CN101034442A - System for judging between identical and proximate goods appearance design based on pattern recognition - Google Patents

System for judging between identical and proximate goods appearance design based on pattern recognition Download PDF

Info

Publication number
CN101034442A
CN101034442A CN 200610018491 CN200610018491A CN101034442A CN 101034442 A CN101034442 A CN 101034442A CN 200610018491 CN200610018491 CN 200610018491 CN 200610018491 A CN200610018491 A CN 200610018491A CN 101034442 A CN101034442 A CN 101034442A
Authority
CN
China
Prior art keywords
image
module
color
shape
information
Prior art date
Application number
CN 200610018491
Other languages
Chinese (zh)
Inventor
刘欣融
Original Assignee
刘欣融
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 刘欣融 filed Critical 刘欣融
Priority to CN 200610018491 priority Critical patent/CN101034442A/en
Publication of CN101034442A publication Critical patent/CN101034442A/en

Links

Abstract

The invention is judgment system of the same and approximate designs of products based on image recognition technology. Being analysed images to be specified,we can identify every image of the described image database in the shape, color, design information through image preprocessing modules, the stated information is put into feature extraction module, the respectively corresponded to the analysis module,to complete extraction of information into a group of eigenvalues of images in shape, pattern and color; after data of image eigenvalue group being input similarity comparison module, will be output of similarity judgment results. The system of this invention enables criteria of appearance design of commodity unified, objective, and more scientifi, and to provide reliability, accuracy for relevant agencies and experts, to improve work efficiency, to effectively reduce the appearance of the infringement case from the source, to cost savings, and to be widely used.

Description

基于图像识别技术的商品的外观设计相同和相近似判断系统 Same design image recognition technique and goods Analyzing System Based approximated

发明领域本发明涉及图像识别技术领域,涉及从存储有商品外观设计的图像上提取所需形状、图案、色彩信息特征值进行相同和相近似的判断系统。 Field of the Invention The present invention relates to a technical field of image recognition, there is extracted from the memory relates to a desired shape, pattern, color information on the image feature values ​​of the same design and product approximating determination system.

发明背景图像识别技术,作为检索图像的方法及程序,应用于较多的是从图像数据库中检索出相似图像,方法一:人们预先对图像赋予反映其内容的关键字,在进行检索时,从图像数据库中将被赋予了与用户所输入的关键字相同的关键字的图像提取出来进行提示;方法二:信息处理装置利用将图像所具有的色彩及形状、纹理等图像的物理特征进行定量化的属性值自动地进行检索,将参照图像和各属性值进行比较,从图像数据库中将相似性高的图像作为检索结果提取出进行提示。 Background of the Invention image recognition technology, as a method for retrieving an image and a program applied to more similar image is retrieved, a method from the image database: it gives advance keywords to the image to reflect its content, during retrieval, from the image in the database is given the keyword entered by the user with the same image is extracted keyword prompt; method two: the physical characteristics of the information processing apparatus having a picture image color and shape, texture, etc. will be quantified using attribute value is automatically retrieved, the reference image and each attribute value, from the image in the image database of high similarity are extracted as a search result prompt.

在检索的相关专利中,中国专利申请CN200410042237.2公开了一种图像检索装置及图像检索程序,图像输入单元;求得图像的属性值的属性值获得单元;图像存储单元;将从输入的或存储的多个图像中所选择的图像作为第1参照图像,根据所述属性值,从所存储的图像中检索与所述第1参照图像相似的至少一个第1图像的第1检索单元;显示所述被检索出的第1图像的缩小图像的单元;使图像检索的要求者选择与所述第1参照图像相似的至少一个第2图像的图像选择单元;对所有的图像,重新设置作为用于赋予表示和所述第1参照图像是否相似的符号的数据区域的类别,对所述被选择的至少一个第2图像赋予表示与该类别相似的符号的符号赋予单元;对应所述类别赋予表示相似的可靠度的数值的数值分配单元。 In a related patent search, Chinese Patent Application CN200410042237.2 discloses an image retrieval apparatus and an image retrieval program, the image input unit; attribute value attribute value obtained by the image obtaining unit; image storing means; or from the input the plurality of images stored in the selected image as the first reference image, according to the attribute value, from the stored image retrieving similar to the first reference image of at least a first retrieval unit of the first image; display It means the reduced image is retrieved first image; that the image retrieval requester selects the first reference image similar to the image selection unit at least one second image; all the images, with re-set as imparting to the category data area and indicates whether or not the reference image similar to the first symbol of the at least a second image represented by the selected given symbol like the symbol category imparting unit; corresponds to the category represented by imparting similar numerical value dispensing unit reliability.

