CN111739006B - Elliptical image detection algorithm and system based on enclosed road integral - Google Patents
Elliptical image detection algorithm and system based on enclosed road integral Download PDFInfo
- Publication number
- CN111739006B CN111739006B CN202010576856.9A CN202010576856A CN111739006B CN 111739006 B CN111739006 B CN 111739006B CN 202010576856 A CN202010576856 A CN 202010576856A CN 111739006 B CN111739006 B CN 111739006B
- Authority
- CN
- China
- Prior art keywords
- ellipse
- integral
- image
- elliptical
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
- G06F17/13—Differential equations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—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
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Quality & Reliability (AREA)
- Algebra (AREA)
- Operations Research (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an ellipse image detection algorithm based on a peripheral path integral, which comprises the following steps: step S1, inputting an image to be detected; step S2, preprocessing the image to be detected to generate a binary edge image; step S3, performing integral operation on the pixel value of the ellipse peripheral coordinate in the binary edge image to obtain an ellipse peripheral road integral pixel density value; step S4, repeating step S3 until all ellipse peripheral coordinates of the binary edge image are traversed to obtain all ellipse peripheral path integral pixel density values; step S5, reading the integral pixel density values of one or more groups of elliptic enclosing channels with the maximum difference between two adjacent elliptic enclosing channels; step S6, determining whether the read density value of the integral pixel of the elliptical peripheral channel forms an ellipse, if so, prompting that the identification is successful, outputting an elliptical parameter, and if not, prompting that the identification is failed. The invention can obviously reduce the calculation amount, is beneficial to saving the system cost, improving the operation precision and improving the operation efficiency.
Description
Technical Field
The invention relates to an appointed image recognition algorithm of a computer vision system, in particular to an ellipse image detection algorithm and system based on a peripheral path integral.
Background
With the continuous development of scientific technology, the application of computer vision (computer vision) in our work and life has become more and more extensive. At present, the application direction of computer vision mainly focuses on relatively macroscopic image analysis, statistics, and comparison, such as real-time traffic flow analysis, license plate recognition, fingerprint recognition, face recognition, and other technologies.
Compared with a macro recognition scene, the micro recognition scene does not require complex analysis and comparison of images, but rather performs identification and positioning on elements with a small difference, such as accurate positioning of an elliptical figure in an image, wherein an ellipse is a common geometric form in many professional fields, and for example, in some high-end manufacturing process flows, computer vision is used for detecting whether a distance between two elliptical parts meets a technical standard, so that the micro recognition scene is a typical application scene.
In the prior art, there are two main common ellipse positioning methods based on computer vision:
first, a Hough Transform (Hough Transform) based positioning algorithm is used. The core idea of the algorithm is that the shape detection problem under the Cartesian coordinate is converted into a peak value statistics problem under the polar coordinate based on the coordinate space transformation between the Cartesian coordinate and the polar coordinate. And under the polar coordinate, calculating whether pixels are located on the same ellipse circle through a voting mechanism. The disadvantage of the hough transform is that a large amount of calculation is required for each point of the image, and the calculation cost is high.
Second, the circle is located by Deep Learning (Deep Learning). At present, a plurality of deep learning frameworks which can be used for commercial use in open source communities, such as Caffe, tensrflow, Keras and the like, can be recognized by developers as long as the frameworks are sufficiently trained. The training refers to a process of taking mass images as training data, then carrying out operation by a neural network model, and finally extracting model parameters. Deep learning generally requires extensive training to perform accurate recognition, and thus cost issues become a significant obstacle if such techniques need to be applied in real-life projects.
Disclosure of Invention
The invention aims to solve the technical problem of providing an elliptical image detection algorithm and system based on a circular track integral, which can obviously reduce the calculation amount, help to save the system cost, improve the operation precision and improve the operation efficiency, aiming at the defects of the prior art.
In order to solve the technical problems, the invention adopts the following technical scheme.
