Vehicle License Plate Recognition | |
| CN107045634B (en) | Text positioning method based on maximum stable extremum region and stroke width | |
| JP5223675B2 (en) | Vehicle detection device, vehicle detection method, and vehicle detection program | |
| KR100975749B1 (en) | Method for recognizing lane and lane departure with Single Lane Extraction | |
| KR101719549B1 (en) | Image preprocessing method for license plate recognition and license plate recognition device using thereof | |
| Paunwala et al. | A novel multiple license plate extraction technique for complex background in Indian traffic conditions | |
| CN107315990B (en) | Pedestrian detection algorithm based on XCS-LBP characteristics | |
| CN111027544B (en) | MSER license plate positioning method and system based on visual saliency detection | |
| CN114972356A (en) | Plastic product surface defect detection and identification method and system | |
| CN114511770A (en) | Road sign plate identification method | |
| Gilly et al. | A survey on license plate recognition systems | |
| CN110348307B (en) | Path edge identification method and system for crane metal structure climbing robot | |
| CN108537815B (en) | Video image foreground segmentation method and device | |
| CN110633635A (en) | ROI-based traffic sign board real-time detection method and system | |
| CN116703895B (en) | Small sample 3D visual detection method and system based on generation countermeasure network | |
| Khin et al. | License plate detection of Myanmar vehicle images captured from the dissimilar environmental conditions | |
| CN106446832B (en) | Video-based pedestrian real-time detection method | |
| CN110838127B (en) | Feature image edge detection method for intelligent automobile | |
| CN111241911A (en) | Self-adaptive lane line detection method | |
| KR102585146B1 (en) | Pothole detection device and method based on deep learning | |
| CN109886120B (en) | Zebra crossing detection method and system | |
| Setiyorini et al. | Vehicle number plate identification using template matching algorithm for automatic parking system |
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 |