WO2010003286A1 - 一种商标检索方法及系统 - Google Patents
一种商标检索方法及系统 Download PDFInfo
- Publication number
- WO2010003286A1 WO2010003286A1 PCT/CN2008/071581 CN2008071581W WO2010003286A1 WO 2010003286 A1 WO2010003286 A1 WO 2010003286A1 CN 2008071581 W CN2008071581 W CN 2008071581W WO 2010003286 A1 WO2010003286 A1 WO 2010003286A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- trademark
- pattern information
- character string
- unit
- string
- 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.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
Definitions
- the invention relates to the field of pattern recognition technology, in particular to the field of image coding technology, and in particular to a trademark search method and system. Background technique
- Any company manufacturer can sign up with the relevant website to find their information on a website, but it is also the company's contact information.
- Trademarks play a very important role in the industrial and commercial society. They are the logo of a company, product or service. They are integrated with the product quality, service quality and management of the enterprise. They become a symbol of corporate reputation and are an intangible asset. Therefore, trademarks are an important attribute of a company and its products and are unique.
- a further embodiment of the present invention provides a trademark search method, the method comprising the steps of: acquiring pattern information of a trademark; encoding the pattern information to generate a string corresponding to the trademark Establishing and storing a mapping relationship of string, trademark, and product information; inputting a string; finding and outputting the trademark and/or product information matching the input string from the stored mapping relationship.
- an embodiment of the present invention further provides a trademark retrieval system, the system comprising: an image acquisition unit for acquiring pattern information of a trademark; and an image coding unit that encodes the pattern information to generate a character string corresponding to the trademark; a storage unit that stores a mapping relationship between a character string, a trademark, and product information; an input unit that inputs a character string of the trademark to be retrieved; and a retrieval unit that finds out from the stored character string Trademark and/or product information that matches the entered string; an output unit that outputs the retrieved trademark and/or product information.
- the existing online operation is to find the address and then transfer it to the other party's website, in order to check it, but there are often many online operations to actually enter the page, wasting a lot of spirit and time.
- the present invention can directly read the data by inputting the correct trademark code. If you want to find information in certain industries, you can read all the provided materials simply and directly at any time, and since the trademark code is a string, Any one of the trademarks can be uniquely and accurately identified, so that retrieval with high stability can be performed.
- the invention not only avoids the obstacles of text input, but also can select the words that it reads, which is very convenient and more effective for international customers.
- Fig. 1 is a block diagram showing the structure of an icon code conversion apparatus according to a first embodiment of the present invention.
- Fig. 2 is a flow chart showing the method of icon code conversion in the first embodiment of the present invention.
- 3A to 31 are schematic views showing the trademarks of the present invention which are separated according to the structure.
- 4 to 9 are schematic views of the geometric figures and lines when the geometric figures and the lines are distinguished by the present invention.
- FIGS. 10A to 10L are schematic views showing the trademark of the present invention illustrating the encoding method.
- Fig. 11 is a block diagram showing the structure of a trademark search system according to a second embodiment of the present invention.
- Fig. 12 is a flow chart showing a trademark search method according to a second embodiment of the present invention. detailed description
- Embodiment 1 of the present invention provides an icon code conversion method and apparatus for uniquely representing an icon in a simple manner, and the icon in Embodiment 1 of the present invention is a trademark.
- FIG. 1 is a structural block diagram of an icon code conversion apparatus according to Embodiment 1 of the present invention, wherein the icon code conversion apparatus includes: an image acquisition unit 101, configured to acquire an image of the icon; and a pre-processing unit 102, configured to The image is pre-processed to extract a plurality of distinguishable features of the image; the coding unit 103 is configured to use the plurality of the distinguishable features to have a sequence of characters (the string includes Numbers, English letters and punctuation) are represented.
- the icon code conversion apparatus includes: an image acquisition unit 101, configured to acquire an image of the icon; and a pre-processing unit 102, configured to The image is pre-processed to extract a plurality of distinguishable features of the image; the coding unit 103 is configured to use the plurality of the distinguishable features to have a sequence of characters (the string includes Numbers, English letters and punctuation) are represented.
- the pre-processing unit 102 includes: an outer shape classification unit 104, configured to classify the image according to an outer shape, and use the classification result as one feature having the separability; and the structure classification unit 105 is configured to follow
- the images are split in order from top to bottom, left to right, and from outside to inside, and each split structure is classified, and each of the classification results is regarded as one of the separable feature.
- Fig. 2 is a flow chart showing the icon code conversion method of the embodiment 1 of the present invention, and the icon code conversion method will be described in detail below with reference to Fig. 2.
- an image of the above trademark is acquired (step S201); then, the image is classified according to the outer shape, and the classification result is taken as one of the distinguishable features (step S202); from top to bottom, from The images are split from left to right and from outside to inside, and each is split The classified structure is classified, and each of the classified results is regarded as one of the distinguishable features
- Step S203 identifying each of the above-mentioned distinguishable features by characters, that is, identifying each of the above-described distinguishable features by numbers, English letters, and punctuation marks (step S204);
- the above-described numbers, English letters, and punctuation marks having the distinguishable features are arranged in a combination, and the above-described trademarks are represented by a sequenced character string (step S205).
- the order in which the images are decomposed in the first embodiment is in order from top to bottom, from left to right, and from outside to inside, as shown in Figs. 3A to 31.
- the classification of the classification is divided into five broad categories:
- the image of each trademark must first be the shape number, and distinguish the figure that belongs to that type; after the shape number is set,
- the structure of the figure can generally be divided into left and right, or left and right, or upper, middle and lower. As shown in the figure, the circular image is divided into inner, inner, central or outer, inner and central points.
- the rule of the fixed number is to set the front left and the right again.