又如,中国专利申请CN02120598.1公开了一种自动检索图像数据库中相似图像的方法,将基于柱状图的特征与子波帧分解特征相结合,以及双流程渐进检索处理。 As another example, Chinese patent application CN02120598.1 discloses an automatic method of similar image retrieval in an image database, combines feature-based and histogram-frame wavelet decomposition characteristics, and bis progressive search processing flow. 所提出的本发明相对于照明的变化以及几何失真而言是稳定的。 The present invention is proposed with respect to changes in lighting, and is stable in terms of geometric distortion.

中国专利申请CN02100946.5公开了一种基于色彩和图像特征组合的图像检索方法,其通过适当组合色彩和纹理特征来检索与查询图像相似的数据图像。 Chinese patent application CN02100946.5 discloses an image retrieval method based on a combination of color and image features, which query to retrieve an image similar to the image data by appropriately combining color and texture features.

诸如此类的专利都不能避免图像检索的准确率问题,同时上述这些检索方法及程序,都未能涉及基于图像内容的形状、色彩、图案及结合的检索及比较,特别是图像的相似程度比较。 Patent like can not avoid the problem of accuracy of image retrieval, while these methods and procedures for retrieval, retrieval and comparison failed relates to image content based on the shape, color, pattern and a combination of particular degree of similarity of the image.

然而,实际生活中,商品的外观设计图片的比较,需要利用图像识别技术,对基于图像内容的形状、色彩、图案及结合进行相似程度比较。 However, in real life, the design images of the item comparison, the need to use an image recognition technique, the shape, color, pattern, based on image content and the coupling degree of similarity is compared.

商品的外观设计富于美感且有特色,是商家区别各种同类产品突显自身特点的一种重要手段。 Product design aesthetically pleasing appearance and characteristics, is the difference between the various businesses of similar products highlights an important means of its own characteristics. 现有市场上存在商品的外观设计比较相近似的问题,造成市场的混乱,消费者很容易混淆,厂家之间并由此引起仿冒产品的知识产权纠纷。 The presence of goods on the existing market design more similar to similar problems, causing confusion in the market, consumers can easily be confused, and the resulting intellectual property disputes between the manufacturers of counterfeit products.

对商品外观设计是否相近似的判断方法都是基于一个判断标准,依据此标准专家对两个商品的外观设计图像进行比较,对图像内容的形状、色彩、图案及结合进行比较,专家对判定标准的理解与适用,存在非统一性、主观性,为了解决这一问题,提出一套基于图像识别技术的商品的外观设计相同和相近似判断系统。 Whether approximating product design methods are based on a determination criterion, according to this standard design experts product compared two images, the shape of the image content, color, pattern, and comparing binding, expert criteria understanding and Application, the presence of non-uniformity, subjective, in order to solve this problem, a design based on the same image recognition technique and goods approximated determination system.

发明目的本发明的目的就是提供一种基于图像识别技术的对图像的形状、色彩、图案及结合的进行识别并进行商品的外观设计相似程度的比较软件系统。 Purpose The purpose of the present invention is designed to provide a degree of similarity comparison software based system on the image shape, color, pattern recognition and binding of image recognition technique and product appearance.

本发明的另一个目的就是提供一种让消费者、商家直接应用,给行政、司法诉讼机关及专利复审委员会提供标准统一的判断原则,具有客观公正性的商品的外观设计相同和相近似程度判断结论系统,缩短工作周期,提高工作效率,节省工作费用。 Another object of the invention is to provide a way for consumers, merchants applied directly to the executive, the judiciary and the judicial bodies of the Patent Reexamination Board to provide a unified standard for rules, have the same design objectivity and fairness of the goods and the approximate degree of judgment phase Conclusion system, shorten cycle times, improve work efficiency, saving the cost of the work.

