201042473 _) jLz^jLWi.doc/n 六、發明說明： 【發明所屬之技術領域】 本發明是有關於一種資料搜尋方法與搜尋器，且特 是有關於一種以輸入影像為查詢標的，進行影像搜尋方法 以及以輸入影像為查詢標的之影像搜尋器。 【先前技術】 〇 纟#訊高度擴充的現在，人們越來越仰賴網際網路所 提供的快速資料搜尋服務。以往，使用者想要於網際網路 之中進行資料檢索時，必須要至特定的入口網站，例如： 奇摩、谷歌(Google)、蕃薯藤，透過鍵盤、滑鼠或是觸控 式螢幕等輸出入裝置，鍵入關鍵字來搜尋相關網頁。也& 是目前熟知的以關鍵文字尋找文字或是圖案資料的常用資 料搜尋模式。 ' 雖然目前谷歌(Google)搜尋引擎開發出相似圖像搜尋 功能，但是其使用方法也是必須藉由輸入查詢關鍵文字， 再經由使用者瀏覽所搜尋到的相關圖片相片，點選特定圖 片相片的“相似影像”功能，才會經由特定圖片連結出所有 •相關的影像。也就是所謂的相似圖案搜尋，也必須經由輸 -入關鍵文字開始，獲得與關鍵文字相關的圖案之後，再拓 展成相似圖案搜哥結果。換句話說，使用者無法經由直接 輸入所欲查詢的影像，達到圖像搜尋圖像或是圖像搜尋文 字的目的。 此外’由於資料搜尋的資料母群過於魔大，對於以圖 3 201042473 j 丄;ziMtwi.aoc/n 像相互比對搜尋圖像而言’需钱大而精密的演算法支 持，才得轉得精確_像_結果，啸尋結果的影像 集合=會更精確的符合制者所要。因此在開發直接以圖 像搜尋圖像或是以目像搜尋文字的搜尋器上，具有軟體程 式撰寫的高複雜度與硬體高速運算支援的難度極需克服。 【發明内容】 本毛月提t、種料搜尋方法，藉由所查詢影像的相關 地理定位資訊，縮小資料掬暮笳 々 貝针孜+乾圍，間化以圖像找圖像的 本發明提供—種影雜料，根據所查詢影像的地理 疋位貝H小資料搜尋範圍，並且 候選資料的影像特徵，快速筛選出影像搜尋結广^ 本發明提出-種資料搜尋方法 =詢資訊，該影像查詢資訊至少包括一：二 目關的一地理定位資訊。根據所接收的地理 資料：二一網際網路中1集複數筆候選資料成-候選 於候選資赫 =筆闕㈣包含至少—候選影像。之後， 結果，對比錢影像與候選影像，以決定一搜尋 ° ，、搜哥結果至少包括候選資料苴中之一 該地Ϊ，: —實施例中，上述之資料=方法，其中 二位置二維條碼、- 包括快逮反應編碼一_二 201042473 d i-dV3iwi.doc/n 明之—實施射，上述之料搜尋方法，其中對 詢影像與該些候選影像的方法，包括對彡 中餘= 像特徵，以由該些候選影像 ^ 喊像最相似的部份該些候選影像。而會 ff徵包括影像顏色、影像_、影像對比訊號與影像ς 寿Ρ 〇 在本發明之-實施例中，上述之資料搜尋方法 〇 括顯示該搜尋結果中每一該些候選資料中的-文字資料。 在本發明之-實施例中，上述之資料搜尋方法，還包 括顯不該搜尋結果令每一該些候選資料的一文字資料與至 少一候選影像。 在本發明之一實施例中，上述之資料搜尋方法，其中 該候選資料庫中的每—筆該些候選資料包括一位置座標， 且該些候選資料的該些位置座標均位於—座標範圍内。 在，發明之一實施例中，上述之資料搜尋方法，其中 帛座標關包括以該地理^位資訊所對應m座標為 中心，距離該定位座標一百公尺以内的範圍。 ^發明另提出-種影像搜尋器’包括：—查詢模組、 -搜尋模組以及-影像對比模組。其中，查詢模組用以接 收-影像查詢資訊，而影像查詢資訊至少包括一查詢影像 以及與該查詢影像相關的一地理定位資訊。搜尋模组，則 =以根據地理定位資訊，於一網際網路中蒐集複數筆候選 貝料成一候選資料庫，其中每一筆該些候選資料包含至少 一候選影像。影像對比模組，是用於對比該查詢影像與該 201042473 ^ i^vjtwT.aoc/n 些候選e像’ a决定—搜尋 括該些候選賢轉其尹之—。 甲4搜+結果至少包 在本發明之—貧施 紐定位資訊包括—經緯度座標、一其中該 球定位資訊舆-輔助全球定位資訊。dr全 括快速反應編碼(咖如SP〇騰code，QRc_)一維條石馬包 象ίίΪ:月之一實施例中，上述之影像搜尋器，其” 對比該查詢影像與該些候選影像還包括^ 以由遠些候選影像中篩選出盥专戊 些候選影像。又，影像特徵包括影像==份？ 像對比訊號與影像輪#。 化4構圖、影 —海明m例中’上述之影像搜尋器’還包括 的-文字資料 搜哥結果中母一該些候選資料中 在树明之一實施例中，上述之影像搜尋器，還包括 -文=二用於顯不該搜尋結果中每一該些候選資料的 文子貝料與至少一候選影像。 本發明之一實施例中’上述之影像搜尋器，其中該 座標範圍與該地理定位資訊，所荒集到的 μ座ί3Γ母—筆該些候選資料的一位置座標均位 旛。X座私靶圍内。而座標範圍包括以該地理定位資訊所對 二的-定位座標為中心，距離該定位座標一百公尺以内的 範圍。 ❹201042473 _) jLz^jLWi.doc/n VI. Description of the Invention: [Technical Field] The present invention relates to a data search method and a searcher, and particularly relates to an image using an input image as a query target Search method and image searcher with input image as the query target. [Prior Art] 〇 纟 讯 讯 讯 讯 讯 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度 高度In the past, when users want to search data on the Internet, they must go to specific portals, such as: Chimo, Google, and potato, through keyboard, mouse, or touch screen. Import and export devices, type keywords to search for related web pages. Also & is a commonly used common search mode for finding text or pattern data in key words. Although Google's Google search engine has developed a similar image search function, its use method must also be based on the input of the query key text, and then through the user to browse the relevant photo photos, click on the specific photo photo. The "similar image" function will link all relevant images through a specific image. The so-called similar pattern search must also start with the input of key words, obtain the pattern related to the key words, and then expand into a similar pattern search results. In other words, the user cannot directly input the image to be queried to achieve the purpose of searching for an image or searching for a text. In addition, because the data of the data search group is too big, for the picture 3 201042473 