CN104834732A - Texture image retrieving method - Google Patents

Texture image retrieving method Download PDF

Info

Publication number
CN104834732A
CN104834732A CN201510250553.7A CN201510250553A CN104834732A CN 104834732 A CN104834732 A CN 104834732A CN 201510250553 A CN201510250553 A CN 201510250553A CN 104834732 A CN104834732 A CN 104834732A
Authority
CN
China
Prior art keywords
image
texture
channel
texture image
hue
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.)
Pending
Application number
CN201510250553.7A
Other languages
Chinese (zh)
Inventor
陈新武
王鹏
钟莉娟
马建忠
张丽
刘真
马文娟
薛静
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xinyang Normal University
Original Assignee
Xinyang Normal University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xinyang Normal University filed Critical Xinyang Normal University
Priority to CN201510250553.7A priority Critical patent/CN104834732A/en
Publication of CN104834732A publication Critical patent/CN104834732A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5862—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G06T7/41—Analysis of texture based on statistical description of texture

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Library & Information Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Image Analysis (AREA)
  • Processing Or Creating Images (AREA)

Abstract

本发明公开了一种纹理图像检索方法，获取目标纹理图像和纹理图像数据库中每幅图像对应的纹理结构模式图像，再进行特征提取得到结构特征序列，计算结构特征序列的相似度，组合第一纹理图像集并获取颜色特征，组合成第二图像集并对第二图像集中的图像进行排列显示。本发明方法通过先对纹理图像进行提取纹理特征序列，再进行图像集的设定、排列以及检索，降低了图像特征的维度，提高了图像处理速度，从而提高了图像的检索速度和准确率。

The invention discloses a texture image retrieval method, which comprises acquiring a target texture image and a texture structure pattern image corresponding to each image in a texture image database, performing feature extraction to obtain a structure feature sequence, calculating the similarity of the structure feature sequence, and combining the first The texture image set and color features are obtained, combined into a second image set, and the images in the second image set are arranged and displayed. The method of the invention first extracts the texture feature sequence from the texture image, and then sets, arranges and retrieves the image set, reduces the dimension of the image feature, improves the image processing speed, and thus improves the image retrieval speed and accuracy.

Description

一种纹理图像检索方法A Texture Image Retrieval Method

技术领域technical field

本发明涉及图像处理技术领域，具体是一种纹理图像检索方法。The invention relates to the technical field of image processing, in particular to a texture image retrieval method.

背景技术Background technique

随着计算机网络技术和成像技术的迅速发展，各种各样的纹理图像数据库在迅速增长，如何从这些纹理图像数据库中快速有效地获取所需图像，是纹理图像检索领域中一个重要的研究课题。With the rapid development of computer network technology and imaging technology, various texture image databases are growing rapidly. How to quickly and effectively obtain the required images from these texture image databases is an important research topic in the field of texture image retrieval. .

现有技术中，利用脉冲耦合神经网络模型对被检索纹理图像进行图像特征提取，将提取到的图像特征和纹理图像数据库中图像的图像特征进行相似性度量，获取同一纹理图像，实现纹理图像的检索。但是利用脉冲祸耦合神经网络模型进行纹理图像检索时，对光照的影响较为敏感，阻碍了脉冲耦合神经网络在纹理图像检索中的应用。而且现有的图像检索方法存在检索速度慢，准确率低的问题。In the prior art, the pulse-coupled neural network model is used to extract the image features of the retrieved texture image, and the similarity measurement is performed between the extracted image features and the image features of the image in the texture image database, and the same texture image is obtained to realize the texture image. search. But when the pulse-coupled neural network model is used for texture image retrieval, it is sensitive to the influence of illumination, which hinders the application of pulse-coupled neural network in texture image retrieval. Moreover, existing image retrieval methods have the problems of slow retrieval speed and low accuracy.

发明内容Contents of the invention

本发明的目的在于提供一种纹理图像检索方法，以解决上述背景技术中提出的问题。The purpose of the present invention is to provide a texture image retrieval method to solve the problems raised in the background art above.