- the irregular map has a point on the top, and there are points on the bottom. It can only be set to the left side, and the left and right sides of the bottom line. Left and right below.
- the above-mentioned regular map includes generally common geometric figures: geometric shapes such as square, rectangle, triangle, polygon, and circular shape, as long as it is a single geometric figure as the outer shape of the trademark image.
- the regular map is numbered as follows:
- the shape of the regular map must be a geometric figure.
- the shape number is first determined, and the lines or other pattern codes in the map are fixed.
- the figure can be divided into left and right, etc., according to the rule number, called the inner shape number.
- the figure that has been fixed in the figure is not shown in the figure. If the internal shape number can be set to the top, bottom, and bottom, then the middle part does not need a fixed number.
- the above-mentioned irregular map is an external map in which two or more geometric figures are separated or overlapped to form a trademark.
- the irregular map is numbered as follows:
- the irregular map must be composed of two geometric figures. Its shape number can only be represented by the English letter C or E. After the shape number is determined, it is judged whether it is left or right or two up and down.
- the picture is composed of the upper and lower parts, some of which are divided into two small pictures, which must be separated by their own.
- the shape of the above-mentioned special regular figure is a single geometric figure.
- the figure may be a special figure mixed with animal and plant text, rather than a line or a small geometric figure.
- the special regular map is numbered as follows:
- the figure is other special figures.
- the special figure is set with the English letter code. These two numbers are used as the shape number, and the internal shape number is divided according to the rules of the drawing. Or the upper and lower, etc., because the special map does not need to be dismantled, only the code of the special map is used. If there is text, only the letter of each of the first bills is taken, or the number is arranged in alphabetical order, such as the order of A is 1, C is Row 3, D is ranked 4th.
- This kind of pattern must have any picture or various lines mixed with characters, animals and plants.
- a line or other In addition to the external number of the D, and then according to the main body of the map, such as the main body in the middle is a line or other, according to the order is left and right to the number, as shown in the figure is divided into upper, lower, first, middle, and lower .
- the special map is numbered as follows:
- the head of the subject Be sure to follow the head of the subject as the direction of the number. If the head is on the top, it will be fixed from the top to the bottom. If the head is on the left or right, it will be left and right. Moreover, only the main lines or figures inside the figure can be used to determine the number, which is the internal shape number, because the shape number has been designated as a special figure code.
- A, B, C, D, E represent the code of the special pattern, and the symbol "+” and ".”, this symbol represents the addition or payment to distinguish the number.
- the thin rectangular shape of the two pointed ends, the insect shape, the figure of the figure must be thin and single and independent.
- a bay-shaped line can not be connected in the same way, or the two ends can be connected in opposite directions (as shown in Figure 4).
- Curves of all, equal and unequal curves, patterns consisting of curves These curves must have more than two corners or more than two bays and must be a connected curve (as shown in Figure 6).
- horn or half-bay is round or cross-shaped, all of this can't be sealed because it can be seen as an independent appearance and figure. It must be in the shape of a line, and it cannot be in a whole or three-dimensional shape (as shown in Figure 8). .
- the special drawing is an independent outline drawing, it may be divided into a large number of re-markings, and some parts of the text are added to a geometric figure, plus + this symbol to distinguish.
- the shape is a regular geometric figure, there is only text in the figure, you can not directly add the + symbol, you can directly type the shape number and the number 6 and then type the text.
- the unconventional picture can only be divided into left and right, and there is no code in the middle, use the + contemporary number, or the pattern has only the shape number, no matter the left and right, only the middle code, you must use + for the connection, such as the outline code + middle Code within.
- A. Represents all living animals and plants, mainly used in the outline drawing.
- the shape is round, the interior is divided into upper and lower parts, and the upper part is a small circle with a T-shape. Therefore, its Shape No.: 8; Top: 8; Bottom: 1, Full picture: 881.
- the shape is a circular interior that can be divided into upper and lower left and right. First left and right, then up and down, below is the M line. Therefore, its shape number: 8; inner left: 1; inner right: 1; inner upper: 4; inner lower: 3, full map fixed number: 81143.
- the shape is a regular figure of the square table, and the inside is a special figure cartoon character. Because the left and right sides are separated, there is also a picture in the middle, the inner part is first right and then right, and the middle part is up and down. Therefore, its shape number: 4A; left: 7; right: 7; upper: 3; bottom: 8, full map fixed number: 4A7738.
- the shape is a square regular picture, the upper part is the human phase, and the bottom is the text. Therefore, its shape number: 4A; Up: 7; Bottom: 6LEF, full picture: 4A76LEF.
- the shape consists of two separate geometric figures. It is not possible to give two figures a code at the same time. It is only necessary to use the English letters as the code name. These two figures become the internal shape number, which is divided into two parts. It is connected with the symbol "+”. Therefore, its shape number: E; left: 4; right: +4, full picture: E4+4.
- the trademark is a cartoon animated picture, the head of the picture is on the right side.
- the fixed number of such a picture must be left and right.
- the shape number is a special code, after the shape number is set, the number cannot be repeated.
- the lines are fixed by the internal shape number, and the inside of the right is the eye. Therefore, its shape number: A7; inner left: 1; inner right: 8, full picture number: A718.
- the shape number can only be used as the English mother code. Because the inside is round, there is no distinction between left and right, only the inner point, the center point, and so on. Therefore, its shape number: D; Xi Bu: 5; Inside: 8; Center point: 1, Full picture: D581.
- the shape number is D, and the internal shape number is divided into upper and lower. Because there are words in the picture below, there are two kinds of fixed numbers, one is +6, the other is +6 and the first pronunciation is written. Therefore, its external shape number: D; upper: 3; medium: 8; bottom: +6 (M is the first single letter), full picture number: D38+6 or D38+6MM.
- Embodiment 2 of the present invention provides a trademark retrieval method and system.