An ellipse image detection algorithm based on a circular track integral is realized based on a system, the system comprises a preprocessing unit, an ellipse circular track integral operation unit, an ellipse identification unit and a result output unit, and the algorithm comprises the following steps: step S1, inputting an image to be detected; step S2, the preprocessing unit preprocesses the image to be detected to generate a binary edge image; step S3, in the binary edge image, the elliptical periphery integral operation unit performs integral operation on the pixel value of the elliptical periphery coordinate to obtain an elliptical periphery integral pixel density value; step S4, repeating step S3 until the ellipse periphery integral operation unit traverses all ellipse periphery coordinates of the binary edge image to obtain all ellipse periphery integral pixel density values; step S5, the ellipse identifying unit reads one or more sets of ellipse contour integral pixel density values with the largest difference between two adjacent sets from the operation result of step S4; step S6, the ellipse identifying unit determines whether the density value of the integrated pixel of the elliptical peripheral road read in step S5 constitutes an ellipse, if so, step S7 is executed, and if not, step S8 is executed; step S7, the result output unit prompts the success of identification and outputs ellipse parameters; in step S8, the result output unit prompts for a recognition failure.
Preferably, in step S2, the preprocessing unit preprocesses the image to be detected by using a preset Canny edge detection operator.
Preferably, a threshold is preset in the ellipse recognizing unit, and in step S6, the determining process of the ellipse recognizing unit includes: step S50, the ellipse recognizing unit determines whether the density value of the integrated pixel of the elliptical peripheral road read in step S5 is higher than the threshold, if yes, step S51 is executed, and if no, step S52 is executed; step S51, the ellipse identifying unit takes the smallest ellipse parameter in the group of ellipse periphery integral pixel density values as the target ellipse which is successfully identified; in step S52, the result output unit prompts that the recognition is failed.
Preferably, the ellipse identifying unit uses 1/3 of the integral value of one complete ellipse lane as the threshold value.
Preferably, when the image to be detected is an arc, it is regarded as a partial circular integration operation, and after the threshold is adjusted, the arc is identified in combination with steps S2 to S8.
An elliptical image detection system based on a circular track integral comprises a preprocessing unit, an elliptical circular track integral operation unit, an ellipse identification unit and a result output unit, wherein: the preprocessing unit is used for preprocessing an image to be detected to generate a binary edge image; the ellipse surrounding track integral operation unit is used for performing integral operation on the pixel value of the ellipse surrounding coordinate in the binary edge image to obtain an ellipse surrounding track integral pixel density value, and traversing all ellipse surrounding coordinates of the binary edge image to obtain all ellipse surrounding track integral pixel density values; the ellipse identification unit is used for reading one or more groups of ellipse surrounding channel integral pixel density values with the largest difference values from the operation results of the ellipse surrounding channel integral operation unit, then judging whether the read ellipse surrounding channel integral pixel density values form an ellipse or not, if so, the result output unit prompts the success of identification and outputs ellipse parameters, and if not, the result output unit prompts the failure of identification.
In the oval image detection algorithm based on the girth integral, a binary edge image is generated according to an image to be detected, integral operation is carried out on pixel values of oval peripheral coordinates in the binary edge image until all the oval peripheral coordinates of the binary edge image are traversed to obtain all oval girth integral pixel density values, then two adjacent oval girth integral pixel density values with the largest difference value are read, whether the read oval girth integral pixel density values form an oval or not is judged, and finally identification success or identification failure information is output. Based on the process, the method not only realizes the detection operation of the elliptical image, but also obviously reduces the calculation amount, is beneficial to saving the system cost, and simultaneously greatly improves the operation precision, is beneficial to improving the operation efficiency and better meets the application requirement compared with the existing algorithm.
Drawings
FIG. 1 is a block diagram of an elliptical image detection system based on the integral of the circular path according to the present invention;
FIG. 2 is a flow chart of an elliptical image detection algorithm based on the circular road integral of the present invention;
FIG. 3 is a schematic diagram of track point coordinates of a perfect circle image;
FIG. 4 is a diagram showing coordinates of trace points of an elliptical image
Fig. 5 is a schematic diagram of track point coordinates of the elliptical image.
Detailed Description
The invention is described in more detail below with reference to the figures and examples.
