CN105701150A - Intuitionistic fuzzy similarity degree based image retrieving method and system - Google Patents

Abstract

The present invention discloses an intuitionistic fuzzy similarity degree based image retrieving method and system. The method comprises: step one, performing intuitionistic fuzzification on images in an image library, so as to obtain an intuitionistic fuzzy set matrix model of each image; step two, according to the intuitionistic fuzzy set matrix model, calculating and obtaining an intuitionistic fuzzy similarity degree of each image in the image library and a retrieved image; and step three, outputting obtained retrieval images sequentially from large to small according to the intuitionistic fuzzy similarity degree, so as to obtain a final retrieval result. According to the method disclosed by the present invention, a to-be-retrieved image can be effectively retrieved, and the method is simple, convenient and quick, and has a better retrieval effect on images with a simple background.

Description

Image retrieval method and system based on intuitionistic fuzzy similarity

Technical Field

The invention relates to an image content retrieval technology, in particular to an image retrieval method and system based on intuitionistic fuzzy similarity.

Background

With the rapid development of the world science and technology, images are widely used with the popularization of multimedia devices, a large number of images are stored in an image database, the image database is larger, and the images contain richer information such as colors, shapes, textures, positions, environments and the like compared with common texts. Meanwhile, the information organization of the images has disorder, and how to realize effective and unified management and query on the images is always a research hotspot in the field of image processing. The image retrieval technology is to realize fast and effective retrieval of required information in a large amount of picture data.

Image retrieval is generally divided into two methods: text-based image retrieval (TBIR) and Content-based image retrieval (CBIR). The TBIR is searched based on the matching degree of comparison keywords and picture labels, the method is easy to understand and simple to implement, but the method is too dependent on subjective perception of people on pictures and has large annotation workload.

Unlike the exact matching of TBIR, CBIR employs similarity matching and combines the technical achievements of multiple fields of computer vision, image processing, image understanding, and databases to avoid the subjectivity of manual description, it allows a user to input a picture to find a picture of similar content. The CBIR mainly utilizes color, texture, shape, space position and their combination features contained in the image to establish a feature index library, and similarity index is carried out through approximate measurement between image feature vector features. Commonly used CBIR is generally classified into a search method based on global features and a search method based on image local information, where the search method based on image local information performs region segmentation on an image, and selects a region of interest for feature extraction. The region segmentation tries to express the content of an image through an object hierarchy, and ideal image segmentation can segment the image into semantic regions so as to extract high-level semantic features.

The characteristics of image bottom layers such as texture, color, shape, semantics and the like are main contents searched by the CBIR, how to effectively extract and utilize the characteristics is a research difficulty of the CBIR, and based on a fuzzy set theory, the bottom layer information generally carries certain fuzziness, so that the image searching by using the fuzzy set theory is reasonable, the intuition fuzzy set theory is a new field of artificial intelligence science and is an optimization of a fuzzy set, at present, the image searching algorithm based on the intuition fuzzy set theory is less researched, most of the image searching algorithms are focused on measuring the distance between images by using the intuition fuzzy set theory, and the algorithm is complicated.

At present, an image retrieval algorithm based on an intuitionistic fuzzy set theory faces the following problems:

(1) the research results are few, and the excellent performance of the intuitionistic fuzzy set is not fully utilized in the field of image retrieval.

(2) Most algorithms prefer to use a distance measure between images, which is tedious.

Disclosure of Invention

The invention aims to provide an image retrieval method and an image retrieval system based on intuitionistic fuzzy similarity, which are used for effectively, simply, conveniently and quickly retrieving an image to be retrieved.

In order to achieve the above object, the present invention provides an image retrieval method based on intuitive fuzzy similarity, comprising:

firstly, carrying out intuitive fuzzification on images in an image library to obtain an intuitive fuzzy set matrix model of the images;

step two, calculating the intuitive fuzzy similarity between each image in the image library and the image to be searched according to the intuitive fuzzy set matrix model;

and step three, outputting the retrieval images in sequence from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result.

The image retrieval method described above, wherein the first step includes: and preprocessing the images in the image library before the intuitive fuzzification is carried out.

The image retrieval method described above, wherein the first step includes:

step 1.1: normalizing the pixel value in each gray level image matrix in the image library, and calculating the membership degree matrix of the image in the following way:

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(1\right)$

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(2\right)$

wherein, i is 1,2, … M, j is 1,2, … N, mu_ij(x_ij) Is a pixel point x_ijDegree of membership of f_maxAnd f_minRespectively, the maximum value and the minimum value in the normalized image matrix, and M, N respectively, the row value and the column value of the image matrix;

step 1.2: after obtaining the membership matrix of the image, calculating the non-membership matrix of the image in the following way:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(3\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijλ > 0.