- Figure 11 is a block diagram showing the structure of the trademark retrieval system of the second embodiment of the present invention, wherein the trademark retrieval system includes: an image acquisition unit 1101 for acquiring an image of the trademark; and an image coding unit 1102 for The image is encoded, and the image is represented by a string having a sequence; the storage unit 1103 is configured to store the image of the trademark and its character string; a symbol 1104, configured to input a code of the trademark image to be retrieved; a retrieval unit 1105, configured to find a character string matching the input code from the stored character string, and display an image of the trademark corresponding to the character string .
- the trademark retrieval system includes: an image acquisition unit 1101 for acquiring an image of the trademark; and an image coding unit 1102 for The image is encoded, and the image is represented by a string having a sequence; the storage unit 1103 is configured to store the image of the trademark and its character string; a symbol 1104, configured to input a code of the trademark image to be retrieved; a retrieval unit
- the encoding unit 1102 includes: a pre-processing unit 1106, configured to perform pre-processing on the image, extracting features of the plurality of the images that are separable; and using the plurality of features that are separable A string with a sequence is represented.
- a pre-processing unit 1106, configured to perform pre-processing on the image, extracting features of the plurality of the images that are separable; and using the plurality of features that are separable A string with a sequence is represented.
- the pre-processing unit 1106 of the second embodiment includes: an outer shape classification unit 1107, configured to classify the image according to an outer shape, and use the classification result as one feature having the separability; a structure classification unit 1108, The image is used to split the image according to the structure, and each of the split structures is classified, and each of the classified results is taken as one of the distinguishable features.
- an outer shape classification unit 1107 configured to classify the image according to an outer shape, and use the classification result as one feature having the separability
- a structure classification unit 1108 The image is used to split the image according to the structure, and each of the split structures is classified, and each of the classified results is taken as one of the distinguishable features.
- Fig. 12 is a flow chart showing the trademark search method of the second embodiment of the present invention, and the trademark search method will be described in detail below with reference to Fig. 12.
- an image of the above-mentioned trademark is acquired (step S1201); then, the image is classified according to the outer shape, and the classification result is taken as one of the distinguishable features (step S1202); from top to bottom, from The images are split from left to right, from outside to inside, and each of the split structures is classified, and each of the classified results is taken as one of the distinguishable features (step S1203)