发明概述本发明针对待比较图像的形状、色彩、图案及结合的整体与要部设置参考点作为判断要素,按照形状、图案、色彩及结合等不同要求,把一张图片转化为提取图像的特征值,对两张图片进行比较时,确定特征值取值范围对应显示相似程度,得到比较图像相似程度结论。 SUMMARY OF THE INVENTION for the comparison image to be the shape, color, pattern and a portion to be bonded integrally with the set reference point is determined as a feature, in accordance with the requirements of the different shapes, patterns, colors and the like in combination, the extracted image into an image feature value, when two images are compared to determine the characteristic value range corresponding to display the degree of similarity, similarity comparison image obtained conclusions.

本发明基于图像识别技术的商品的外观设计相同和相近似判断系统,包括:一个图像的输入模块,用于指定待比较分析图像,或一个图像数据库单元,包含一个或多个待比较分析图像。 The present invention is based on the same design image recognition technique and goods approximated determination system, comprising: an input module of an image to be compared for designating image analysis, an image or a database unit, comprising one or more image analysis to be compared.

图像预处理模块,用于确定所述图像数据库存中的每一个图像的形状、色彩、图案信息;图像预处理模块采用图像边缘检测模块提取图像的边缘信息,采用图像分割模块将边缘信息二值化,在分割后的二值图像中,采用图像区域填充模块,应用图像填充算法,计算图像面积,在填充的图像中,采用边缘提取模块,完成在图像中提取出所需要的外形、图案和颜色信息。 The image pre-processing module, the shape of each image of the image data in the stock, the color, pattern information for determining; image preprocessing module using an image edge detection module extracts the image edge information, using the edge information of the image segmentation module binary of, the binary image after segmentation using image area filling module, apply image filling algorithm to calculate an image area in the filled image using edge detection module to complete the extracted shape desired in the images, patterns and colors information.

特征提取模块,包括图像形状分析模块、图案分析模块和颜色分析模块,所述图像的形状、色彩、图案信息分别进入各对应分析模块,完成提取的信息转化为图像的外形、图案和颜色一组特征值,特征量化的属性值单元。 Feature extraction module, an analysis module includes an image shape, color and pattern analysis module analyzing module, the shape of the image, color, respectively, into the pattern data corresponding to each analysis module to complete the extracted information into shape, and color of the image pattern group eigenvalue, the quantization unit attribute values.

其中,特征提取模块的图像形状分析模块采用分割和填充算法完成图像周长和面积的计算;图像图案分析模块采用分割算法,用二维向量进行描述;图像色彩分析模块采用各个分量直方图的均值和方差作为比较的数值,可以提取三个二维的向量的向量来描述。 Wherein the feature extraction module using image segmentation and shape analysis module to complete the filling algorithm and calculating the image area perimeter; image pattern analysis module using the segmentation algorithm, described by two-dimensional vector; image analysis module uses color histograms of each component of the mean and a variance value comparison can be extracted three-dimensional vector of the vectors described.

相似度比较模块,图像特征值组数据输入相似度比较模块,所述相似度比较模块含有比较原则确定的比较算法;相似度比较模块可以采用模糊交操作来定义度量的差别单位。 Similarity comparison module, the image data input feature value group similarity comparison module, the comparison module contains a similarity comparison algorithm comparison rules determined; similarity comparison module can be operative to define fuzzy post difference metric units.

一个输出模块,用于输出相似程度判断结果且可以用百分比表示。 An output module for outputting the degree of similarity and the determination result can be expressed as a percentage.

本发明的优点将通过下面所述的详细说明清楚地表现出来,可以通过给出的详细的说明和特定的实施例予以理解,并对本领域的人员而言是显而易见。 Advantages of the invention will be clearly demonstrated by the following detailed description, given by can be appreciated that the detailed description and specific embodiments, and the terms will be apparent to those skilled.