The invention discloses an elliptical image detection algorithm based on a surrounding road integral, which is realized based on a system by combining with a figure 1 and a figure 2, wherein the system comprises a preprocessing unit 1, an elliptical surrounding road integral operation unit 2, an ellipse recognition unit 3 and a result output unit 4, and the algorithm comprises the following steps:
step S1, inputting an image to be detected;
step S2, the preprocessing unit 1 preprocesses the image to be detected to generate a binary edge image;
step S3, in the binary edge image, the elliptical periphery integral operation unit 2 performs integral operation on the pixel value of the elliptical periphery coordinate to obtain an elliptical periphery integral pixel density value;
step S4, repeating step S3 until the ellipse periphery integral operation unit 2 traverses all ellipse periphery coordinates of the binary edge image to obtain all ellipse periphery integral pixel density values;
step S5, the ellipse identifying unit 3 reads one or more sets of ellipse surround channel integrated pixel density values with the largest difference values from the operation result of the step S4;
step S6, the ellipse identifying unit 3 determines whether the density value of the integrated pixel of the elliptical peripheral road read in step S5 constitutes an ellipse, if so, step S7 is executed, and if not, step S8 is executed;
step S7, the result output unit 4 prompts the success of identification and outputs ellipse parameters;
in step S8, the result output unit 4 prompts for a failure in recognition.
In the algorithm, a binary edge image is generated according to an image to be detected, integral operation is carried out on pixel values of elliptical peripheral coordinates in the binary edge image until all elliptical peripheral coordinates of the binary edge image are traversed to obtain all elliptical peripheral track integral pixel density values, then two adjacent elliptical peripheral track integral pixel density values with the largest difference value are read, whether the read elliptical peripheral track integral pixel density values form an ellipse or not is judged, and finally information of successful identification or failed identification is output. Based on the process, the method not only realizes the detection operation of the elliptical image, but also obviously reduces the calculation amount, is beneficial to saving the system cost, and simultaneously greatly improves the operation precision, is beneficial to improving the operation efficiency and better meets the application requirement compared with the existing algorithm.
In this embodiment, each two adjacent elliptical periphery integrated values are used as a set of data, then the difference between the two elliptical periphery integrated values in the set of data is calculated, and finally, the set with the larger difference is read from the sets of data, that is, the step S5 is as follows: the ellipse identification unit 3 reads the density values of one or more groups of ellipse contour integral pixels with the maximum difference between two adjacent ellipse contour integral pixels. The reading mode has the effect of finding a group of data with the largest peripheral integration difference, the group of data is the position of the ellipse, the transition change of the edge of the ellipse and the background pixel is strongest due to the fact that the image is subjected to binarization processing, and the calculation of the difference value is the strongest transition change of the edge of the ellipse and the background pixel.
In a preferable mode, in step S2, the preprocessing unit 1 preprocesses the image to be detected by using a preset Canny edge detection operator.
In this embodiment, the ellipse circular path integral operation is actually to add the pixel values of the ellipse circumference coordinates in the image to obtain the ellipse circular path integral density value. For the perfect circle graph, please refer to fig. 3, first, the circular trace point coordinates are uniformly distributed on the circle as the constant of the circle center angle θ increases, that is, the circular trace point coordinates can correctly calculate the pixel density value of the circle according to the circular trace point coordinates.
However, in an ellipse, see fig. 4, neither from the ellipse center point C, nor from the focus S, the constant increase in the angle θ is discretely and uniformly distributed on the trajectory as in a perfect circle. Therefore, in order to calculate a more accurate value of the circular track integral, the present embodiment needs to develop a set of elliptical circular track integral algorithm to calculate the circular track integral value of the elliptical track point more accurately.
In this regard, in step S3, the operation procedure of the elliptical periphery integral operation unit 2 includes:
step S30, the coordinates of the center point of the ellipse are set to (0, 0), and the ellipse is expressed by the following formula:
wherein x and y are Cartesian coordinates of any point P of the ellipse circumference, a is the ellipse major semi-axis, and b is the ellipse minor semi-axis;
step S31, after parameterizing the formula, introducing an angle t that increases from the center of the ellipse to obtain:
x＝a cos(t)；
y＝b sin(t)。
based on the above operators, please refer to fig. 5, the embodiment needs to perform the circle path integration on as many P points on the ellipse circumference as possible, and the implementation process of the embodiment does not make the trajectory points uniformly distributed on the ellipse circumference along with the constant increase of the angle t, but tries the angle t as many as possible to calculate whether the corresponding P points are located on the ellipse circumference. The purpose of this is to locate as many trace points as possible on the ellipse circumference, and since there is a pixel accuracy problem in the image domain, if multiple points P are located at the same coordinate during the course of the circle integration, the pixel value of only one point P is taken into the circle integration.
In addition, the angle t is from 0 during the peripherical path integration0Increase to 3600The number of steps required in the process of (a) needs to be considered especially, increasing the number of steps by a value t too small reduces the accuracy of the value of the integrated value of the roundabout, and increasing the number of steps by a value t too large slows down the efficiency of the algorithm.