j 丄; ziMtwi.aoc / n like the mutual comparison of the search for images, the need for large and sophisticated algorithm support, only to turn Accurate _ like _ results, the image collection of the whistling results = will be more accurate to meet the requirements of the system. Therefore, in developing a searcher that directly searches for images by image or searches for text by visual image, the difficulty of high-complexity and hardware high-speed computing support with software programming is extremely difficult to overcome. [Summary of the Invention] The present invention provides a method for searching for a gross image, and by narrowing the information about the geolocation information of the image to be inquired, reducing the data, acupuncture and acupuncture, and interpolating the image to find an image. Providing a kind of groceries, according to the geographical position of the image to be queried, and the image search feature of the candidate data, and the image feature of the candidate data, quickly screening out the image search sufficiency ^ The present invention proposes a data search method = inquiry information, The image query information includes at least one: a geolocation information of two eyes. According to the received geographic data: a set of multiple candidate data in the Internet is a candidate - candidate for the candidate = pen (4) contains at least - candidate images. Afterwards, the result is a comparison between the money image and the candidate image to determine a search, and the search result includes at least one of the candidate data: - in the embodiment, the above data = method, wherein the two positions are two-dimensional Barcode, - including fast catch response code _ 2 201042473 d i-dV3iwi.doc / n - the implementation of the above, the material search method, the method of querying the image and the candidate images, including the 彡 余 = = The feature is to image the candidate images by the most similar portions of the candidate images. And the ff sign includes the image color, the image_, the image contrast signal, and the image. In the embodiment of the present invention, the data searching method includes displaying the candidate data in the search result. text document. In the embodiment of the present invention, the data searching method further includes displaying, by the search result, a text data of each of the candidate materials and at least one candidate image. In an embodiment of the present invention, the data search method, wherein each of the candidate data in the candidate database includes a location coordinate, and the location coordinates of the candidate data are all within the coordinate range . In one embodiment of the invention, the data searching method, wherein the 标 coordinate mark comprises a range centered on the m coordinate corresponding to the geographic information, and within a range of 100 meters from the positioning coordinate. The invention further proposes that the image searcher includes: - an inquiry module, a search module, and an image comparison module. The query module is configured to receive image query information, and the image query information includes at least one query image and a geolocation information associated with the query image. The search module is configured to collect a plurality of candidate candidates into a candidate database in an internet according to the geolocation information, wherein each of the candidate materials includes at least