首先,该系统的应用优点:1、使用现代化高科技手段提高司法鉴定工作层次,应用判断系统软件提高鉴定工作效率与准确率;2、判定标准统一,具有客观公正性,排除各种非专业性因素的干扰;3、避免了专家或合议组成员对判断原则的适用与理解的非统一性,主观意见的主导地位;4、鉴定机构专家组鉴定意见文字描述与相近似程度判断结论有机结合,提供可靠性和准确性的咨询意见。 First, the application advantages of the system: 1, the use of modern high-tech means to improve the forensic work level, the application judge system software to improve the efficiency and accuracy of identification; 2, unified criteria, with objectivity and fairness, to remove all non-professional disturbing factors; 3, to avoid a collegial group of experts or members of non-uniformity of the application of the principle of judgment and understanding, subjective opinions of dominance; 4, accreditation bodies and expert group views the text description was similar conclusions to determine the extent of organic combination, providing the reliability and accuracy of advice.

5、消费者、专利权人、商家直接应用,缩短工作周期,提高工作效率,对专利权人可以事先预知自己的外观设计是否受到侵权,对于进入诉讼,可以预测自己的外观设计是否侵权,对商家可以做到产品的前景预测;给商家和专利权人带来巨大的商业价值。 5, the consumer, the patentee, direct business applications, shorten cycle times, improve work efficiency, the patentee can predict in advance whether by their own design infringement, for litigation, can predict whether their design infringement, for businesses can do to predict the prospects of the product; tremendous business value to merchants and the patentee.

附图概述通过下面给出的详细说明结合附图将更全面的理解本发明,本发明仅给出示例性的用途,并不仅限于本发明,也可在本领域有其它应用。 BRIEF DESCRIPTION OF THE DRAWINGS The invention fully understood from the detailed description will be given below in conjunction with the present invention are given merely illustrative purposes, and is not limited to the present invention, there may be other applications in the art.

图1本发明示意性实施例的系统方框图。 The system of the present invention Figure 1 a schematic block diagram of exemplary embodiments.

图2本发明示意性实施例的图像预处理模块方框图。 The present invention FIG. 2 a schematic block diagram of the image pre-processing module embodiment.

实施例的详细说明本发明判断系统第一步:一个图像的输入模块,用于指定待比较分析图像;一个图像数据库单元,包含一个或多个待比较分析图像。 Detailed description of the invention The first step in determining system embodiment: an input module of an image, analyzing the image to be compared for designating; a unit image database, contains one or more image analysis to be compared. 用扫描或拍摄将商品的外观设计图像通过计算机输入模块,指定原始图像与待比较匹配图像进入工作状态。 Scanning or photographing the image of goods design computer input module, specify the original image and the matching image to be compared into operation.

第二步,图像预处理模块确定每一个图像的形状、色彩、图案信息,采用Sobel算子提取图像的边缘信息,采用OTSU自适应分割算法将边缘信息二值化,便于后面填充算法的处理。 The second step, image preprocessing module determines the shape of each image, the color, pattern information, using the Sobel operator extracting the image edge information, using the adaptive segmentation algorithm OTSU binary edge information, to facilitate handling of the back filling algorithm. 在分割后的二值图像中,采用基于八领域扫描的图像填充算法,从图像的左上角开始扫描图像,按顺时针方向扫描领域范围,填充物品的内部区域,便于计算图像的面积。 In the binary image segmentation using the image filling algorithm based on eight field scanning, the image is scanned top-left corner of the image, the field of scanning range in a clockwise direction, filling the interior region of the article, to facilitate the calculation of the area of ​​the image. 在填充的图像中,采用提取图像轮廓的方法提取物品的周长。 Filling the image, an image extraction method to extract the contour of the perimeter of the article. 操作系统软件指令通过图像预处理模块确定获取图像形状信息,图像色彩信息,图像图案信息,形状、色彩结合信息、图像形状、色彩、图案及结合等多种信息。 Operating system software instructions determined by the preprocessing module obtains the image information of image shape, image color information, the image information of the pattern, shape, color with a variety of information, image shape, color, pattern and the like in combination.

经过预处理的图像形状、色彩、图案信息处理的信息进入特征提取模块,进入图像形状分析模块信息,采用三个要素刻画商品的外型,周长、面积和周长和面积的比值,也可采用预处理算法可以比较精确的完成图像周长和面积的计算,提取上面一个三维向量对形状进行描述。 After the pre-image shape, color, pattern information into the feature information extraction processing module, into the shape of the image analysis module information using three exterior elements portrayed goods, perimeter, area and perimeter and area ratio, can be preprocessing algorithm can be employed to complete more precise calculation of perimeter and area of ​​the image, extracting a three-dimensional vector of the above shape will be described.