The image retrieval method described above, wherein in the second step, the method includes:

and solving the intuitionistic fuzzy similarity between each image in the image library and the searched image in the following way:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(4\right)$

wherein, ω is_j＝(1/n，1/n，·..，1/n)^TIs a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iOf the jth coordinate point of (a), theta (a)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

The image retrieval method, wherein in the third step, the method comprises:

and selecting the image with the maximum intuitive fuzzy similarity with the retrieval image as the searched image.

In order to achieve the above object, the present invention provides an image retrieval system based on intuitive fuzzy similarity, comprising:

the blurring processing module is used for performing intuitive blurring on the images in the image library to obtain an intuitive fuzzy set matrix model of the images;

the similarity acquisition module is used for calculating the intuitive fuzzy similarity between each image in the image library and the image to be searched according to the intuitive fuzzy set matrix model;

and the retrieval result acquisition module is used for sequentially outputting the retrieval images from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result.

The image retrieval system described above, further comprising: and the preprocessing module is used for preprocessing the images in the image library before the intuitive fuzzification is carried out.

The image retrieval system, wherein the blurring processing module further comprises:

the membership matrix module is used for normalizing the pixel values in each gray level image matrix in the image library and calculating the membership matrix of the image in the following way;

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(1\right)$

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(2\right)$

wherein, i is 1,2, … M, j is 1,2, … N, mu_ij(x_ij) Is a pixel point x_ijDegree of membership of f_maxAnd f_minRespectively, the maximum value and the minimum value in the normalized image matrix, and M, N respectively, the row value and the column value of the image matrix;

the non-membership matrix module is used for calculating the non-membership matrix of the image in the following mode after obtaining the membership matrix of the image:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(3\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijλ > 0.

The image retrieval system, wherein the similarity obtaining module finds out the intuitionistic fuzzy similarity between each image in the image library and the retrieved image in the following way:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(4\right)$

wherein, ω is_j＝(1/n，1/n，，1/n)^TIs a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iOf the jth coordinate point of (a), theta (a)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

The image retrieval system is characterized in that the retrieval result acquisition module selects an image with the largest intuitive fuzzy similarity with the retrieved image as the searched image.

Compared with the prior art, the invention has the beneficial technical effects that:

in order to solve the technical problems in the prior art, the invention provides an image retrieval method based on an intuitionistic fuzzy theory, which can effectively improve the retrieval precision and efficiency. Experiments show that: the algorithm provided by the invention can effectively retrieve the image to be retrieved, is simple, convenient and quick, and has better image retrieval effect on the image with simple background.

Drawings

FIG. 1 is a flowchart of an image retrieval method based on intuitive fuzzy similarity according to the present invention.

FIG. 2 is a diagram showing the effect of the example flower search according to the present invention.

Fig. 3 is a diagram of the effect of bus example retrieval according to the invention.

FIG. 4 is a diagram illustrating the searching effect of an example elephant according to the present invention.

FIG. 5 is a diagram of the image retrieval system based on the intuitive fuzzy similarity according to the present invention.

Detailed Description

The invention is described in detail below with reference to the drawings and specific examples, but the invention is not limited thereto.

Fig. 1 is a flowchart of an image retrieval method based on intuitive fuzzy similarity according to the present invention. The flow chart describes an image retrieval method based on an intuitionistic fuzzy theory, which is carried out according to the following steps:

step 1: and preprocessing the images in the image library.

Step 2: and performing intuitive fuzzification on the preprocessed images to obtain an intuitive fuzzy set matrix model of each image.

And step 3: and calculating the intuitive fuzzy similarity between each image in the image library and the searched image.

And 4, step 4: and sequentially outputting the obtained retrieval images from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result.

The step 1 is as follows:

step 1.1: an averaging filter is used to filter the images in the image library.

Step 1.2: and carrying out graying processing on the filtered image to obtain a grayscale image matrix.

The step 2 is as follows:

step 2.1: and solving a membership matrix of the image.

Normalizing the pixel values in each gray level image matrix, and then solving the fuzzy membership degree of the image by using a Gamma function, wherein the Gamma function is defined as follows:

$G\left(x\right)=\frac{{(x-m)}^{\mathrm{\γ}-1}\exp (-\frac{x-m}{\mathrm{\β}})}{{\mathrm{\β}}^{\mathrm{\γ}}\mathrm{\Γ}\left(\mathrm{\γ}\right)}---\left(1\right)$

wherein γ is a shape parameter, m is a position parameter, β is a scale parameter, and (1) ═ 1.

