Image edge detection algorithm based on fuzzy set

By using inverse cosine value and fuzzy set algorithms in image edge detection, redundant features are eliminated and important features are highlighted, and the problem of unclear edge detection in the prior art is solved, achieving a more complete and clear edge extraction effect.

CN118982550APending Publication Date: 2024-11-19BEIJING POLYTECHNIC
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411013124.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The prior art is prone to ignore or lose some important features in image edge detection, resulting in unclear or distorted edge detection.

Method used

By converting the pixel values ​​in the image to be detected into inverse cosine values ​​as the parameters of TDFP, redundant features are cleared, important features are highlighted, and edge extraction is performed using fuzzy sets and maximization and minimization operators to ensure the integrity of the image edges. .

Benefits of technology

Effectively eliminate redundant features in the image, preserve the tiny edge information of the image, and improve the clarity and integrity of edge detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118982550A_ABST
    Figure CN118982550A_ABST
Patent Text Reader

Abstract

The invention provides an image edge detection algorithm based on a fuzzy set, and the algorithm comprises the steps: taking a pixel point on a to-be-detected image as a feature point, carrying out the normalization processing, and obtaining a normalized feature value; performing anti-cosine trigonometric function calculation on the normalized characteristic value to obtain an anti-cosine value; solving a triangular difference value feature of the plurality of arc cosine values, and obtaining an intermediate feature map; and constructing a fuzzy set according to the intermediate feature map, and extracting a target edge in the image according to the fuzzy set to obtain the target edge. According to the method, the pixel values corresponding to the pixel points in the to-be-detected image are converted, the anti-cosine values of the pixel values are extracted to serve as parameters of the triangular difference value features, redundant features in the to-be-detected image are eliminated through the triangular difference value features, important features in the to-be-detected image are emphasized, and the effect of improving the edge detection effect is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of image processing, and in particular relates to an image edge detection algorithm based on fuzzy sets. Background Art

[0002] Edge detection is a basic technology in the fields of image processing and target recognition. In the existing technology, fuzzy sets are constructed through fuzzification algorithms, different membership functions are adopted for different targets on the fuzzy feature plane, and edge extraction algorithms are used to detect image edges. However, it is easy to ignore and blur some important features in the original image during processing, resulting in unclear edge detection or loss of some edge lines. Summary of the invention

[0003] In view of the above technical problems, the present invention converts the pixel values ​​corresponding to the pixel points in the image to be detected, extracts their arc cosine values ​​as the parameters of TDFP, removes the redundant features in the image to be detected through TDFP, and highlights the important features in the image to be detected, thereby achieving the purpose of improving the edge detection effect.

[0004] The first aspect of the present invention provides: an image edge detection algorithm based on fuzzy sets,

[0005] The pixel points on the image to be detected are taken as feature points, normalized, and normalized feature values ​​are obtained;

[0006] Perform arccosine trigonometric function calculation on the normalized eigenvalue to obtain an arccosine value;

[0007] Calculate trigonometric difference features from the arc cosine values ​​and obtain an intermediate feature map;

[0008] A fuzzy set is constructed according to the intermediate feature map, and the edge of the target in the image is extracted according to the fuzzy set to obtain the edge of the target.

[0009] The second aspect of the present invention provides: a fuzzy set-based image edge detection system, the system comprising at least one processor; and a memory storing instructions, which, when executed by at least one processor, implement the steps of the method described in the first aspect of the present invention.

[0010] The third aspect of the present invention provides: an image edge detection algorithm based on fuzzy sets and a system computer-readable storage medium, on which a computer program / instruction is stored, and when the computer program / instruction is executed by a processor, the steps of the method described in the first aspect of the present invention are implemented.

[0011] The fourth aspect of the present invention provides: an image edge detection algorithm based on fuzzy sets and a system computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the method described in the first aspect of the present invention.