进入图案分析模块信息量采用OTSU分割算法来完成,由分割算法自适应确定分割的阈值,和图像的亮度直方图的均值组成一个二维向量进行描述。 Using pattern information into the analysis module to complete the OTSU segmentation algorithm, the luminance is divided by a partitioning algorithm adaptively determined threshold value, and a histogram of the image of a two-dimensional vector consisting of the mean will be described. 进入颜色分析模块信息,采用R,G,B三个分量各自的直方图来完成,由相似图大致可以判断三个分量的直方图具有相似性,对于直方图,采用各个分量直方图的均值和方差作为比较的数值,即对于颜色可以提取三个二维的向量来描述。 Entering color information analysis module, using R, G, B components each of three histograms accomplished by a generally similar histogram based on three components having similar properties, for the histogram, the histogram using each component and the mean as a comparison the value of variance, i.e., the color can be described with respect to the extracted three-dimensional vectors.

因此,图像的形状、色彩、图案信息分别进入各对应分析模块,完成提取的信息转化为图像的外形、图案和颜色一组特征值,特征量化的属性值单元。 Therefore, the shape of the image, color, respectively, into the pattern data corresponding to each analysis module to complete the information into a shape, a pattern and a color of an image characteristic value set extracted characteristic property values ​​quantized units.

相似度比较模块,图像特征值组数据输入相似度比较模块,所述相似度比较模块采用模糊数学中的模糊交操作来定义度量的差别单位,确定特征值取值范围对应显示相似程度,采用如下的公式计算图像的相似度,调整权值门限,使计算机相似度尽可能的符合客观的判别标准:ρ=|IΛw||I|]]>其中I为原始图像的特征值向量,W为比较图像的特征值向量,p为最终的相似度比较结果,Λ为模糊交操作,为(pΛq)=min(piΛqi),|p|为L1范数:|p|=Σi=1N|pi|]]>系统软件对产品的特征值取值范围对应的产品的相似程度及侵权程度的判断表 Similarity comparison module, the image data input feature value group similarity comparison module, the comparison module fuzzy similarity deposit operation fuzzy mathematics to define different units of measurement, determined characteristic value range corresponding to the display similarity, as follows similarity formula image, adjust the weights threshold, causing a computer similarity as much as possible in line with the objective criterion: & rho; = | I & Lambda; w || I |]]> where I characteristic value vector of the original image, W is a comparison image feature value vector, p is the final similarity comparison, Lambda fuzzy post operation, is (pΛq) = min (piΛqi), | p | is the L1 norm: | p | = & Sigma; i = 1N | pi |]]> and the degree of similarity of the characteristics of the product software infringement value range determination table corresponding to the product

第三步,通过输出模块输出相似程度百分比判断结果。 A third step, the output percentage similarity determination result through the output module.

Claims (7)