- Each of the above-mentioned distinguishable features is identified by a character, that is, each of the above-described divisible features is identified by a number, an English letter, and a punctuation mark (step S1204);
- the number of the characteristic features, the English alphabet and the punctuation are arranged in a combination, and the above-mentioned trademark is represented by a string with a sequence (step S1205); the above-mentioned trademark and character string are stored (step S1206); and the code of the trademark to be retrieved is input (step S1207); retrieving the matching character string from the stored character
- the existing online operation is to find the address and then transfer it to the other party's website, in order to check it, but there are often many online operations to actually enter the page, wasting a lot of spirit and time.
- the present invention can directly read the data by inputting the correct trademark code. If you want to find information in certain industries, you can read all the provided materials simply and directly at any time, and since the trademark code is a string, Any one of the trademarks can be uniquely and accurately identified, so that retrieval with high stability can be performed.
- the invention not only avoids the obstacles of text input, but also can select the words that it reads, which is very convenient and more effective for international customers.
- the manufacturer may not be able to find the manufacturer accurately and conveniently, but the manufacturer of the product can be found and seen by anyone on the Internet by the invention, as long as the product is Wherever the advertisement is, you can see the trademark code at any time, and you can find the manufacturer by entering the above trademark code through the existing search engine or method.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Processing Or Creating Images (AREA)
Description
一种商标检索方法及系统 技术领域
本发明有关于模式识别技术领域,尤其是有关于图像编码技术领域,具 体地讲是一种商标检索方法及系统。 背景技术
目前,用户在需要了解一家公司或该公司产品或服务时,往往是通过以 下的方式来进行查询:
1,以文字输入法找到相关资料。
2.找到要找的地址, 还必须再转入该公司网址或打电话联系, 才能更详 细了解该公司的详情, 可能该找到的公司根本不是自己心中所耍的厂家。
3.也许进入该公司网页之后, 因语言文化差异, 自己看不懂, 或因时间 差异, 一时也无法电话联络上。
4.任何公司厂家跟相关网站有签约才能在某个网站查到他们的信息, 但 也是公司的联络方式而以。
而商标在工商业社会中起着非常重要的作用,是公司、产品或服务的标 志, 与企业的商品质量、 服务质量、 经营管理融为一体, 成为企业信誉的象 征, 是一种无形的资产。 因此, 商标是公司及其产品的一个重要属性, 具有 唯一性。
用户在了解一家公司或其产品、服务的时候除了记住该公司的名称以外 往往更容易记住该公司的商标,而且在标识某产品或服务时商标比名称更容 易让人识别。
因此,用户在想了解该公司及其该公司的所有产品或服务时,完全可以 通过检索其商标来进行查询。
然而, 通过商标来进行查询时, 由于商标一般是图标, 因此很难用一般
的检索引擎或匹配办法来进行检索。现有的技术当中存在对商标的图标进行 处理而降维, 以降维后的图像来匹配比对的检索方法。然而, 这种方法的稳 定性偏低,且编码降维的算法极其复杂不适合推广。另外这种方法归根结底 还是将图标与图像进行比对,并没有解决上述的很难用一般的检索引擎或匹 配办法来进行检索的缺点。
因此,就需要一种可以唯一的标识商标且还可以使用一般的检索引擎或 匹配办法来进行检索的编码方法, 以解决现有技术当中的缺陷。
在此, 将中国专利 ZL03152653.5"商标检索方法 "的专利文献引用于此, 作为本发明的现有技术。 发明内容
本发明的目的在于,提供一种商标检索方法及系统,从而可以以一种简 单、 快捷的方式准确的检索到所需的商标及相关信息。
为了实现上述本发明的目的, 本发明的又实施例提供一种商标检索方 法,该方法包括以下步骤:获取商标的图案信息;对所述图案信息进行编码, 生成与所述商标对应的字符串;建立并存储字符串、商标及产品信息的映射 关系; 输入字符串; 从存储的所述的映射关系中找出与所输入的字符串相 匹配的商标和 /或产品信息并输出。
为了实现上述本发明的目的, 本发明的实施例还提供一种商标检索系 统,该系统包括: 图像获取单元,用于获取商标的图案信息; 图像编码单元, 对所述图案信息进行编码, 生成与所述商标对应的字符串; 存储单元, 存储 字符串、商标及产品信息的映射关系; 输入单元, 用于输入所要检索的商标 的字符串;检索单元,从已存储的字符串中找出与所输入的字符串相匹配的 商标和 /或产品信息; 输出单元, 输出检索到的所述的商标和 /或产品信息。
本发明的有益效果:
一.在现有的世界各大网络查询任何公司地址,都必需输入相关名称或