为实现上述目的，本发明提供如下技术方案：To achieve the above object, the present invention provides the following technical solutions:

一种纹理图像检索方法，包括如下步骤：A texture image retrieval method, comprising the steps of:

(1)根据旋转不变局部二值模式算法对目标纹理图像和纹理图像数据库中的每幅图像进行处理，获取目标纹理图像和纹理图像数据库中每幅图像对应的纹理结构模式图像；(1) process each image in the target texture image and the texture image database according to the rotation invariant local binary pattern algorithm, and obtain the texture structure pattern image corresponding to each image in the target texture image and the texture image database;

(2)根据脉冲耦合神经网络模型对目标纹理图像和纹理数据库的每幅图像对应的纹理结构模式图像进行特征提取，得到目标纹理图像和纹理图像数据库中的每幅图像的纹理结构特征序列；(2) According to the pulse-coupled neural network model, the texture structure pattern image corresponding to each image of the target texture image and the texture database is extracted to obtain the texture structure feature sequence of each image in the target texture image and the texture image database;

(3)计算目标纹理图像的纹理结构特征序列和纹理图像数据库中的每幅图像的纹理结构特征序列进行相似度，将纹理图像数据库中的每幅图像的纹理结构特征序列相似度大于第一设定阈值的图像组合成第一纹理图像集；(3) Calculate the similarity between the texture structure feature sequence of the target texture image and the texture structure feature sequence of each image in the texture image database, and make the texture structure feature sequence similarity of each image in the texture image database greater than the first set The images with a fixed threshold are combined into a first texture image set;

(4)获取目标纹理图像和第一纹理图像集中每一纹理图像的颜色特征；(4) Obtain the color feature of each texture image in the target texture image and the first texture image set;

(5)计算目标纹理图像与第一纹理图像集中每一图像之间的颜色特征相似度，将颜色特征相似度大于第二设定阈值的图像组合成第二图像集；(5) Calculate the color feature similarity between the target texture image and each image in the first texture image set, and combine the images with the color feature similarity greater than the second preset threshold into a second image set;

(6)根据与目标纹理图像之间的颜色特征相似度由高到低的顺序，对第二图像集中的图像进行排列显示。(6) Arranging and displaying the images in the second image set according to the descending order of the color feature similarity with the target texture image.

作为本发明进一步方案：所述获取目标图像和第一图像集中每一图像的颜色特征具体包括：As a further solution of the present invention: the acquisition of the target image and the color features of each image in the first image set specifically include:

将目标纹理图像与第一纹理图像集中每一纹理图像转换为色调、饱和度和亮度图像格式，获取每一格式转换后的图像的色调通道、饱和度通道和亮度通道；converting each texture image in the target texture image and the first texture image set into a hue, saturation and brightness image format, and obtaining the hue channel, saturation channel and brightness channel of the converted image in each format;

对饱和度通道进行二值化处理，得到饱和度通道的亮区域和暗区域，将饱和度通道的亮区域对色调通道进行投影获得色调通道的色调区域，以及将饱和度通道的暗区域对亮度通道进行投影获得亮度通道中与饱和度通道暗区域对应的区域，并统计色调通道中的色调区域的灰度直方图以及所述亮度通道中与饱和度通道暗区域的对应区域灰度直方图；Binarize the saturation channel to obtain the bright and dark areas of the saturation channel, project the bright area of the saturation channel to the hue channel to obtain the hue area of the hue channel, and convert the dark area of the saturation channel to the brightness The channel is projected to obtain the region corresponding to the dark region of the saturation channel in the brightness channel, and the grayscale histogram of the hue region in the hue channel and the grayscale histogram of the region corresponding to the dark region of the saturation channel in the brightness channel;

根据色调通道中的色调区域的灰度直方图设定色调数组，以及根据亮度通道中与饱和度通道暗区域的对应区域灰度直方图设定亮度数组，并根据色调数组和亮度数组获取对应图像的颜色信息；Set the hue array according to the grayscale histogram of the hue area in the hue channel, and set the brightness array according to the grayscale histogram of the corresponding area in the brightness channel and the dark area of the saturation channel, and obtain the corresponding image according to the hue array and the brightness array color information;

根据纹理图像的颜色信息获取纹理图像的颜色特征。The color features of the texture image are obtained according to the color information of the texture image.

与现有技术相比，本发明的有益效果是：本发明方法通过先对纹理图像进行提取纹理特征序列，再进行图像集的设定、排列以及检索，降低了图像特征的维度，提高了图像处理速度，从而提高了图像的检索速度和准确率。Compared with the prior art, the beneficial effect of the present invention is: the method of the present invention first extracts the texture feature sequence from the texture image, and then sets, arranges and retrieves the image set, thereby reducing the dimension of the image feature and improving the image quality. Processing speed, thereby improving the image retrieval speed and accuracy.