When m ≠ 0 and γ ═ 1, the Gamma function can be simplified as:

$G\left(x\right)=\frac{1}{\mathrm{\β}}\exp (-\frac{x-m}{\mathrm{\β}})---\left(2\right)$

normalizing the image, and then solving a membership matrix of the image on the basis:

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(3\right)$

wherein, i is 1,2, … M, j is 1,2, … N, f_maxAnd f_minRespectively, the maximum and minimum values in the normalized image matrix f (x). M, N are row and column values of the image matrix, respectively, and M, N varies with the number of rows and columns of the image matrix, and has no specific value range.

Further, a pixel point x can be obtained_ijThe membership matrix of (a) is:

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(4\right)$

wherein, mu_ij(x_ij) Is a pixel point x_ijThe membership degree of (b) is obtained based on Gamma function substitution, simplification and transformation.

Step 2.2: and solving a non-membership matrix of the image.

After obtaining the membership matrix of the image, using a Sugeno fuzzy complementary set to obtain a non-membership matrix of the image, wherein the Sugeno complementary set is defined as follows:

$N\left(g\right(x\left)\right)=\frac{1-g\left(x\right)}{1+\mathrm{\λ}g\left(x\right)}---\left(5\right)$

wherein N (1) ═ 0 and N (0) ═ 1.

The non-membership matrix of the image is calculated as follows:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(6\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijDegree of non-membership, mu_ij(x_ij) Is a pixel point x_ijThe membership degree of (b) is more than 0, and the value of (b) is 0.5.

The obtained membership matrix and the obtained non-membership matrix are an intuitive fuzzy set matrix model. The step 3 is as follows:

and solving the intuitionistic fuzzy similarity between each image in the image library and the searched image in the following way:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(7\right)$

wherein, ω is_j＝(1/n，1/n，，1/n)^TIs a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iOf the jth coordinate point of (a), theta (a)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

The step 4 is as follows:

and sequentially outputting the obtained retrieval images from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result, and selecting the image with the maximum intuitive fuzzy similarity with the retrieval image as the searched image.

The technical effect of the technical solution of the present invention will be further explained by a specific embodiment.

To verify the performance of the inventive algorithm, the proposed intuitive fuzzy theory based search algorithm was experimented with a picture library with 1000 generic images. Three image retrieval examples are given below, and the retrieval images are respectively: flowers, buses, elephants. The search results are shown in fig. 2, 3, and 4.

The general effect evaluation of the image retrieval is recall ratio and precision ratio, which are defined as follows:

recall-number of retrieved related images/number of all related images

Precision ratio-number of retrieved related images/number of retrieved images

The method selects 10 types from 1000 images, randomly selects 10 images in each type to form an image set with 100 images, performs retrieval experiments on all images in each type of image, calculates the recall ratio and precision ratio of the retrieval, and finally takes an average value as the recall ratio and precision ratio of the type of image, wherein the obtained result is shown in table 1.

Table 1 evaluation of the results of the algorithm

Categories	Image of a person	Recall ratio of	Precision ratio
				1	Character	0.14	0.14
2	Beach sand	0.22	0.22
				3	Construction of buildings	0.2	0.2
4	Bus with a movable rail	0.21	0.21
				5	Dinosaur	0.79	0.79
6	Elephant	0.16	0.16
				7	Flower (A. B. A	0.35	0.35
8	Horse	0.34	0.34
				9	Snow mountain	0.21	0.21
10	Food product	0.16	0.16

The results of the embodiment can remarkably show that the algorithm provided by the invention obtains considerable retrieval results, and can effectively retrieve partial images of the same kind in addition to the images to be retrieved, and can show that the recall ratio and the precision ratio are higher for the images with simple backgrounds such as dinosaurs, and the retrieval algorithm provided by the invention is very effective for the images with simple backgrounds.

Fig. 5 is a structural diagram of the image retrieval system based on the intuitive fuzzy similarity according to the present invention. The system 500 corresponds to the method described in fig. 1, and in connection with fig. 1, the system 500 comprises the following modules:

and a preprocessing module 51, configured to preprocess the images in the image library.

And the blurring processing module 52 is configured to perform intuitive blurring on the preprocessed image to obtain an intuitive fuzzy set matrix model of each image.