[0012] The beneficial effects of the present invention are as follows: the edge features of the image are extracted by highlighting the triangular difference features, fuzzy sets, and maximization operators and minimization operators, the important features of the image to be detected are highlighted by TDFP, the image is preprocessed before edge detection, and the redundant features in the image are effectively eliminated; the image to be detected is converted into different membership functions by using fuzzy sets, the grayscale information of different targets in the image is preserved, and for a certain target, the fuzzy enhancement operator is used to perform attenuation operations on most pixels in the low grayscale area and enhancement operations on most pixels in the high grayscale area, so that the edge information of different grayscale levels of each target in the image is preserved and depicted, so that the edge of the acquired image is comprehensive and complete, and a large number of fine edges of the image are retained. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Flow chart of an embodiment of the present invention;

[0014] Figure 2 Normalized flow chart of an embodiment of the present invention;

[0015] Figure 3 Flowchart of obtaining arccosine value according to an embodiment of the present invention;

[0016] Figure 4 Flow chart of fuzzy set edge extraction according to an embodiment of the present invention;

[0017] Figure 5 A schematic diagram of the structure of a detection system according to an embodiment of the present invention;

[0018] Figure 6 A schematic diagram of the connection of processor modules of a detection system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following examples further illustrate the content of the present invention, but should not be construed as limiting the present invention. Without departing from the spirit and substance of the present invention, modifications or substitutions made to the methods, steps or conditions of the present invention all fall within the scope of the present invention.

[0020] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0021] In the present invention, the pixel values ​​corresponding to the pixel points in the image to be detected are converted, and the arc cosine values ​​are extracted as the parameters of TDFP (trigonometric difference feature), and the redundant features in the image to be detected are removed by TDFP, so as to highlight the important features in the image to be detected;

[0022] A fuzzy set is established for the image preprocessed by TDFP, and after fuzzy enhancement, the “min” and “max” operators are used to detect its edge.

[0023] "Redundant features" specifically refer to those features in the image to be detected that are not important for subsequent analysis or recognition tasks and may interfere with the recognition process. These features include but are not limited to the following aspects:

[0024] 1. Noise: Random, meaningless changes in pixel values ​​in an image, which may be caused by various interferences during the image capture process.

[0025] 2. Background information: If the object to be detected is a specific object in the image, then the background information may be regarded as redundant features because they do not provide useful information for identifying the object.

[0026] 3. Repeating textures or patterns: Images may contain recurring textures or patterns, which may not be important for identifying a specific object and can therefore also be considered as redundant features.

[0027] 4. Lighting and shadows: Uneven lighting and shadows may cause changes in pixel values ​​in the image, but these changes may not reflect the essential characteristics of the object.

[0028] 5. Irrelevant details: For some recognition tasks, some details in the image may not be important and may even interfere with the recognition process.

[0029] The purpose of TDFP (triangular difference feature) preprocessing is to highlight the important features in the image and reduce or eliminate the influence of these redundant features to improve the accuracy of subsequent image analysis and recognition.

[0030] In some embodiments of the present invention, Figure 1 As shown: An image edge detection algorithm based on fuzzy sets, including:

[0031] The pixel points on the image to be detected are taken as feature points, normalized, and normalized feature values ​​are obtained;

[0032] Perform arccosine trigonometric function calculation on the normalized eigenvalue to obtain an arccosine value;

[0033] Calculate trigonometric difference features from the arc cosine values ​​and obtain an intermediate feature map;

[0034] A fuzzy set is constructed according to the intermediate feature map, and the edge of the target in the image is extracted according to the fuzzy set to obtain the edge of the target.

[0035] In these embodiments:

[0036] Normalization is used for image preprocessing, such as normalizing the pixel values ​​of an image to the range of [0, 1] or [-1, 1] to facilitate subsequent processing and analysis, especially in algorithm implementation, to avoid computational problems caused by large differences in data magnitude.

[0037] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that: performing normalization processing, the step of obtaining the normalized eigenvalue includes:

[0038] Perform maximum normalization on the image to obtain the maximum normalized eigenvalue;

[0039] The image is normalized to the minimum value to obtain the minimum normalized eigenvalue.

[0040] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that: the normalized eigenvalue is subjected to arccosine trigonometric function calculation to obtain the arccosine value, which includes:

[0041] Perform arccosine trigonometric function calculation on the maximum normalized eigenvalue to obtain the maximum arccosine value;

[0042] The minimum normalized eigenvalue is subjected to arccosine trigonometric function calculation to obtain a minimum arccosine value.