附图说明Description of drawings

图1为纹理图像检索方法的流出图。Figure 1 shows the flow diagram of the texture image retrieval method.

具体实施方式Detailed ways

下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

一种纹理图像检索方法，包括如下步骤：A texture image retrieval method, comprising the steps of:

(1)根据旋转不变局部二值模式算法对目标纹理图像和纹理图像数据库中的每幅图像进行处理，获取目标纹理图像和纹理图像数据库中每幅图像对应的纹理结构模式图像；(1) process each image in the target texture image and the texture image database according to the rotation invariant local binary pattern algorithm, and obtain the texture structure pattern image corresponding to each image in the target texture image and the texture image database;

(2)根据脉冲耦合神经网络模型对目标纹理图像和纹理数据库的每幅图像对应的纹理结构模式图像进行特征提取，得到目标纹理图像和纹理图像数据库中的每幅图像的纹理结构特征序列；(2) According to the pulse-coupled neural network model, the texture structure pattern image corresponding to each image of the target texture image and the texture database is extracted to obtain the texture structure feature sequence of each image in the target texture image and the texture image database;

(3)计算目标纹理图像的纹理结构特征序列和纹理图像数据库中的每幅图像的纹理结构特征序列进行相似度，将纹理图像数据库中的每幅图像的纹理结构特征序列相似度大于第一设定阈值的图像组合成第一纹理图像集；(3) Calculate the similarity between the texture structure feature sequence of the target texture image and the texture structure feature sequence of each image in the texture image database, and make the texture structure feature sequence similarity of each image in the texture image database greater than the first set The images with a fixed threshold are combined into a first texture image set;

(4)获取目标纹理图像和第一纹理图像集中每一纹理图像的颜色特征；(4) Obtain the color feature of each texture image in the target texture image and the first texture image set;

(5)计算目标纹理图像与第一纹理图像集中每一图像之间的颜色特征相似度，将颜色特征相似度大于第二设定阈值的图像组合成第二图像集；(5) Calculate the color feature similarity between the target texture image and each image in the first texture image set, and combine the images with the color feature similarity greater than the second preset threshold into a second image set;

(6)根据与目标纹理图像之间的颜色特征相似度由高到低的顺序，对第二图像集中的图像进行排列显示。(6) Arranging and displaying the images in the second image set according to the descending order of the color feature similarity with the target texture image.

在上述步骤中，获取目标图像和第一图像集中每一图像的颜色特征具体包括：In the above steps, obtaining the color features of each image in the target image and the first image set specifically includes:

将目标纹理图像与第一纹理图像集中每一纹理图像转换为色调、饱和度和亮度图像格式，获取每一格式转换后的图像的色调通道、饱和度通道和亮度通道；converting each texture image in the target texture image and the first texture image set into a hue, saturation and brightness image format, and obtaining the hue channel, saturation channel and brightness channel of the converted image in each format;

对饱和度通道进行二值化处理，得到饱和度通道的亮区域和暗区域，将饱和度通道的亮区域对色调通道进行投影获得色调通道的色调区域，以及将饱和度通道的暗区域对亮度通道进行投影获得亮度通道中与饱和度通道暗区域对应的区域，并统计色调通道中的色调区域的灰度直方图以及所述亮度通道中与饱和度通道暗区域的对应区域灰度直方图；Binarize the saturation channel to obtain the bright and dark areas of the saturation channel, project the bright area of the saturation channel to the hue channel to obtain the hue area of the hue channel, and convert the dark area of the saturation channel to the brightness The channel is projected to obtain the region corresponding to the dark region of the saturation channel in the brightness channel, and the grayscale histogram of the hue region in the hue channel and the grayscale histogram of the region corresponding to the dark region of the saturation channel in the brightness channel;

根据色调通道中的色调区域的灰度直方图设定色调数组，以及根据亮度通道中与饱和度通道暗区域的对应区域灰度直方图设定亮度数组，并根据色调数组和亮度数组获取对应图像的颜色信息；Set the hue array according to the grayscale histogram of the hue area in the hue channel, and set the brightness array according to the grayscale histogram of the corresponding area in the brightness channel and the dark area of the saturation channel, and obtain the corresponding image according to the hue array and the brightness array color information;

根据纹理图像的颜色信息获取纹理图像的颜色特征。The color features of the texture image are obtained according to the color information of the texture image.