In this regard, in step S31 of the present embodiment, the number of increasing steps of the value of the angle t is determined according to the ellipse circumference 2 π b +4 (a-b). Here, the number of steps is the maximum value of the for cycle in the engineering, and t originally represents the angle when the round-trip integration is performed, but the number of increases from 0 degree to 360 degrees needs to be determined manually, so the ellipse circumference is used as the maximum value of the for cycle in the present embodiment.
In practical applications, the ellipse identification operator in this embodiment defines an ellipse by using 4 parameters, i.e., a, b, x0, and y0, where a is the semi-major axis of the ellipse, b is the semi-minor axis of the ellipse, and x0 and y0 are cartesian coordinate values of the center point of the ellipse. The embodiment utilizes a, b, x0 and y0 to define the coordinate P of the ellipse circle, and the essence of the invention is to apply the channel integral in the image, and continuously and circularly try different a, b, x0 and y0 until finding the P-point pixel value with the largest difference.
Based on the above parameter basis, the present embodiment adopts an important processing means, that is: and performing multiple times of the girth integral operation in the binary edge image, and then finding out the maximum girth integral difference value so as to identify the target ellipse.
Specifically, in step S6, the ellipse identifying unit 3 identifies an ellipse based on the following operator:
wherein:
c (x, y) is a binary edge image domain generated after Canny edge detection is carried out on an image to be detected;
2 π b +4(a-b) is the ellipse perimeter;
As a preferable mode, a threshold is preset in the ellipse recognizing unit 3, and the determining process of the ellipse recognizing unit 3 in step S6 includes:
in step S50, the ellipse recognizing unit 3 determines whether the density value of the integrated pixel of the elliptical peripheral road read in step S5 is higher than the threshold, if so, performs step S51, and if not, performs step S52;
step S51, the ellipse identifying unit 3 takes the smallest ellipse parameter in the group of ellipse periphery integral pixel density values as the target ellipse for successful identification; specifically, the elliptical periphery integral pixel density value is a value obtained by adding pixel values of all P points on an elliptical circumference, and the coordinates of the P points in the image are calculated by a, b, x0 and y 0. When a group of ellipse parameters with the maximum peripheral integration difference is found, the smallest ellipse is the target ellipse, in a binary image, each pixel value of an ellipse circumference is white 255, and the background is black 0, in the engineering realization, because a and b are increased along with the for cycle, when the ellipse is detected, the ellipse circumference and the background have a change from 255 to 0, and for the group of parameters of 255, the smaller a and b parameters are the smaller ellipse parameters, namely the target ellipse;
in step S52, the result output unit 4 presents a recognition failure.
Further, the ellipse recognizing unit 3 takes 1/3 of the integral value of one complete ellipse lane as the threshold value.
The threshold value can be set or adjusted according to actual needs. Further, when the image to be detected is a perfect circle, initial values of the parameters a and b are set to a ═ b, and the perfect circle is identified in conjunction with steps S2 to S8. Similarly, when the image to be detected is an arc, it is regarded as a partial circular integration operation, and after the threshold is adjusted, the arc is identified in combination with steps S2 to S8.
It can be seen that the algorithm of the present invention can be applied not only to elliptical image detection, but also to the identification of a true circle or arc (i.e., a partial corridor).
In order to better describe the technical solution of the present invention, the present invention further relates to an elliptical image detection system based on the circular track integral, please refer to fig. 1, the system includes a preprocessing unit 1, an elliptical circular track integral operation unit 2, an ellipse recognition unit 3 and a result output unit 4, wherein:
the preprocessing unit 1 is used for preprocessing an image to be detected to generate a binary edge image;
the elliptical peripheral path integral operation unit 2 is used for performing integral operation on the pixel values of the elliptical peripheral coordinates in the binary edge image to obtain an elliptical peripheral path integral pixel density value, and traversing all elliptical peripheral coordinates of the binary edge image to obtain all elliptical peripheral path integral pixel density values;
the ellipse identification unit 3 is configured to read one or more sets of maximum difference values of the two adjacent ellipse-contour integral pixel density values from the operation result of the ellipse-contour integral operation unit 2, and then determine whether the read ellipse-contour integral pixel density values form an ellipse, if so, the result output unit 4 prompts successful identification and outputs ellipse parameters, and if not, the result output unit 4 prompts failed identification.