[0043] In some embodiments of the present invention, Figure 1As shown, the difference from the above embodiment is that the image is normalized to the maximum value, and the maximum normalized eigenvalue is obtained according to the following formula (I):

[0044]

[0045] In the formula, X represents the feature value set of all feature pixels on the image to be detected, x i represents the i-th feature pixel, x i ' represents the normalized eigenvalue, max represents the maximum value, and min represents the minimum value.

[0046] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that the image is normalized to the minimum value, and the minimum normalized eigenvalue is obtained according to the following formula (II):

[0047]

[0048] In the formula, X represents the feature value set of all feature pixels on the image to be detected, x i represents the i-th feature pixel, x i ' represents the normalized eigenvalue, max represents the maximum value, and min represents the minimum value.

[0049] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that the maximum normalized eigenvalue is subjected to arccosine trigonometric function calculation according to the following formula (III):

[0050] θ i (max) = arccos(x i ′(max))(III)

[0051] In the formula, x i ' represents the normalized eigenvalue, θ i Represents x i 'The corresponding inverse cosine trigonometric function value, arccos represents the inverse cosine trigonometric function, and max represents the maximum value.

[0052] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that the minimum normalized eigenvalue is calculated by the inverse cosine trigonometric function according to the following formula (IV):

[0053] θ i (min) = arccos(x i ′(min))(IV)

[0054] In the formula, x i' represents the normalized eigenvalue, θ i Represents x i 'The corresponding inverse cosine trigonometric function value, arccos represents the inverse cosine trigonometric function, and max represents the maximum value.

[0055] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that: the trigonometric difference feature is obtained according to the several arc cosine values ​​according to the following formula:

[0056] x i ”=cos[θ i (max)]sin[θ i (min)]-sin[θ i (max)]cos[θ i (min)](V)

[0057] Where: x i " represents the triangular difference feature of the i-th feature pixel, θ i (max) represents the maximum arc cosine value corresponding to the maximum normalized eigenvalue of the i-th feature pixel, θ i (min) represents the minimum arc cosine value corresponding to the minimum normalized eigenvalue of the i-th feature pixel.

[0058] In these embodiments:

[0059] Before constructing the fuzzy set of the feature image to be detected, the important features are first highlighted through TDFP. The specific implementation process is as follows:

[0060] The pixel point on the image to be detected is used as the feature point, and its value x i As the eigenvalue, the image is normalized to the maximum value and the minimum value respectively:

[0061]

[0062] Where X represents the feature value set of all feature pixels on the image to be detected;

[0063] Use the two normalized eigenvalues ​​as the parameter values ​​of the inverse cosine trigonometric function:

[0064] θ i (max) = arccos(x i ′(max))

[0065] θ i (min) = arccos(x i ′(min))

[0066] Substitute the above arccosine value into the formula to obtain its trigonometric difference characteristics:

[0067] x i ”=cos[θ i (max)]sin[θ i (min)]-sin[θ i (max)]cos[θ i (min)]

[0068] The pixel values ​​corresponding to the pixel points in the image to be detected are converted, and their arc cosine values ​​are extracted as the parameters of TDFP. The redundant features in the image to be detected are removed through TDFP, and the important features in the image to be detected are highlighted.

[0069] In some embodiments of the present invention, Figure 1 As shown, the difference from the above embodiment is that: the step of constructing a fuzzy set on the intermediate feature map, extracting edges, and obtaining image edges includes:

[0070] Grayscale the intermediate feature map to obtain a grayscale image and a matrix expression of the grayscale image;

[0071] Fuzzifying the grayscale image to obtain a fuzzy set;

[0072] Use the maximization operator and the minimization operator as the basis for edge extraction, define the edge matrix of the image, and extract the image edge.

[0073] In these embodiments:

[0074] Construct fuzzy sets for intermediate feature maps:

[0075] The brightness of each pixel in the image corresponds to a value in the integer matrix, which forms the gray value matrix of the pixel. Each row and column of the integer matrix corresponds to the two coordinates of the sampling point when scanning the image, that is, the position of the image pixel.