Claims (2)

1.一种纹理图像检索方法，其特征在于，包括如下步骤：1. a texture image retrieval method, is characterized in that, comprises the steps: (1)根据旋转不变局部二值模式算法对目标纹理图像和纹理图像数据库中的每幅图像进行处理，获取目标纹理图像和纹理图像数据库中每幅图像对应的纹理结构模式图像；(1) process each image in the target texture image and the texture image database according to the rotation invariant local binary pattern algorithm, and obtain the texture structure pattern image corresponding to each image in the target texture image and the texture image database; (2)根据脉冲耦合神经网络模型对目标纹理图像和纹理数据库的每幅图像对应的纹理结构模式图像进行特征提取，得到目标纹理图像和纹理图像数据库中的每幅图像的纹理结构特征序列；(2) According to the pulse-coupled neural network model, the texture structure pattern image corresponding to each image of the target texture image and the texture database is extracted to obtain the texture structure feature sequence of each image in the target texture image and the texture image database; (3)计算目标纹理图像的纹理结构特征序列和纹理图像数据库中的每幅图像的纹理结构特征序列进行相似度，将纹理图像数据库中的每幅图像的纹理结构特征序列相似度大于第一设定阈值的图像组合成第一纹理图像集；(3) Calculate the similarity between the texture structure feature sequence of the target texture image and the texture structure feature sequence of each image in the texture image database, and make the texture structure feature sequence similarity of each image in the texture image database greater than the first set The images with a fixed threshold are combined into a first texture image set; (4)获取目标纹理图像和第一纹理图像集中每一纹理图像的颜色特征；(4) Obtain the color feature of each texture image in the target texture image and the first texture image set; (5)计算目标纹理图像与第一纹理图像集中每一图像之间的颜色特征相似度，将颜色特征相似度大于第二设定阈值的图像组合成第二图像集；(5) Calculate the color feature similarity between the target texture image and each image in the first texture image set, and combine the images with the color feature similarity greater than the second preset threshold into a second image set; (6)根据与目标纹理图像之间的颜色特征相似度由高到低的顺序，对第二图像集中的图像进行排列显示。(6) Arranging and displaying the images in the second image set according to the descending order of the color feature similarity with the target texture image. 2.根据权利要求1所述的纹理图像检索方法，其特征在于，所述获取目标图像和第一图像集中每一图像的颜色特征具体包括：2. The texture image retrieval method according to claim 1, wherein said acquiring the target image and the color features of each image in the first image set specifically comprises: 将目标纹理图像与第一纹理图像集中每一纹理图像转换为色调、饱和度和亮度图像格式，获取每一格式转换后的图像的色调通道、饱和度通道和亮度通道；converting each texture image in the target texture image and the first texture image set into a hue, saturation and brightness image format, and obtaining the hue channel, saturation channel and brightness channel of the converted image in each format; 对饱和度通道进行二值化处理，得到饱和度通道的亮区域和暗区域，将饱和度通道的亮区域对色调通道进行投影获得色调通道的色调区域，以及将饱和度通道的暗区域对亮度通道进行投影获得亮度通道中与饱和度通道暗区域对应的区域，并统计色调通道中的色调区域的灰度直方图以及所述亮度通道中与饱和度通道暗区域的对应区域灰度直方图；Binarize the saturation channel to obtain the bright and dark areas of the saturation channel, project the bright area of the saturation channel to the hue channel to obtain the hue area of the hue channel, and convert the dark area of the saturation channel to the brightness The channel is projected to obtain the region corresponding to the dark region of the saturation channel in the brightness channel, and the grayscale histogram of the hue region in the hue channel and the grayscale histogram of the region corresponding to the dark region of the saturation channel in the brightness channel; 根据色调通道中的色调区域的灰度直方图设定色调数组，以及根据亮度通道中与饱和度通道暗区域的对应区域灰度直方图设定亮度数组，并根据色调数组和亮度数组获取对应图像的颜色信息；Set the hue array according to the grayscale histogram of the hue area in the hue channel, and set the brightness array according to the grayscale histogram of the corresponding area in the brightness channel and the dark area of the saturation channel, and obtain the corresponding image according to the hue array and the brightness array color information; 根据纹理图像的颜色信息获取纹理图像的颜色特征。The color features of the texture image are obtained according to the color information of the texture image.
CN201510250553.7A 2015-05-13 2015-05-13 Texture image retrieving method Pending CN104834732A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510250553.7A CN104834732A (en) 2015-05-13 2015-05-13 Texture image retrieving method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510250553.7A CN104834732A (en) 2015-05-13 2015-05-13 Texture image retrieving method