And a similarity obtaining module 53, configured to obtain an intuitive fuzzy similarity between each image in the image library and the retrieved image. And

and the retrieval result obtaining module 54 is configured to sequentially output the obtained retrieval images according to the similarity from large to small, so as to obtain a final retrieval result.

The preprocessing module 51 further includes:

and a filtering module 511, configured to filter the images in the image library by using an average filter.

And the graying module 512 is configured to perform graying processing on the filtered image to obtain a grayscale image matrix.

The blurring processing module 52 further includes:

and the membership matrix module 521 is used for solving a membership matrix of the image.

Normalizing the pixel values in each gray level image matrix, and then solving the fuzzy membership degree of the image by using a Gamma function, wherein the Gamma function is defined as follows:

$G\left(x\right)=\frac{{(x-m)}^{\mathrm{\γ}-1}\exp (-\frac{x-m}{\mathrm{\β}})}{{\mathrm{\β}}^{\mathrm{\γ}}\mathrm{\Γ}\left(\mathrm{\γ}\right)}---\left(1\right)$

wherein γ is a shape parameter, m is a position parameter, β is a scale parameter, and (1) ═ 1.

When m ≠ 0 and γ ═ 1, the Gamma function can be simplified as:

$G\left(x\right)=\frac{1}{\mathrm{\β}}\exp (-\frac{x-m}{\mathrm{\β}})---\left(2\right)$

normalizing the image, and then solving a membership matrix of the image on the basis:

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(3\right)$

wherein, i is 1,2, … M, j is 1,2, … N, f_maxAnd f_minRespectively, the maximum and minimum values in the normalized image matrix f (x). M, N are row and column values of the image matrix, respectively, and M, N varies with the number of rows and columns of the image matrix, and has no specific value range.

Further, a pixel point x can be obtained_ijThe membership degree is as follows:

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(4\right)$

wherein, mu_ij(x_ij) Is a pixelPoint x_ijThe membership degree of (b) is obtained based on Gamma function substitution, simplification and transformation.

And a non-membership matrix module 522 for solving a non-membership matrix of the image.

After obtaining the membership matrix of the image, using a Sugeno fuzzy complementary set to obtain a non-membership matrix of the image, wherein the Sugeno complementary set is defined as follows:

$N\left(g\right(x\left)\right)=\frac{1-g\left(x\right)}{1+\mathrm{\λ}g\left(x\right)}---\left(5\right)$

wherein N (1) ═ 0 and N (0) ═ 1.

The non-membership matrix of the image is calculated as follows:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(6\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijDegree of non-membership, mu_ij(x_ij) Is a pixel point x_ijThe membership degree of (b) is more than 0, and the value of (b) is 0.5.

The obtained membership matrix and the obtained non-membership matrix are an intuitive fuzzy set matrix model.

The similarity obtaining module 53 obtains the intuitive fuzzy similarity between each image in the image library and the image to be retrieved by the following method:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(7\right)$

wherein,ω_j＝(1/n，l/n，，1/n)^Tis a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iThe value at the jth coordinate point, θ (A)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

The retrieval result obtaining module 54 obtains the final retrieval result by the following method:

and sequentially outputting the obtained retrieval images from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result, and selecting the image with the maximum intuitive fuzzy similarity with the retrieval image as the searched image.

The invention provides an image retrieval method based on intuitionistic fuzzy similarity, which constructs an intuitionistic fuzzy set matrix model of an image through a Gamma function and a Sugeno fuzzy complement, measures the intuitionistic fuzzy similarity of the image and uses the intuitionistic fuzzy similarity to carry out image retrieval. Experiments show that the algorithm provided by the invention can effectively retrieve the image to be retrieved, is simple, convenient and quick, and has better image retrieval effect on simple backgrounds.

The present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof, and it should be understood that various changes and modifications can be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (10)

1. An image retrieval method based on intuitionistic fuzzy similarity is characterized by comprising the following steps:

firstly, carrying out intuitive fuzzification on images in an image library to obtain an intuitive fuzzy set matrix model of the images;

step two, calculating the intuitive fuzzy similarity between each image in the image library and the image to be searched according to the intuitive fuzzy set matrix model;

and step three, outputting the retrieval images in sequence from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result.

2. The image retrieval method according to claim 1, wherein the first step comprises: and preprocessing the images in the image library before the intuitive fuzzification is carried out.