[0076] If I is used to describe a grayscale image with a grayscale level of L and a size of M×N, its matrix expression is:

[0077]

[0078] The fuzzification of an image is to convert the image grayscale matrix into a fuzzy set to form a fuzzy feature matrix, and convert the grayscale value of each pixel of the image into a membership degree relative to a specific grayscale level. In this way, a fuzzy value can represent the brightness of a part of the image. In fuzzy set theory, fuzzy sets are mainly characterized by their membership functions.

[0079] The definition is as follows: If there is a set A on the domain U, then for any x∈U, there is a number UA ∈[0,1] corresponds to this, which is the degree of membership of x to the fuzzy set A. Let the fuzzy set of the image be A1. After the image is fuzzified, the fuzzy set A1 can be expressed as:

[0080] A1={U A1 (X mn ) m=0,1,...,M-1;n=0,1,...,N-1}

[0081] in,

[0082]

[0083] Then the matrix expression can be converted to:

[0084]

[0085] Among them, X T is the threshold, u mn The gray level X of the representative point (m, n) mn Membership function relative to a specific gray level.

[0086] In the present invention, the threshold value X is used T Define the membership function by region. When defining the elements of the fuzzy matrix, the pixel value X mn Less than a certain threshold X T The pixel is defined as relative to the threshold X T The membership degree is greater than the threshold X T The pixel relative to the maximum gray level X max Degree of membership.

[0087] Finally, the "min" and "max" operators proposed by Nakagama and Rosenfeld are used as the basis for edge extraction, and the edge of the image is defined as the following matrix:

[0088] Edges=|x mn -min(x i )|

[0089] Implement edge detection on images.

[0090] Fuzzy sets are used to perform different membership function transformations on the image to be detected, which preserves the grayscale information of different targets in the image. For a certain target, the fuzzy enhancement operator is used to attenuate most pixels in the low grayscale area and enhance most pixels in the high grayscale area. This preserves and depicts the edge information of different grayscale levels of each target in the image, so that the edge of the acquired image is comprehensive and complete, and a large number of fine edges of the image are retained.

[0091] In some embodiments of the present invention, Figure 1As shown, the difference from the above embodiment is: a fuzzy set-based image edge detection system, the system includes at least one processor; and a memory, which stores instructions, and when the instructions are executed by at least one processor, the steps of the method described in the first aspect of the present invention are implemented.

[0092] The present invention extracts image edge features through triangular difference feature highlighting, fuzzy sets, and maximization operators and minimization operators ("min" and "max" operators). Compared with other image edge detection algorithms, the present invention highlights important features of the image to be detected through TDFP, pre-processes the image before edge detection, and effectively eliminates redundant features in the image; uses fuzzy sets to perform different membership function conversions on the image to be detected, preserves the grayscale information of different targets in the image, and for a certain target, uses a fuzzy enhancement operator to perform attenuation operations on most pixels in the low grayscale area and performs enhancement operations on most pixels in the high grayscale area, so as to preserve and characterize the edge information of different grayscale levels of each target in the image, so that the acquired image edge is comprehensive and complete, and a large number of fine edges of the image are retained.

[0093] The embodiments and functional operations of the subject matter described in this specification may be implemented in: digital electronic circuitry, tangibly implemented computer software or firmware, computer hardware, including the structures disclosed in this specification and their structural equivalents, or a combination of more than one of the above. The embodiments of the subject matter described in this specification may be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on one or more tangible non-transitory program carriers, for execution by a data processing device or to control the operation of a data processing device.

[0094] Alternatively or additionally, the program instructions may be encoded on an artificially generated propagation signal, such as a machine-generated electrical, optical or electromagnetic signal, which is generated as encoded information for transmission to an appropriate receiver device for execution by a data processing device. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or one of the above devices, an image edge detection algorithm and system based on fuzzy sets, or a combination of multiple thereof.