正确的名称,而本发明是把商标转换为数字并当作文字来输入,而这些数字 的来源就是商标,就是公司厂家把它刊登在报刊或印在产品上随货品的流通 让人们认识,这样以后任何人都不会因文字差异如读不懂、或不懂怎样输入 文字等障碍, 只要记下号码, 或看到商标下的代号就可以找到厂家。
二.现有的网上操作,都是找到地址之后再转入对方的网址,才能查阅, 但往往也会出现许多网上各种操作才能真正进入该网页,浪费不少精神和时 间。而本发明是只要输入正确的商标代码, 就可直接阅读资料, 如果要找某 些行业的资料,也可以随时地简单又直接地阅读所有提供的资料,并且由于 该商标代码是字符串, 因此可以唯一地、准确地标识任何一个商标, 从而能 稳定性极高的进行检索。
三. 本发明除了免去文字输入的障碍, 更可以选择自己读懂的文字, 对 国际客户是非常方便更有效的推广产品。
在现有技术当中, 虽然知道商标, 但从产品未必能准确、方便的找到厂 家, 但通过本发明任何人都能上网查到、 看到该产品的厂家, 对厂家来说, 只要产品所到之处, 就是广告所到之处, 随时看到商标代码, 并在通过现有 的检索引擎或方法且通过输入上述商标代码就能找到厂家。 附图说明
此处所说明的附图用来提供对本发明的进一步理解,构成本申请的一部 分, 并不构成对本发明的限定。 在附图中:
图 1所示的是本发明实施例 1的图标代码转换装置的结构框图。
图 2所示的是本发明实施例 1的图标代码转换方法的流程图。
图 3A至图 31所示的是本发明的举例说明按照结构拆分的商标示意图。 图 4至图 9所示的是本发明的对几何图和线条进行区别定号时的几何 图、 线条的示意图。
图 10A至图 10L所示的是本发明的举例说明编码方法的商标示意图。
图 11所示的是本发明的实施例 2的商标检索系统的结构框图 图 12所示的是本发明实施例 2的商标检索方法的流程图。 具体实施方式
为使本发明的目的、技术方案和优点更加清楚明白,下面结合实施方式 和附图, 对本发明做进一步详细说明。在此, 本发明的示意性实施方式及其 说明用于解释本发明, 但并不作为对本发明的限定。
实施例 1
本发明的实施例 1提供一种图标代码转换方法及装置,以通过简单的方 式唯一的表示图标, 在本发明实施例 1中的图标为商标。
图 1所示的是本发明实施例 1的图标代码转换装置的结构框图,其中该 图标代码转换装置包括: 图像获取单元 101, 用于获取所述图标的图像; 预 处理单元 102, 用于对所述图像进行预处理, 抽取多个所述图像的具有可分 性的特征; 代码化单元 103, 用于将多个所述具有可分性的特征以具有顺序 的字符串 (该字符串包括数字、 英文字母和标点符号)进行表示。 并且, 上 述预处理单元 102包括: 外形分类单元 104, 用于按照外形对所述图像进行 分类,并将该分类结果作为一个所述具有可分性的特征;结构分类单元 105, 用于按照从上到下、从左到右、从外到里的顺序对所述图像进行拆分, 并对 每个拆分出的结构进行分类,将每个该分类结果作为一个所述具有可分性的 特征。
图 2所示的是本发明实施例 1的图标代码转换方法的流程图,以下结合 图 2详细说明图标代码转换方法。
首先, 获取上述商标的图像 (步骤 S201 ) ; 接着, 按照外形对所述图 像进行分类,并将该分类结果作为一个所述具有可分性的特征(步骤 S202) ; 以从上到下、从左到右、从外到里的顺序对所述图像进行拆分, 并对每个拆
分出的结构进行分类, 将每个该分类结果作为一个所述具有可分性的特征
(步骤 S203) ; 将上述每个具有可分性的特征以字符来进行标识, 即以数 字、 英文字母和标点符号标识上述的每个具有可分性的特征(步骤 S204) ; 最后,将标识上述具有可分性的特征的数字、英文字母和标点符号组合排列, 以带顺序的字符串表示上述商标 (步骤 S205) 。
在本实施例 1中对图像进行分解的顺序是以从上到下、从左到右、从外 到里的顺序, 如图 3A至 31所示。
以下,举例说明本发明的对图像编码的方法,但本发明的编码方法不限 于此。
1.首先, 对商标的图像进行分类, 在本发明中将该分类的种类分成 5大 类形:
(1)单一个几何图的外形叫正规图。
(2)二个以上的几何图组合成商标外形叫非正规图。
(3)正规图内有动植物或其它特殊图形的称特殊正规图。
(4)商标外形图是动植物或文字等相混合有几何图或线条称特殊非正规 图。
(5)单独一个动植物、 文字、 实物等作为商标外形图称特殊图。
然后, 在根据图像的结构进行拆分, 拆图的规则和定号方法如下: 每个商标的图像首先要定外形号, 分清它是属于那一种类形的图; 定了外形号之后, 按照图的结构一般可分左右上下, 或者左中右, 或者 上中下, 如图是圆形图就分内、 内中、 中心点或者外、 内、 中心点。 定号的 规则是先左后右再定上后下;
非正规图是上方有分左右, 下方也有分左右, 那只能定上方的左边, 后 定下方的左右两边;
后定下方的左和右。
(1)上述正规图包括一般常见的几何图形:如四方形、 长方形、 三角形、 多边形、 园形等几何图形, 只要是单一个几何图作为商标图的外表形状。
正规图是以如下方式定号:
正规图的外形一定是几何图形,先定了外形号,再定图内的线条或其它 图案代号, 图内可分左右上下等, 按规则定号, 称内形号。 已被定内形号的 图如图中还有图不可再重复定号。如内形号可定左右上下四部分,那么中间 部分不用定号。
(2)上述非正规图是二个以上的几何图分开或者重迭组成一个商标的外 表图。
非正规图是以如下方式定号:
首先,非正规图一定是由二个几何图组成的,它的外形号只能用英文字 母 C或 E来代表, 定外形号之后要判断这是左右分开或者是两个上下组成 的。 如图是上下二部分组成, 其中一部分又分左右 2个小图, 必须是各自独 立分开的才可定号。
(3)上述特殊正规图的外形是单一的几何图形, 图内可能是混有动植物 文字等特殊图, 而不是线条或小的几何图构成。
特殊正规图是以如下方式定号:
因外形是几何图, 图内是其它特殊图, 定外形号就先定几何图代号后定 特殊图用英文字母代号,这二个号当作外形号,而内形号按拆图规则分左右 或者上下等因特殊图不用再拆图,只用特殊图的代号,如有文字只取最前面 每个发单的字母, 或者按字母的排列秩序打出号码, 如 A的顺序是 1、 C 是排第 3、 D是排第 4。
(4)特殊非正规图的商标图像的组合是二种以上不相同的图组成, 有几 何图混合动植物特殊图, 外形号只能用 D作代号, 表示困难复杂。
特殊非正规图是以如下方式定号:
这种图案一定有人物动植物等混合有何图或各种线条等。除了给外表定 号为 D, 然后按照图的主体来分, 如主体在中间左右是线条或其它, 按秩序 是左中右来定号, 如图是分上、 中下就先上中后下。
(5)凡是人物卡通图、 动植物、 实物、 文字等单独做为商标图这通称特 殊图, 外形代号是英文字母 A、 B、 C、 D、 E。
特殊图是以如下方式定号:
一定要按照主体的头部作为定号方向,如头在上就由上到下来定号,如 头在左边或右边,就由先左后右来定号。而且只能取图内部主要线条或图来 定号, 既内形号, 因外形号已经定为特殊图代号。
2.商标内的几何图和线条的区别定号
规划之后给予固定代号, 用数字 1、 2、 3、 4、 5, 6、 7、 8、 9、 0 作代号。
另外 A、 B、 C、 D、 E代表特殊图案的代号, 还有" +"这符号和 ".", 这符号代表增加或者付来区分号码。
(1) 1号的定位
一画过, 无论是横、 直、 斜、 撇、 捺、 半湾月形, 细扁的长方形两头细 尖、 虫形、 总之图的线条一定要细而且是单一又独立。如一划过的月湾形线 条不能有两条相同又相连接, 或两头反方向相连在一起 (如图 4所示) 。
(2) 2号的定位