3. The image retrieval method according to claim 1 or 2, wherein the first step comprises:

step 1.1: normalizing the pixel value in each gray level image matrix in the image library, and calculating the membership degree matrix of the image in the following way:

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(1\right)$

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(2\right)$

wherein, i is 1,2, … M, j is 1,2,…N，μ_ij(x_ij) Is a pixel point x_ijDegree of membership of f_maxAnd f_minRespectively, the maximum value and the minimum value in the normalized image matrix, and M, N respectively, the row value and the column value of the image matrix;

step 1.2: after obtaining the membership matrix of the image, calculating the non-membership matrix of the image in the following way:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(3\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijλ > 0.

4. The image retrieval method according to claim 3, wherein the second step includes:

and solving the intuitionistic fuzzy similarity between each image in the image library and the searched image in the following way:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(4\right)$

wherein, ω is_j＝(1/n，1/n，···，1/n)^TIs a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iOf the jth coordinate point of (a), theta (a)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

5. The image retrieval method according to claim 1,2 or 4, wherein in the third step, the method comprises:

and selecting the image with the maximum intuitive fuzzy similarity with the retrieval image as the searched image.

6. An image retrieval system based on intuitive fuzzy similarity, comprising:

the blurring processing module is used for performing intuitive blurring on the images in the image library to obtain an intuitive fuzzy set matrix model of the images;

the similarity acquisition module is used for calculating the intuitive fuzzy similarity between each image in the image library and the image to be searched according to the intuitive fuzzy set matrix model;

and the retrieval result acquisition module is used for sequentially outputting the retrieval images from large to small according to the intuitive fuzzy similarity to obtain a final retrieval result.

7. The image retrieval system of claim 6, further comprising: and the preprocessing module is used for preprocessing the images in the image library before the intuitive fuzzification is carried out.

8. The image retrieval system of claim 6 or 7, wherein the blurring processing module further comprises:

the membership matrix module is used for normalizing the pixel values in each gray level image matrix in the image library and calculating the membership matrix of the image in the following way;

$\mathrm{\μ}=\underset{i=1}{\overset{M}{\Σ}}\underset{j=1}{\overset{N}{\Σ}}{x}_{ij}/(M\×N),\mathrm{\β}=1/(f_{max}-f_{min})---\left(1\right)$

${\mathrm{\μ}}_{ij}\left(x_{ij}\right)=(f_{max}-f_{min})\·\exp (-(f_{max}-f_{min})\·|x_{ij}-\mathrm{\μ}\left|\right)=\frac{1}{\mathrm{\β}}\·\exp (-\frac{x_{ij}-\mathrm{\μ}}{\mathrm{\β}})---\left(2\right)$

wherein, i is 1,2, … M, j is 1,2, … N, mu_ij(x_ij) Is a pixel point x_ijDegree of membership of f_maxAnd f_minRespectively, the maximum value and the minimum value in the normalized image matrix, and M, N respectively, the row value and the column value of the image matrix;

the non-membership matrix module is used for calculating the non-membership matrix of the image in the following mode after obtaining the membership matrix of the image:

${\mathrm{\υ}}_{ij}\left(x_{ij}\right)=\frac{1-{\mathrm{\μ}}_{ij}\left(x_{ij}\right)}{1+{\mathrm{\λ\μ}}_{ij}\left(x_{ij}\right)}---\left(3\right)$

wherein upsilon is_ij(x_ij) Is a pixel point x_ijλ > 0.

9. The image retrieval system of claim 8, wherein the similarity acquisition module finds the intuitive blurred similarity of each image in the image library to the retrieved image as follows:

$\begin{array}{c}\mathrm{\θ}\left(A_1,A_2\right)\\ =1-\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\ω}}_j\[\frac{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|+\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|}{4}\\ +\frac{\max \left\{\left|{\mathrm{\μ}}_{A_1}\left(x_j\right)-{\mathrm{\μ}}_{A_2}\left(x_j\right)\right|,\left|{\mathrm{\υ}}_{A_1}\left(x_j\right)-{\mathrm{\υ}}_{A_2}\left(x_j\right)\right|\right\}}{2}\]\end{array}---\left(4\right)$

wherein, ω is_j＝(1/n，1/n，···，1/n)^TIs a weight value, n is the total number of pixel points of the image to be searched, A₁And A₂Respectively representing an intuitive fuzzy set matrix model of the image to be searched and an intuitive fuzzy set matrix model of any one image in the search image library,A_i(j) set matrix model A is blurred for intuition_iOf the jth coordinate point of (a), theta (a)₁,A₂) Is represented by A₁And A₂The intuitive fuzzy similarity of the corresponding original images.

10. The image retrieval system according to claim 6, 7 or 9, wherein the retrieval result acquisition module selects an image having a maximum intuitive blur similarity with the retrieved image as the retrieved image.