In the elliptical image detection algorithm and system based on the peripheral integration, firstly, a Canny edge detection operator is applied to pre-process an image to be detected to generate a binary edge image; then, a group of parameters a, b, x0 and y0 are used in the binary edge image, and ellipse surrounding track integration is carried out and a primary integration value is obtained in combination with the increase of the angle t; repeating the integration process, traversing the binary edge image along with the change of the parameters a, b, x0 and y0, finding out parameter combinations as much as possible, and acquiring the integral value of the peripheral road as much as possible; then, one or more groups of the enclosure integral values with the largest difference values are taken out, whether the taken enclosure integral values are an ellipse or not is judged, and the specific judgment process is as follows: 1/3 of a complete ellipse circular path integrated value is used as a threshold value, if the taken circular path integrated value is higher than the threshold value, the smaller ellipse parameter in the group of circular path integrated values is the target ellipse which is successfully identified, and if the circular path integrated value is lower than the threshold value, the identification is failed. Based on the principle, the invention adopts a brand-new ellipse image detection algorithm, compared with the prior art, the invention has higher identification precision, smaller calculation and lower operation cost, better fills the blank of the technology for rapidly operating and processing ellipse images, perfect circle images and arc lines in the computer vision field, and can meet the detection requirements of various industries on the ellipse images, the perfect circle images and the arc line images.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents or improvements made within the technical scope of the present invention should be included in the scope of the present invention.
Claims (6)
1. An ellipse image detection algorithm based on a circular track integral is characterized in that the algorithm is realized based on a system, the system comprises a preprocessing unit (1), an ellipse circular track integral operation unit (2), an ellipse identification unit (3) and a result output unit (4), and the algorithm comprises the following steps:
step S1, inputting an image to be detected;
step S2, the preprocessing unit (1) preprocesses the image to be detected to generate a binary edge image;
step S3, in the binary edge image, the ellipse circular track integral operation unit (2) performs integral operation on the pixel value of the ellipse circular coordinate to obtain the ellipse circular track integral pixel density value;
step S4, repeating step S3 until the ellipse periphery integral operation unit (2) traverses all ellipse periphery coordinates of the binary edge image to obtain all ellipse periphery integral pixel density values;
step S5, the ellipse identifying unit (3) reads one or more groups of ellipse circular channel integral pixel density values with the maximum difference from the operation result of the step S4;
step S6, the ellipse identifying unit (3) determines whether the density value of the integrated pixel of the elliptical peripheral channel read in step S5 constitutes an ellipse, if so, step S7 is executed, and if not, step S8 is executed;
step S7, the result output unit (4) prompts the successful recognition and outputs the ellipse parameters;
step S8, the result output unit (4) prompts the recognition failure;
in step S3, the operation procedure of the elliptical trajectory integral operation unit (2) includes:
step S30, the coordinates of the center point of the ellipse are set to (0, 0), and the ellipse is expressed by the following formula:
wherein x and y are Cartesian coordinates of any point P of the ellipse circumference, a is the ellipse major semi-axis, and b is the ellipse minor semi-axis;
step S31, after parameterizing the formula, introducing an angle t that increases from the center of the ellipse to obtain:
x＝a cos(t)；
y＝b sin(t)；
in step S31, determining the number of growth steps of the value of the angle t according to the ellipse perimeter 2 π b +4 (a-b);
in step S6, the ellipse recognition unit (3) recognizes an ellipse based on the following operator:
wherein:
c (x, y) is a binary edge image domain generated after Canny edge detection is carried out on an image to be detected;
2. The elliptical image detection algorithm based on the circle integral of claim 1, characterized in that in step S2, the preprocessing unit (1) preprocesses the image to be detected by using a preset Canny edge detection operator.
3. The elliptical image detection algorithm based on the circular tract integration as claimed in claim 1, wherein a threshold is preset in the elliptical recognition unit (3), and in step S6, the determination process of the elliptical recognition unit (3) comprises:
step S50, the ellipse recognizing unit (3) determines whether the density value of the ellipse contour integral pixel read in step S5 is higher than the threshold, if yes, step S51 is performed, and if no, step S52 is performed;
step S51, the ellipse identifying unit (3) takes the smallest ellipse parameter in the pixel density values of the group of ellipse surrounding channels as the target ellipse which is successfully identified;
in step S52, the result output unit (4) prompts that the recognition is failed.