除了两个圆形重叠不可定为 2号,其它一切几何图形都必须是二个以上 相同图形重叠而成, 这两个重叠图必须是头对头, 或者头对底, 或者是星星 图几个角互相连接 (如图 5A所示) 。
注意各种几何图的中间线条产生的错觉图案:任何几何图中间如果有一 条直线横线曲线等,都能改变整个图的错觉一分为二的图,这些能改变图内
部的线条规划如图 5B所示。 这些图都不能因中间的线条而被当作二个图或 分成左右上下来定号, 只能取外形号之后再定图内的线条代号作为内形号。
(3) 3号的定位
一切的曲线,相等和不相等的曲线, 由曲线构成的图案这些曲线必须有 二个曲角以上或者二个湾以上, 必须是相连接的一条曲线 (如图 6所示) 。
(4) 4号的定位
一切正规几何图形和多边不规则图形,只要是三边或四边多边等组成的 几何图是封口的, 而且有角的, 圆形和半圆形不包括在 4号定位之内 (如图 7所示) 。
(5) 5号的定位
凡有角或半湾圆形或交叉形,这一切因形不可封口,看去变成一个独立 的外表、 图形, 必须是呈线条状, 不可能呈整体或立体状 (如图 8所示) 。
(6) 6号的定位
各国文字、 字母, 如 、 B、 C、 中文字、 阿拉伯文字等, 有时因这些 文字是坠属在外形图内, 要付加十符号一起之后才输入文字。
(7) 7号的定位
各种大小飞禽走兽昆虫鱼虾动生物及卡通图,如这些图是附属在外图内 就要加上十符号在前面。
(8) 8号的定位
单一个圆形图, 或两个重叠的圆形, 或两个相连的圆形, 或圆形边有一 直线或椭圆形等。其中, 圆形图内中间有一条横线或曲线交叉线分隔的不能 当作 2个不同图来定号。因是在独立的图内,这些线条跟 2号的定位图一样, 只能当作线条不能当作两个或多个图案来定号 (如图 9所示) 。
(9) 9号的定位
一切物品、 用具、 工具、 实体物。 如飞机、 汽车、 建筑物等, 在特殊情
况下要先加上 +符号一起使用, 例如 + 9。
(10) 0的定位
一切花草树木果实、 植物、 谷类、 食品等。
(11) "+"的符号定位
凡遇到 6、 7、 9、 0这些特殊困的时候, 如它们是付层在外形图的内部, 我们就要用 +这符号。 如特殊图是独立的外形图, 可能要分折成几大部分再 定号,而其中某部分有文字付属在一个几何图内时也要加上 +这符号来区别。 另外, 如果外形是常规几何图, 图内单独只有文字, 可以不用加上 +这符号, 可直接打出外形号和 6号之后打上文字。如果非常规图只能分左右, 而中间 没代号, 就用 +当代号, 或者那图案只有外形号, 没分左右上下, 只有中间 有代号, 就要用 +作连接, 如外形图代号 +中间内的代号。
(12) "."的用途定位和 L字母相配合
如果遇到文字图案,我们不想用文字输入法,就只能用数字来代表文字 字母的顺序排位, 那么从 10以上, 在数字后面加上。 等於区别数字之间的 号码, 如 1、 10、 9、 23就如小数点一样的胜任, 但必须在前面先输入 L 字母 L、 1、 10、 9、 23
(13)字母 A、 B、 C、 D、 E、 作为特殊代号的定位用於外形号代号:
A、 代表一切有生命的动植物, 主要用在外形图定号。
B、 代表几何图和特殊图二种混组成的非正规特殊图外形号代表。
C、 由二个以上的图形或半圆形图组成的外形图代号。
D、 看去很难辨清是动物或其它图分不清的外形代号。
E、 两个以上有角的不同几何图组成的非正规图外形号代号。
以下, 举具体的例子说明本发明的编码方法。
图 10A:
外形是圆形, 内部呈上下分部分, 上方是小圆形下是 T字形。 因此, 其
外形号: 8; 上: 8; 下: 1, 全图定号: 881。
图 10B:
外形是圆形内部可分上下左右中间可以不理。先左后右再上后下,下方 是 M线条。 因此, 其外形号: 8; 内左: 1 ; 内右: 1 ; 内上: 4; 内下: 3, 全图定号: 81143。
图 10C:
外形是四方表正规图, 内部是特殊图卡通人物, 因左右各自分开, 中间 也有图, 内部分先右后右, 中间分上下。 因此, 其外形号: 4A; 左: 7; 右: 7; 上: 3; 下: 8, 全图定号: 4A7738。
图 10D:
外形是四方形正规图, 内部上方是人相,下方是文字。因此,其外形号: 4A; 上: 7; 下: 6LEF, 全图定号: 4A76LEF。
图廳:
外形由二个分开的几何图构成,不能同时给二个图一个代号,只好用英 文字母作代号。 这二个图也就变成了内形号, 分左右二部分, 要用" +"这符 号把它连接。 因此, 其外形号: E; 左: 4; 右: +4, 全图定号: E4+4。
图 10F:
商标是卡通动画图, 图的头部在右边,凡这类图的定号就必先左后右因 外形号是特别代号,定了外形号之后就不能再重复定号,只能把图内的线条 作内形号来定号, 右边内部是眼。 因此, 其外形号: A7; 内左: 1 ; 内右: 8, 全图定号: A718。
图 10G:
跟上图一样, 差别在中间有比较明显的分隔线, 必须定号, 而内形号是 一样先左、 中、 右。 因此, 其外形号: A7; 内左: 3; 中: 5; 内右: 8, 全 图定号: A7358。
图 10H:
跟上图一样,差别在左边没有代号,右边下方才有代号,定号方法不变, 先左后右。 因此, 其外形号: A7; 左: (没有图不用写代号); 右: 8, 全图 代号: A78。
图 101:
由二个几何图组成, 只能用英文字母作代号, 分上下定号。 因此, 其外 形号: CE; 上: 1 ; 下: 4, 全图定号: CE14。
图 10J:
这是正规图, 内部可分左右上下。 因此, 其外形号: 8; 左: 1 ; 右: 3; 上: 3; 下: 1, 全图定号: 81331。
图 10K:
外形号只能用英文母作代号, 因内部是圆形, 所以没分左、 右只分内、 中心点等。 因此, 其外形号: D; 夕卜: 5; 内: 8; 中心点: 1, 全图定号: D581。
图 10L:
跟上图一样, 外形号是 D, 内形号分上中下。 因下方图内有文字, 有二 种定号法, 一种是 +6, 另一种是 +6后把第一读音字母写出来。 因此, 其外 形号: D; 上: 3; 中: 8; 下: +6 (M是第一发单字母), 全图定号: D38+6 或 D38+6MM。
实施例 2
本发明的实施例 2提供一种商标检索方法及系统。
图 11所示的是本发明的实施例 2的商标检索系统的结构框图, 其中该 商标检索系统包括: 图像获取单元 1101, 用于获取所述商标的图像; 图像 编码单元 1102, 用于对所述图像进行编码, 以具有顺序的字符串表示所述 图像; 存储单元 1103, 用于存储所述商标的图像及其字符串; 代码输入单
元 1104, 用于输入所要检索的商标图像的代码; 检索单元 1105, 用于从已 存储的字符串中找出与所输入的代码相匹配的字符串,显示与该字符串相应 的商标的图像。 其中, 所述编码单元 1102包括: 预处理单元 1106, 用于对 所述图像进行预处理,抽取多个所述图像的具有可分性的特征;将多个所述 具有可分性的特征以具有顺序的字符串进行表示。
另外, 本实施例 2的预处理单元 1106包括: 外形分类单元 1107, 用于 按照外形对所述图像进行分类,并将该分类结果作为一个所述具有可分性的 特征; 结构分类单元 1108, 用于按照结构对所述图像进行拆分, 并对每个 拆分出的结构进行分类, 将每个该分类结果作为一个所述具有可分性的特 征。
图 12所示的是本发明实施例 2的商标检索方法的流程图, 以下结合图 12详细说明商标检索方法。