[0095] Although the present invention has been described in detail above by general description, specific implementation methods and experiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. An image edge detection algorithm based on fuzzy sets, It is characterized in that include: The pixel points on the image to be detected are taken as feature points, normalized, and normalized feature values ​​are obtained; Perform arccosine trigonometric function calculation on the normalized eigenvalue to obtain an arccosine value; Calculate trigonometric difference features from the arc cosine values ​​and obtain an intermediate feature map; A fuzzy set is constructed according to the intermediate feature map, and the edge of the target in the image is extracted according to the fuzzy set to obtain the edge of the target.

2. The image edge detection method based on fuzzy sets according to claim 1, It is characterized in that Perform normalization processing to obtain normalized eigenvalues. The steps include: Perform maximum normalization on the image to obtain the maximum normalized eigenvalue; The image is normalized to the minimum value to obtain the minimum normalized eigenvalue.

3. The image edge detection method based on fuzzy sets according to claim 2, It is characterized in that The step of performing arccosine trigonometric function calculation on the normalized eigenvalue to obtain the arccosine value comprises: Perform arccosine trigonometric function calculation on the maximum normalized eigenvalue to obtain the maximum arccosine value; The minimum normalized eigenvalue is subjected to arccosine trigonometric function calculation to obtain a minimum arccosine value.

4. The image edge detection method based on fuzzy sets according to claim 2, It is characterized in that The image is normalized to the maximum value, and the maximum normalized eigenvalue is obtained according to the following formula (I): In the formula, X represents the feature value set of all feature pixels on the image to be detected, x i represents the i-th feature pixel, x i ' represents the normalized eigenvalue, max represents the maximum value, and min represents the minimum value.

5. The image edge detection method based on fuzzy sets according to claim 1, It is characterized in that The image is normalized to the minimum value, and the minimum normalized eigenvalue is obtained according to the following formula (II): In the formula, X represents the feature value set of all feature pixels on the image to be detected, x i represents the i-th feature pixel, x i ' represents the normalized eigenvalue, max represents the maximum value, and min represents the minimum value.

6. The image edge detection method based on fuzzy sets according to claim 1, It is characterized in that The maximum normalized eigenvalue is subjected to arccosine trigonometric function calculation according to the following formula (III): θ i (max)=arccos(x i ′(max))(III) In the formula, x i ' represents the normalized eigenvalue, θ i Represents x i 'The corresponding inverse cosine trigonometric function value, arccos represents the inverse cosine trigonometric function, and max represents the maximum value.

7. The image edge detection method based on fuzzy sets according to claim 1, It is characterized in that The minimum normalized eigenvalue is subjected to arccosine trigonometric calculation according to the following formula (IV): i i (min)=arccos(x i ′(min))(IV) In the formula, x i ' represents the normalized eigenvalue, θ i Represents x i 'The corresponding inverse cosine trigonometric function value, arccos represents the inverse cosine trigonometric function, and max represents the maximum value.

8. The image edge detection method based on fuzzy sets according to claim 3, It is characterized in that The trigonometric difference feature is obtained according to the arc cosine values ​​in accordance with the following formula: x i ”=cos[θ i (max)]sin[θ i (min)]-sin[θ i (max)]cos[θ i (min)](V) Where: x i " represents the triangular difference feature of the i-th feature pixel, θ i (max) represents the maximum arc cosine value corresponding to the maximum normalized eigenvalue of the i-th feature pixel, θ i (min) represents the minimum arc cosine value corresponding to the minimum normalized eigenvalue of the i-th feature pixel.

9. The image edge detection method based on fuzzy sets according to claim 1, It is characterized in that A fuzzy set is constructed according to the intermediate feature map, and an edge of a target in an image is extracted according to the fuzzy set. The step of obtaining the edge of the target includes: Get the matrix expression of the grayscale image; Fuzzifying the grayscale image to obtain a fuzzy set; Use the maximization operator and the minimization operator as the basis for edge extraction, define the edge matrix of the image, and extract the target edge in the image.

10. An image edge detection system based on fuzzy sets, It is characterized in that The system comprises at least one processor; and a memory storing instructions, which, when executed by the at least one processor, implement the steps of the method according to any one of claims 1 to 9.

Citation Information

Patent Citations

  • Fuzzy edge detection method based on object cloud

    CN101286233A

  • Image enhancement and partition method

    CN103871029A

  • Image edge detection method based on fuzzy set

    CN106097339A