4. The surround-integral-based elliptical image detection algorithm according to claim 3, characterized in that the ellipse identifying unit (3) uses 1/3 of one complete elliptical surround integral value as the threshold value.
5. The elliptical image detection algorithm based on the circular road integral as claimed in claim 4, characterized in that when the image to be detected is a perfect circle, the initial values of the parameters a and b are set as a-b, and the perfect circle is identified in combination with the steps S2 to S8.
6. The elliptical image detection algorithm based on the circle integrals as claimed in claim 4, characterized in that when the image to be detected is an arc, it is regarded as a partial circle integral operation, and after adjusting the threshold, the arc is identified in combination with steps S2 to S8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010576856.9A CN111739006B (en) | 2020-06-22 | 2020-06-22 | Elliptical image detection algorithm and system based on enclosed road integral |
| PCT/CN2020/106036 WO2021258498A1 (en) | 2020-06-22 | 2020-07-31 | Ellipse image detection algorithm and system based on contour integration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010576856.9A CN111739006B (en) | 2020-06-22 | 2020-06-22 | Elliptical image detection algorithm and system based on enclosed road integral |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111739006A CN111739006A (en) | 2020-10-02 |
| CN111739006B true CN111739006B (en) | 2021-07-13 |
Family
ID=72650473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010576856.9A Active CN111739006B (en) | 2020-06-22 | 2020-06-22 | Elliptical image detection algorithm and system based on enclosed road integral |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111739006B (en) |
| WO (1) | WO2021258498A1 (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113313724B (en) * | 2021-05-27 | 2022-04-08 | 深圳企业云科技股份有限公司 | Line detection processing method for resisting resampling of mobile phone camera |
| CN113793309B (en) * | 2021-08-27 | 2024-04-09 | 西北工业大学 | Subpixel level ellipse detection method based on morphological characteristics |
| CN115115607B (en) * | 2022-07-19 | 2024-06-07 | 重庆大学 | Image shape feature extraction and recognition method based on image analysis |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923645A (en) * | 2009-06-09 | 2010-12-22 | 黑龙江大学 | Iris splitting method suitable for low-quality iris image in complex application context |
| CN102034101A (en) * | 2010-10-22 | 2011-04-27 | 广东工业大学 | Method for quickly positioning circular mark in PCB visual detection |
| CN103164854A (en) * | 2013-02-25 | 2013-06-19 | 东南大学 | Processing method of atherosclerotic plaque medical image |
| CN103903284A (en) * | 2014-03-24 | 2014-07-02 | 中山新诺科技股份有限公司 | Multi-circle detecting method based on least square method |
| CN104915973A (en) * | 2015-06-30 | 2015-09-16 | 南京大学 | Method for solving center of regular circle in image |
| CN110705569A (en) * | 2019-09-18 | 2020-01-17 | 西安交通大学 | Image local feature descriptor extraction method based on texture features |
| CN110942092A (en) * | 2019-11-21 | 2020-03-31 | 大庆思特传媒科技有限公司 | Graphic image recognition method and recognition system |
| CN111311593A (en) * | 2020-03-14 | 2020-06-19 | 深圳市维森软件股份有限公司 | Multi-ellipse detection and evaluation algorithm, device, terminal and readable storage medium based on image gradient information |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005503619A (en) * | 2001-09-14 | 2005-02-03 | エステーミクロエレクトロニクス ソシエテ アノニム | How to identify the ring position |
| US7583823B2 (en) * | 2006-01-11 | 2009-09-01 | Mitsubishi Electric Research Laboratories, Inc. | Method for localizing irises in images using gradients and textures |
| US7970179B2 (en) * | 2006-09-25 | 2011-06-28 | Identix Incorporated | Iris data extraction |
| US8161775B2 (en) * | 2009-07-16 | 2012-04-24 | Stoneferry Technology, LLC | Integrated hollow fabric structure |
| CN102651145B (en) * | 2012-04-06 | 2014-11-05 | 哈尔滨工业大学 | Three-dimensional femoral model visualization method |
| CN109405979B (en) * | 2018-09-14 | 2019-11-15 | 中国科学院西安光学精密机械研究所 | A kind of Michelson's interferometer image width of fringe detection method and system |
-
2020
- 2020-06-22 CN CN202010576856.9A patent/CN111739006B/en active Active
- 2020-07-31 WO PCT/CN2020/106036 patent/WO2021258498A1/en active Application Filing