one candidate image. The image comparison module is used to compare the query image with the candidate image of the 201042473 ^ i^vjtwT.aoc/n 'a decision--search for the candidates to turn their Yin--. A4 search + results are at least included in the present invention - the poor position information includes: latitude and longitude coordinates, one of which is the positioning information of the ball - assisted global positioning information. Dr all-inclusive fast response coding (cafe such as SP〇腾code, QRc_) one-dimensional stone horse bag icon ίίΪ: In one embodiment of the month, the above image searcher, "contrast the query image with the candidate images Including ^ to select the candidate images from the far-off candidate images. Moreover, the image features include image == share? Like contrast signal and image wheel #. 4 composition, shadow - Haiming m in the case of the above The image searcher further includes a text data search result in the parent data. In one of the candidate materials, in the embodiment of the tree, the image searcher further includes - text = two for displaying the search result. The image searcher of the candidate material and the at least one candidate image. In an embodiment of the present invention, the image searcher, wherein the coordinate range and the geographic location information are collected, the μ ί Γ Γ — — — The position coordinates of the candidate data are all located in the X-seat private target. The coordinate range includes the center of the positioning coordinate of the geolocation information, and the range within 100 meters of the positioning coordinate. ❹
201042473 jizujiwi.doc/n 小搜尋，提高理定位資訊缩 簡易程度與搜尋結果的正圖像搜尋文字的 景物或文物的當下或是事後，养 ^以在觀賞實際 便於使用者立即規劃行程“ @物件資訊， 富其旅遊觀戶或是簡後記升使財轉魏度或是豐 為讓本發明之上述特徵和優點 舉實施例，並配合所關式作詳細說明如;、純’下文待 【實施方式】 S。^ ίΐΓ之一實施例的一種資料搜尋方法流程簡 夫日1 lit發明之—貫施例的—種影像搜尋器簡圖。請 、口 於步驟S101中，首先經由影像指聋哭 200的一查詢模組2〇2，接 σ 绐次却s丨a, 侪叹衫像查珣貧訊。上述影像查 理二括一查詢影像以及與該查詢影像相關的-地 貝訊。而上述影像搜尋器例如是-影像搜尋引 一圖Μ是依照本發明之一實施例的一種影像輸入方法 示心圖胡參照圖3Α，上述的查詢模組2〇2例如是圖3Α 中所,，、、員示@使用者介面·，其包括可上傳所欲查詢的 圖片或是影像的影像輸入攔位302，以及供使用者填入與 查5旬圖片或是影像相關的地理定位資訊的定位資訊輸入攔 位304甚至還包括一查詢按鍵可供使用者在上傳查 询圖片或影像以及輸入地理定位資訊之後，啟動查詢流 7 201042473 3i^U3twr.aoc/n 程。於本實施例中，是以一世界文化遺產，印度的泰姬瑪 哈陵(The mausoleum of the Taj Mahal)之影像 308 作為上傳 的查詢影像’而與此泰姬瑪哈陵影像308相關的地理定位 資訊包括泰姬瑪哈陵所在地的緯度為27.174604，而經度 為78.041814。使用者可藉由主動上傳影像3〇8，並且同時 輪入與所上傳查詢影像相關的地理定位資訊31〇,完成影 像查詢資訊輸入步驟(亦即是影像查詢器2〇〇完成接收影 像查詢資訊步驟S101)。於此實施例中，使用者在取得影 像後，格配額外輸入影像相關地理定位資訊完成影像資料 搜尋方法的步驟S101。然而，本發明並不受限於此。 一圖3B是依照本發明之另一實施例的一種影像輸入方 ^示意圖。請參照圖3B，使用者於拍攝景物取得影像的同 a守，即可同時取得所拍攝影像的相關地理定位資訊，並與 影像結合成為影像的摘要資訊之一。於此實施例中，仍舊 以拍攝實際景物312為泰姬瑪哈陵為例，當使 有地理定位裝㈣可攜絲置314拍攝泰姬瑪哈陵3 = 同時，可攜式裝置3M❸也理定位裝置同時 地較位資料，並且將影像爾與同步獲:= 理疋位貧料組成的影像查詢資訊輸入安裝於可攜 =中的#像搜尋H 完成上述#料搜尋方法的^驟 於此實施例中’可攜式裝置314例如是相機呈 文位影像功能的通訊裝置。而可攜式 的理 位裝置例如越衛絲置316通訊，可提供=== 201042473 j iz.u〇Lwi.doc/n 的全球定位系統(Global Positioning System，GPS)或是可提 供輔助全球定位資訊的輔助全球定位系統(Assisted G1〇bal42 ji 2010 2010 42 Information, rich in tourism, or post-sales, etc., to make the above-mentioned features and advantages of the present invention, and to elaborate on the relevant aspects, such as; The method of S. ^ ΐΓ 实施 实施 实施 实施 实施 实施 实施 实施 实施 1 1 1 lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit lit A query module of 2200 is connected to σ 绐 却 却 , , , , , , , , , 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 For example, the image search device is an image input method according to an embodiment of the present invention. The query module 2〇2 is, for example, shown in FIG. Member Face-to-face, which includes an