首先, 获取上述商标的图像(步骤 S1201 ) ; 接着, 按照外形对所述图 像进行分类, 并将该分类结果作为一个所述具有可分性的特征 (步骤 S1202) ; 以从上到下、 从左到右、 从外到里的顺序对所述图像进行拆分, 并对每个拆分出的结构进行分类,将每个该分类结果作为一个所述具有可分 性的特征(步骤 S1203 );将上述每个具有可分性的特征以字符来进行标识, 即以数字、 英文字母和标点符号标识上述的每个具有可分性的特征 (步骤 S1204) ; 最后, 将标识上述具有可分性的特征的数字、 英文字母和标点符 号组合排列, 以带顺序的字符串表示上述商标(步骤 S1205) ; 将上述商标 和字符串进行存储 (步骤 S1206);输入所要检索的商标的代码 (步骤 S1207); 根据该所输入的代码从所存储的字符串中检索匹配的字符串,并检索出与该 字符串相应的商标 (步骤 S1208 ) 。
本发明具有如下效果:
一.在现有的世界各大网络查询任何公司地址,都必需输入相关名称或
正确的名称, 而本发明是把数字当作文字来输入, 而这些数字代号的来源, 就是公司厂家把它刊登在报刊或印在产品上随货品的流通让人们认识,这样 以后任何人都不会因文字差异如读不懂、或不懂怎样输入文字等障碍,只要 记下号码, 或看到商标下的代号就可以找到厂家。
二.现有的网上操作,都是找到地址之后再转入对方的网址,才能查阅, 但往往也会出现许多网上各种操作才能真正进入该网页,浪费不少精神和时 间。而本发明是只要输入正确的商标代码, 就可直接阅读资料, 如果要找某 些行业的资料,也可以随时地简单又直接地阅读所有提供的资料,并且由于 该商标代码是字符串, 因此可以唯一地、准确地标识任何一个商标, 从而能 稳定性极高的进行检索。
三. 本发明除了免去文字输入的障碍, 更可以选择自己读懂的文字, 对 国际客户是非常方便更有效的推广产品。
在现有技术当中, 虽然知道商标, 但从产品未必能准确、方便的找到厂 家, 但通过本发明任何人都能上网查到、 看到该产品的厂家, 对厂家来说, 只要产品所到之处, 就是广告所到之处, 随时看到商标代码, 并在通过现有 的检索引擎或方法且通过输入上述商标代码就能找到厂家。
以上所述的具体实施方式,对本发明的目的、技术方案和有益效果进行 了进一步详细说明,所应理解的是, 以上所述仅为本发明的具体实施方式而 已, 并不用于限定本发明的保护范围, 凡在本发明的精神和原则之内, 所做 的任何修改、 等同替换、 改进等, 均应包含在本发明的保护范围之内。
Claims
1. 一种商标检索方法, 其特征在于, 该方法包括以下步骤:
获取商标的图案信息;
对所述图案信息进行编码, 生成与所述商标对应的字符串;
建立并存储字符串、 商标及产品信息的映射关系;
输入字符串;
从存储的所述的映射关系中找出与所输入的字符串相匹配的商标和 /或 产品信息并输出。
2. 根据权利要求 1所述的商标检索方法, 其特征在于, 所述对图案信 息进行编码包括:
对所述图案信息进行预处理,抽取多个所述图案信息的具有可分性的特 征;
将多个所述具有可分性的特征以具有顺序的字符串进行表示。
3. 根据权利要求 2所述的商标检索方法, 其特征在于, 所述预处理包 括:
按照外形对所述图案信息进行分类,并将该分类结果作为一个所述具有 可分性的特征;
按照结构对所述图案信息进行拆分, 并对每个拆分出的结构进行分类, 将每个该分类结果作为一个所述具有可分性的特征。
4. 根据权利要求 3所述的商标检索方法, 其特征在于, 所述按照结构 对图案信息进行拆分是以从上到下、 从左到右、 从外到里的顺序进行拆分。
5. 根据权利要求 1所述的商标检索方法, 其特征在于, 所述字符串包 括: 数字、 英文字母和标点符号。
6. 根据权利要求 1所述的商标检索方法, 其特征在于, 所述建立并存
储字符串、 商标及产品信息的映射关系包括:
在产品上附上所述字符串。
7. —种商标检索系统, 其特征在于, 该系统包括:
图像获取单元, 用于获取商标的图案信息;
图像编码单元,对所述图案信息进行编码, 生成与所述商标对应的字符 串;
存储单元, 存储字符串、 商标及产品信息的映射关系;
输入单元, 用于输入所要检索的商标的字符串;
检索单元,从存储的所述的映射关系中找出与所输入的字符串相匹配的 商标和 /或产品信息;
输出单元, 输出检索到的所述的商标和 /或产品信息。
8. 根据权利要求 7所述的商标检索系统, 其特征在于, 所述编码单元 包括:
预处理单元,用于对所述图案信息进行预处理,抽取多个所述图案信息 的具有可分性的特征;
将多个所述具有可分性的特征以具有顺序的字符串进行表示。
9. 根据权利要求 8所述的商标检索系统, 其特征在于, 所述预处理单 元包括:
外形分类单元,用于按照外形对所述图案信息进行分类,并将该分类结 果作为一个所述具有可分性的特征;
结构分类单元,用于按照结构对所述图案信息进行拆分,并对每个拆分 出的结构进行分类, 将每个该分类结果作为一个所述具有可分性的特征。
10. 根据权利要求 9所述的商标检索系统, 其特征在于, 所述按照结构 对图案信息进行拆分是以从上到下、 从左到右、 从外到里的顺序进行拆分。
1 1 . 根据权利要求 7所述的商标检索系统, 其特征在于, 所述字符串包
括: 数字、 英文字母和标点符号。
12. 根据权利要求 7所述的商标检索系统, 其特征在于, 所述建立并存 储字符串、 商标及产品信息的映射关系是指, 在产品上附上所述字符串。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2008/071581 WO2010003286A1 (zh) | 2008-07-08 | 2008-07-08 | 一种商标检索方法及系统 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2008/071581 WO2010003286A1 (zh) | 2008-07-08 | 2008-07-08 | 一种商标检索方法及系统 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2010003286A1 true WO2010003286A1 (zh) | 2010-01-14 |
Family
ID=41506654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2008/071581 Ceased WO2010003286A1 (zh) | 2008-07-08 | 2008-07-08 | 一种商标检索方法及系统 |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2010003286A1 (zh) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| USD819066S1 (en) | 2016-05-05 | 2018-05-29 | Corsearch, Inc. | Display panel with a graphical user interface |
| US10437845B2 (en) | 2016-05-05 | 2019-10-08 | Corsearch, Inc. | System and method for displaying search results for a trademark query in an interactive graphical representation |