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923645A (en) * | 2009-06-09 | 2010-12-22 | 黑龙江大学 | Iris splitting method suitable for low-quality iris image in complex application context |
| CN102034101A (en) * | 2010-10-22 | 2011-04-27 | 广东工业大学 | Method for quickly positioning circular mark in PCB visual detection |
| CN103164854A (en) * | 2013-02-25 | 2013-06-19 | 东南大学 | Processing method of atherosclerotic plaque medical image |
| CN103903284A (en) * | 2014-03-24 | 2014-07-02 | 中山新诺科技股份有限公司 | Multi-circle detecting method based on least square method |
| CN104915973A (en) * | 2015-06-30 | 2015-09-16 | 南京大学 | Method for solving center of regular circle in image |
| CN110705569A (en) * | 2019-09-18 | 2020-01-17 | 西安交通大学 | Image local feature descriptor extraction method based on texture features |
| CN110942092A (en) * | 2019-11-21 | 2020-03-31 | 大庆思特传媒科技有限公司 | Graphic image recognition method and recognition system |
| CN111311593A (en) * | 2020-03-14 | 2020-06-19 | 深圳市维森软件股份有限公司 | Multi-ellipse detection and evaluation algorithm, device, terminal and readable storage medium based on image gradient information |
Non-Patent Citations (4)
| Title |
|---|
| Edge curvature and convexity based ellipse detection method;Dilip K.Prasad 等;《Pattern Recognition》;20120930;第45卷(第9期);第3204-3221页 * |
| 一种基于分数阶次微积分模板的新型边缘检测方法;何春 等;《自动化学报》;20120531;第38卷(第5期);第776-787页 * |
| 基于Hough变换的椭圆检测算法;赵京东;《光电子技术》;20080930;第28卷(第3期);第161-164页 * |
| 基于边界跟踪及Hough变换的椭圆检测算法;韦宏强 等;《长春理工大学学报（自然科学版）》;20100930;第33卷(第3期);第133-136页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111739006A (en) | 2020-10-02 |
| WO2021258498A1 (en) | 2021-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111739006B (en) | Elliptical image detection algorithm and system based on enclosed road integral | |
| WO2019169532A1 (en) | License plate recognition method and cloud system | |
| US10410354B1 (en) | Method and apparatus for multi-model primitive fitting based on deep geometric boundary and instance aware segmentation | |
| Shan et al. | Real-time 3D single object tracking with transformer | |
| CN102646193A (en) | Segmentation method of character images distributed in ring shape | |
| CN112200884B (en) | Lane line generation method and device | |
| CN111310737B (en) | Lane line detection method and device | |
| CN114743259A (en) | Pose estimation method, pose estimation system, terminal, storage medium and application | |
| CN113393524B (en) | Target pose estimation method combining deep learning and contour point cloud reconstruction | |
| Chang et al. | Real-time vehicle tracking mechanism with license plate recognition from road images | |
| CN115526864A (en) | Steel rail surface defect detection method based on improved characteristic pyramid network and metric learning | |
| CN115810133A (en) | Welding control method based on image processing and point cloud processing and related equipment | |
| KR20140093407A (en) | Recognition device, vehicle model recognition apparatus and method | |
| CN117495891B (en) | Point cloud edge detection method and device and electronic equipment | |
| Chen et al. | Eyes localization algorithm based on prior MTCNN face detection | |
| Guo et al. | Machine vision-based intelligent manufacturing using a novel dual-template matching: a case study for lithium battery positioning | |
| CN113989308A (en) | Polygonal target segmentation method based on Hough transform and template matching | |
| Jiang et al. | Mobile robot gas source localization via top-down visual attention mechanism and shape analysis | |
| EP4375924A1 (en) | Cell alignment degree measurement method, controller, detection system, and storage medium | |
| CN116824333A (en) | Nasopharyngeal carcinoma detecting system based on deep learning model | |
| CN112183485B (en) | Deep learning-based traffic cone detection positioning method, system and storage medium | |
| Zhu et al. | Visual campus road detection for an ugv using fast scene segmentation and rapid vanishing point estimation | |
| CN113920155A (en) | Moving target tracking algorithm based on kernel correlation filtering | |
| CN109492685B (en) | Target object visual detection method for symmetric characteristics | |
| CN113591863A (en) | License plate character recognition method with self-correction consciousness |
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 |