1.基于图像识别技术的商品的外观设计相同和相近似判断系统,包括:一个图像的输入模块,用于指定待比较分析图像;或一个图像数据库单元,包含一个或多个待比较分析图像;图像预处理模块,用于确定所述图像数据库中的每一个图像的形状、色彩、图案信息;特征提取模块,包括图像形状分析模块、图案分析模块和颜色分析模块,所述图像的形状、色彩、图案信息分别进入各对应分析模块,完成提取的信息转化为图像的外形、图案和颜色一组特征值,特征量化的属性值单元;相似度比较模块,图像特征值组数据输入相似度比较模块,所述相似度比较模块含有比较原则确定的比较算法;一个输出模块,用于输出相似程度判断结果。 1. Designs based on the same image recognition technique and goods approximated determination system, comprising: an input module of an image to be compared for designating image analysis; or a unit image database, contains one or more image analysis to be compared; the image pre-processing module, the shape of each image in the image database, the color, pattern information for determining; feature extraction module, the image shape analysis module comprises a shape, pattern, color analysis module and analysis module, the image, color , respectively, into the information pattern information corresponding to each analysis module to complete the conversion of the extracted shape, and color of the image pattern set eigenvalue, the quantization unit attribute values; similarity comparison module, the image data input feature value group similarity comparison module , the similarity comparison module comprising the principle of comparing the determined comparison algorithm; an output module for outputting the degree of similarity determination result.
2.根据权利要求1所述的系统,其中图像预处理模块采用图像边缘检测模块提取图像的边缘信息,采用图像分割模块将边缘信息二值化,在分割后的二值图像中,采用图像区域填充模块,应用图像填充算法,计算图像面积,在填充的图像中,采用边缘提取模块,完成在图像中提取出所需要的外形、图案和颜色信息。 2. The system according to claim 1, wherein the image pre-processing module using an image edge detection module extracts the image edge information, using the edge information of the image segmentation module binarized, the binary image after segmentation using image area filling module using an image filling algorithm to calculate an image area in the filled image, using an edge extraction module extracts the complete shape, pattern and color information in the image required.
3.根据权利要求1所述的系统,其中特征提取模块的图像形状分析模块采用分割和填充算法完成图像周长和面积的计算。 3. System according to claim 1, wherein the feature extraction module using image segmentation and shape analysis module to complete the filling algorithm and calculating the image area perimeter.
4.根据权利要求1所述的系统,其中特征提取模块的图像图案分析模块采用分割算法,用二维向量进行描述。 4. The system of claim 1, wherein the image feature extraction module using the pattern analysis module segmentation algorithm, described by a two-dimensional vector.
5.根据权利要求1所述的系统,其中特征提取模块的图像色彩分析模块采用各个分量直方图的均值和方差作为比较的数值,可以提取三个二维的向量的向量来描述。 5. The system according to claim 1, wherein the image feature extraction module each component color analysis module uses the histogram mean and variance values ​​as the comparison, three-dimensional vectors may be extracted vectors will be described.
6.根据权利要求1所述的系统,其中相似度比较模块可以采用模糊交操作来定义度量的差别单位。 6. The system according to claim 1, wherein the difference between the similarity comparison module unit operable to define cross-fuzzy metrics may be employed.
7.根据权利要求1所述的系统,其中输出模块输出的相似程度判断结果可以用百分比表示。 7. The system according to claim 1, wherein the output module outputs the degree of similarity determination result can be expressed as a percentage.
CN 200610018491 2006-03-08 2006-03-08 System for judging between identical and proximate goods appearance design based on pattern recognition CN101034442A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200610018491 CN101034442A (en) 2006-03-08 2006-03-08 System for judging between identical and proximate goods appearance design based on pattern recognition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200610018491 CN101034442A (en) 2006-03-08 2006-03-08 System for judging between identical and proximate goods appearance design based on pattern recognition

Publications (1)

Publication Number Publication Date
CN101034442A true CN101034442A (en) 2007-09-12

Family

ID=38730989

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200610018491 CN101034442A (en) 2006-03-08 2006-03-08 System for judging between identical and proximate goods appearance design based on pattern recognition

Country Status (1)

Country Link
CN (1) CN101034442A (en)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100550038C (en) 2007-12-26 2009-10-14 腾讯科技(深圳)有限公司 Image content recognizing method and recognition system
CN102542470A (en) * 2011-12-22 2012-07-04 钱袋网(北京)信息技术有限公司 Card reading method and card reading device of mobile terminal
CN102930537A (en) * 2012-10-23 2013-02-13 深圳市宜搜科技发展有限公司 Image detection method and system
CN102955784A (en) * 2011-08-19 2013-03-06 北京百度网讯科技有限公司 Equipment and method for judging similarity of various images on basis of digital signatures
CN102982332A (en) * 2012-09-29 2013-03-20 顾坚敏 Retail terminal goods shelf image intelligent analyzing system based on cloud processing method
CN103198317A (en) * 2011-12-14 2013-07-10 韩国电子通信研究院 Image processing device and image processing method
CN103295346A (en) * 2012-02-29 2013-09-11 东芝泰格有限公司 Information processing apparatus, store system and information processing method
WO2014031022A1 (en) * 2012-08-21 2014-02-27 Общество С Ограниченной Ответственностью "Кузнеч" Method and system for searching for copyright infringements in images
WO2015027805A1 (en) * 2013-08-28 2015-03-05 上海合合信息科技发展有限公司 Query method, apparatus, system and client for product description
CN105139012A (en) * 2015-08-25 2015-12-09 长沙市麓智信息科技有限公司 Appearance patent retrieving system based on image pre-processing
CN107451879A (en) * 2017-06-12 2017-12-08 北京小度信息科技有限公司 Information judgment method and device
CN107908685A (en) * 2017-10-31 2018-04-13 西安交通大学 The retrieval of various visual angles commodity image and recognition methods based on transfer learning