image input block 302 for uploading the image or image to be queried, and a location information input block 304 for the user to fill in the geolocation information related to the 5-day picture or image, or even The utility model comprises a query button for the user to start the query flow 7 201042473 3i^U3twr.aoc/n after uploading the query picture or image and inputting the geolocation information. In this embodiment, it is a world cultural heritage, India The image of the mausoleum of the Taj Mahal 308 as an uploaded query image's geolocation information related to this Taj Mahal image 308 includes the latitude of the Taj Mahal at 27.174604, and The longitude is 78.041814. The user can upload the image 3〇8 and simultaneously rotate the geolocation information 31 related to the uploaded query image to complete the image query information input step (that is, the image finder 2 is completed). Receiving the image query information step S101). In this embodiment, after the user obtains the image, the image is provided with additional input image related geolocation information to complete the image. Step S101 of the data searching method. However, the present invention is not limited thereto. Figure 3B is a schematic diagram of an image input method according to another embodiment of the present invention. Referring to Figure 3B, the user obtains an image by shooting a scene. The same geographical position information can be obtained at the same time, and combined with the image to become one of the summary information of the image. In this embodiment, the actual scene 312 is still taken as the Taj Mahal as an example. When the geolocation device is installed (4), the portable wire can be placed 314 to shoot the Taj Mahal 3 = at the same time, the portable device 3M ❸ 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位 定位The image query information input of the poor material is installed in the portable device. The search device 314 is, for example, a communication device for the camera image display function. The portable position device, such as the Wisdom 316 communication, can provide the Global Positioning System (GPS) of === 201042473 j iz.u〇Lwi.doc/n or can provide assisted global positioning. Information Assisted Global Positioning System (Assisted G1〇bal
Positioning System ’ AGPS)。 此外，於此實施例中，在可攜式裝置314拍攝泰姬瑪 哈陵312的同時，可攜式裝置314藉由其地理定位裝置與 喊生裝置316通訊同步得到與拍攝影像308當時可攜式裝 置314所在地的地理定位資訊318。於此實施例中，此地 理定位資訊318則登錄成影像308的影像摘要32〇之中。 亦即，於此實施例中，影像308於成像同時， 地理定位資訊318，而影像期本身即可成為一影像^ 貧訊322,直接經由使用者介面3〇〇a(如圖3B所示）中的圖 片/影像輪入欄位302a上傳。換句話說，使用者在取得影 像的同時，可將影像與同步獲得的地理定位資訊組成—影 像查詢資訊322 ’立即上傳至影像搜尋n 完成影像 料搜尋方法的步驟S101 ’以於拍攝相片同時’進行所拍 景物的同步資料搜尋。 於上述兩實施例中，地理定位資訊例如是—經緯度座 標或是經由可攜式裝置上的地理定位裝置與衛星通訊 得的一全球定位資訊或是一辅助全球定位資訊。然而，^ 發明並不受限於此。地理定位f訊的獲取還可包括辨 拍攝景物的地點或是文物所提供的一位置二維條碼。^ 辨識並解碼此位置二維條碼，可獲得所拍攝景物的相^ 理定位資料。如圖4所示，上述二維條瑪包括快速反應 碼(quick response code，QR code)。 … 9 201042473Positioning System ’ AGPS). In addition, in this embodiment, while the portable device 314 captures the Taj Mahal 312, the portable device 314 is synchronized with the shunt device 316 by its geolocation device to obtain the captured image 308 at the time. Geolocation information 318 at the location of device 314. In this embodiment, the geographic location information 318 is registered as the image summary 32 of the image 308. That is, in this embodiment, the image 308 is simultaneously positioned to geo-locate the information 318, and the image period itself can become an image 322, directly via the user interface 