| US10942973B2 (en) | 2017-05-20 | 2021-03-09 | Corsearch, Inc. | Automatically generating and evaluating candidate terms for trademark clearance |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1458613A (zh) * | 2003-06-01 | 2003-11-26 | 王辉 | 人工智能商品防伪技术 |
| CN1581159A (zh) * | 2003-08-04 | 2005-02-16 | 中国科学院自动化研究所 | 商标检索方法 |
| WO2007072947A1 (ja) * | 2005-12-22 | 2007-06-28 | Matsushita Electric Industrial Co., Ltd. | 画像検索装置および画像検索方法 |
-
2008
- 2008-07-08 WO PCT/CN2008/071581 patent/WO2010003286A1/zh not_active Ceased
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1458613A (zh) * | 2003-06-01 | 2003-11-26 | 王辉 | 人工智能商品防伪技术 |
| CN1581159A (zh) * | 2003-08-04 | 2005-02-16 | 中国科学院自动化研究所 | 商标检索方法 |
| WO2007072947A1 (ja) * | 2005-12-22 | 2007-06-28 | Matsushita Electric Industrial Co., Ltd. | 画像検索装置および画像検索方法 |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| USD819066S1 (en) | 2016-05-05 | 2018-05-29 | Corsearch, Inc. | Display panel with a graphical user interface |
| USD852821S1 (en) | 2016-05-05 | 2019-07-02 | Corsearch, Inc. | Portion of display panel with a graphical user interface |
| US10437845B2 (en) | 2016-05-05 | 2019-10-08 | Corsearch, Inc. | System and method for displaying search results for a trademark query in an interactive graphical representation |
| US10942973B2 (en) | 2017-05-20 | 2021-03-09 | Corsearch, Inc. | Automatically generating and evaluating candidate terms for trademark clearance |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102770862B (zh) | 混合使用位置传感器数据和视觉查询来返回视觉查询的本地收录 | |
| CN102822817B (zh) | 用于视觉查询的可采取动作的搜索结果 | |
| CN108431801B (zh) | 为与用户关联的图像生成标签 | |
| CN103617422B (zh) | 一种基于名片识别的社交关系管理方法 | |
| KR101226751B1 (ko) | 디지털기기로 해독되는 정보를 포함한 디자인코드 패턴 및 그 운영시스템 | |
| US20110035406A1 (en) | User Interface for Presenting Search Results for Multiple Regions of a Visual Query | |
| KR20180041202A (ko) | 사용자와 연관된 이미지들의 조직화 | |
| CN102667764A (zh) | 用于为视觉查询的多个区域展示搜索结果的用户接口 | |
| JP2013501975A (ja) | 視覚クエリに応答するためのアーキテクチャ | |
| CN106097794A (zh) | 基于增强现实的汉语拼音组合认读学习系统及认读方法 | |
| Foka et al. | AI, Cultural Heritage, and Bias: Some Key Queries That Arise from the Use of GenAI | |
| CN109857912A (zh) | 一种字形识别方法、电子设备及存储介质 | |
| WO2010003286A1 (zh) | 一种商标检索方法及系统 | |
| Benefiel et al. | Documenting Ancient Graffiti: Text, Image, Support and Access | |
| CN111078724A (zh) | 学习系统中的测试题检索方法、装置、设备及存储介质 | |
| Majid et al. | Digitization of Handwritten Chess Scoresheets with a BiLSTM Network | |
| CN110046273B (zh) | 一种基于绘图的书籍查询推荐方法、电子设备及存储介质 | |
| CN106934336A (zh) | 一种幻灯片识别的方法及装置 | |
| CN110414497A (zh) | 对象电子化的方法、装置、服务器及存储介质 | |
| Benefiel et al. | Images and text on the walls of Herculaneum: Designing the Ancient Graffiti Project | |
| AU2014200923B2 (en) | Hybrid use of location sensor data and visual query to return local listings for visual query | |
| US10776656B2 (en) | Methods and systems for applying content aware stickers onto a layout | |
| CN114580420A (zh) | 一种中文实体识别方法、装置、设备及介质 | |
| Benefiel et al. | Wall Inscriptions in the Ancient City: The Ancient Graffiti Project | |
| JP5263825B2 (ja) | 情報クリッピング装置、情報クリッピング方法、および、情報クリッピングプログラム |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 08773135 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 08773135 Country of ref document: EP Kind code of ref document: A1 |