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100550038C (en) 2007-12-26 2009-10-14 腾讯科技(深圳)有限公司 Image content recognizing method and recognition system
CN102955784A (en) * 2011-08-19 2013-03-06 北京百度网讯科技有限公司 Equipment and method for judging similarity of various images on basis of digital signatures
CN103198317A (en) * 2011-12-14 2013-07-10 韩国电子通信研究院 Image processing device and image processing method
CN102542470A (en) * 2011-12-22 2012-07-04 钱袋网(北京)信息技术有限公司 Card reading method and card reading device of mobile terminal
CN103295346A (en) * 2012-02-29 2013-09-11 东芝泰格有限公司 Information processing apparatus, store system and information processing method
CN103295346B (en) * 2012-02-29 2016-04-27 东芝泰格有限公司 Signal conditioning package, store system and information processing method
WO2014031022A1 (en) * 2012-08-21 2014-02-27 Общество С Ограниченной Ответственностью "Кузнеч" Method and system for searching for copyright infringements in images
CN102982332A (en) * 2012-09-29 2013-03-20 顾坚敏 Retail terminal goods shelf image intelligent analyzing system based on cloud processing method
CN102930537A (en) * 2012-10-23 2013-02-13 深圳市宜搜科技发展有限公司 Image detection method and system
WO2015027805A1 (en) * 2013-08-28 2015-03-05 上海合合信息科技发展有限公司 Query method, apparatus, system and client for product description
CN105139012A (en) * 2015-08-25 2015-12-09 长沙市麓智信息科技有限公司 Appearance patent retrieving system based on image pre-processing
CN107451879A (en) * 2017-06-12 2017-12-08 北京小度信息科技有限公司 Information judgment method and device
CN107451879B (en) * 2017-06-12 2018-11-02 北京小度信息科技有限公司 Information judgment method and device
CN107908685A (en) * 2017-10-31 2018-04-13 西安交通大学 The retrieval of various visual angles commodity image and recognition methods based on transfer learning

Similar Documents

Publication Publication Date Title
Pentland et al. Photobook: Tools for content-based manipulation of image databases
Van De Weijer et al. Learning color names for real-world applications
Mojsilovic et al. Matching and retrieval based on the vocabulary and grammar of color patterns
Mery et al. Segmentation of colour food images using a robust algorithm
Alajlan et al. Geometry-based image retrieval in binary image databases
Karayev et al. Recognizing image style
Kim et al. Content-based trademark retrieval system using a visually salient feature
Ma et al. EdgeFlow: a technique for boundary detection and image segmentation
Rao A taxonomy for texture description and identification
Cheng et al. Global contrast based salient region detection
Paquet et al. Nefertiti: a query by content software for three-dimensional models databases management
US7689044B2 (en) Method and apparatus for representing a group of images
Nastar et al. Efficient query refinement for image retrieval
Pan et al. Region duplication detection using image feature matching
Li et al. Scale-and rotation-invariant local binary pattern using scale-adaptive texton and subuniform-based circular shift
KR100724932B1 (en) apparatus and method for extracting human face in a image
Meilhac et al. Relevance feedback and category search in image databases
US7127106B1 (en) Fingerprinting and recognition of data
CN101601287B (en) Apparatus and methods of producing photorealistic image thumbnails
Zhong et al. Object localization using color, texture and shape
JP2010507139A (en) Face-based image clustering
US8908996B2 (en) Methods and apparatus for automated true object-based image analysis and retrieval
JP2008097607A (en) Method to automatically classify input image
JP2009187581A (en) Information search and retrieval system
Fauqueur et al. Region-based image retrieval: Fast coarse segmentation and fine color description

Legal Events

Date Code Title Description
C06 Publication
C57 Notification of unclear or unknown address
C57 Notification of unclear or unknown address
C02 Deemed withdrawal of patent application after publication (patent law 2001)