Publications (1)

Publication Number Publication Date
CN104834732A true CN104834732A (en) 2015-08-12

Family

ID=53812618

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510250553.7A Pending CN104834732A (en) 2015-05-13 2015-05-13 Texture image retrieving method

Country Status (1)

Country Link
CN (1) CN104834732A (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6594386B1 (en) * 1999-04-22 2003-07-15 Forouzan Golshani Method for computerized indexing and retrieval of digital images based on spatial color distribution
CN101551823A (en) * 2009-04-20 2009-10-07 浙江师范大学 Comprehensive multi-feature image retrieval method
CN102663391A (en) * 2012-02-27 2012-09-12 安科智慧城市技术（中国）有限公司 Image multifeature extraction and fusion method and system
CN103106265A (en) * 2013-01-30 2013-05-15 北京工商大学 Method and system of classifying similar images
CN103605811A (en) * 2013-12-10 2014-02-26 三峡大学 Texture image retrieval method and device
CN104462481A (en) * 2014-12-18 2015-03-25 浪潮(北京)电子信息产业有限公司 Comprehensive image retrieval method based on colors and shapes
CN104572971A (en) * 2014-12-31 2015-04-29 安科智慧城市技术（中国）有限公司 Image retrieval method and device

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6594386B1 (en) * 1999-04-22 2003-07-15 Forouzan Golshani Method for computerized indexing and retrieval of digital images based on spatial color distribution
CN101551823A (en) * 2009-04-20 2009-10-07 浙江师范大学 Comprehensive multi-feature image retrieval method
CN102663391A (en) * 2012-02-27 2012-09-12 安科智慧城市技术（中国）有限公司 Image multifeature extraction and fusion method and system
CN103106265A (en) * 2013-01-30 2013-05-15 北京工商大学 Method and system of classifying similar images
CN103605811A (en) * 2013-12-10 2014-02-26 三峡大学 Texture image retrieval method and device
CN104462481A (en) * 2014-12-18 2015-03-25 浪潮(北京)电子信息产业有限公司 Comprehensive image retrieval method based on colors and shapes
CN104572971A (en) * 2014-12-31 2015-04-29 安科智慧城市技术（中国）有限公司 Image retrieval method and device

Similar Documents

Publication Publication Date Title
CN108562589B (en) Method for detecting surface defects of magnetic circuit material
CN106203430B (en) A kind of conspicuousness object detecting method based on foreground focused degree and background priori
CN104966085B (en) A kind of remote sensing images region of interest area detecting method based on the fusion of more notable features
CN104463195B (en) Printing digit recognizing method based on template matches
CN110210387B (en) Insulator target detection method, system and device based on knowledge graph
CN106610969A (en) Multimodal information-based video content auditing system and method
CN108388905B (en) A Light Source Estimation Method Based on Convolutional Neural Network and Neighborhood Context
CN104899877A (en) Image foreground extraction method based on super-pixels and fast three-division graph
Huang et al. Automated hemorrhage detection from coarsely annotated fundus images in diabetic retinopathy
CN105761260B (en) A kind of skin image affected part dividing method
CN107369162B (en) Method and system for generating insulator candidate target area
CN105608454A (en) Text structure part detection neural network based text detection method and system
CN109447111B (en) Remote sensing supervision classification method based on subclass training samples
CN106096542A (en) Image/video scene recognition method based on range prediction information
CN104462481A (en) Comprehensive image retrieval method based on colors and shapes
CN104766344B (en) Vehicle checking method based on movement edge extractor
CN106780428B (en) Chip quantity detection method and system based on color recognition
CN104809245A (en) Image retrieval method
CN106296670A (en) A kind of Edge detection of infrared image based on Retinex watershed Canny operator
CN106570515A (en) Method and system for treating medical images
CN110738218A (en) Method and device for identifying hidden danger of smoke and fire of power transmission line channels
CN118799919A (en) A full-time multimodal person re-identification method based on simulation augmentation and prototype learning
CN106485266A (en) A kind of ancient wall classifying identification method based on extraction color characteristic
CN101866422A (en) A method of extracting image attention based on image multi-feature fusion
CN107818341A (en) A kind of color extraction method based on improvement K means algorithms

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
EXSB Decision made by sipo to initiate substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20150812