3〇〇a (as shown in FIG. 3B). The picture/image in the wheel is uploaded to field 302a. In other words, the user can acquire the image and the geolocation information obtained by the synchronization - the image query information 322 'immediately upload to the image search n to complete the image material search method step S101 'to take the photo at the same time' Perform a synchronized data search of the captured scene. In the above two embodiments, the geolocation information is, for example, a latitude and longitude coordinate or a global positioning information communicated with the satellite via a geolocation device on the portable device or an auxiliary global positioning information. However, the invention is not limited to this. The acquisition of the geolocation information may also include identifying the location of the scene or a location 2D barcode provided by the artifact. ^ Identify and decode the 2D barcode at this location to obtain the relative positioning data of the captured scene. As shown in FIG. 4, the above two-dimensional bar includes a quick response code (QR code). ... 9 201042473
Ji^Uitwr.aoc/n 圖5是本發明之—實施例的由候選資料組成的候選資 枓庫不意圖:接著，請參照圖卜圖2與圖5，根據所换收 的影像查詢資訊巾所包含的地理定位資訊310/318(分別參 照圖3A與圖犯），影像搜尋器2〇〇的搜尋模組綱於'網 ，網路206中ϋ集數筆候選資料5〇2(如圖5所示)成—候選 資料庫500。（步驟sl〇3)其中每—筆候選資料地包含至 ^候選影像5G4。每—筆候選資料5()2除了包 外’還包含—文字資料獨。另外，上述候 Ϊ Id Ϊ - 〇中的母—筆候選資料5〇2還包括對應候選影 、一位置座標’且每一候選資料502的位置座標均 以所接收到的地理定位資訊對應的-定位座標為中 距離定位練—特定距離以_座標顚内。 述的特定距離約為一百公尺。 一 丁 的亍S Ϊ本ΪΓ之—實施例崎比查詢影像與候選影像Ji^Uitwr.aoc/n FIG. 5 is a candidate asset library composed of candidate materials according to an embodiment of the present invention. Next, please refer to FIG. 2 and FIG. 5, and query the information towel according to the received image. The included geolocation information 310/318 (refer to FIG. 3A and FIG. 3 respectively), the search module of the image searcher 2〇〇 is outlined in the 'net, and the network 206 collects a plurality of candidate data 5〇2 (as shown in the figure). 5) shows the candidate database 500. (Step sl3) wherein each of the candidate data is included to the candidate image 5G4. Each of the candidate materials 5 () 2 in addition to the package also contains - text data alone. In addition, the parent-pen candidate data 5〇2 in the candidate Id Ϊ - 还 further includes a corresponding candidate shadow, a position coordinate 'and the position coordinates of each candidate material 502 are corresponding to the received geographic positioning information - The positioning coordinates are the middle distance positioning practice—the specific distance is within the _ coordinate. The specific distance described is approximately one hundred meters. A 亍 亍 S Ϊ ΪΓ — - implementation of the case than the query image and candidate image
It參照圖1、圖2與圖6，之後，於步驟_ 知象搜兮益200的影像比對模組2 有候選影像— 丨Γ 果至少包減選資料观其中之—。在 二影像308與候選影㈣(至少包括 =所不的候選影像50如至5041)的方法，包括 :由:J影像504a至5〇41的至少-影像特徵， 5〇41)中舖^ L 1包括圖6所示的候選影像5〇4a至 上述影像碰包括影像祕、影像_、影 10 201042473 ^izu^iwi.doc/n 汾像輪廓其巾’影像輪廓例如是影像或是圖片中的物 輪廓。 請參照圖6，以查詢影像308、候選影像5〇4g與5〇4k 為例查為衫像308 +的泰姬瑪哈陵的輪廊遍讀候選影 像通中的物件輪廓508幾乎重疊。反之，查詢影像3〇8 中的泰，瑪哈陵的輪#通與候選影像遍g中的物件輪 廢 元王不相付。因此，經由影像比對模組208進行對 ❹比驟S105之後，具有候選影像5晰的候選資料搬則 被筛選入搜尋結果。相反的，經過影像對比之後，候選影 像504g因與查騎像3G8差異較大，因此具有候選影像 5〇4g的候選資料502則被踢除在搜尋結果之外。 圖7疋本發明之一實施例的顯示搜尋結果示意簡圖。 請參照圖1、圖2與圖7,於步驟§1〇7中，由影像搜尋器 200的-顯示模組210顯示搜尋結果中每一候選資料的一 文字資料5G6(如圖7所示至少包括文字資料遍、黯、 5〇6c與506m)與至少-候選影像5〇4(如圖7所示至少包括 〇 ，選影像5〇4h、5G4k、5〇4c與504m)。然而，本發明中搜 尋結果的顯示方式並不受限於上述候選影像搭配文字資料 的顯不方式。於另一實施例中，顯示模組21〇顯示搜尋結 果的方式包括僅顯示搜尋結果中每一候選資料中的一文字 資料。 於本發明中，使用者可以在拍攝影像之後，將影像以 及與影像相關的地理定位資訊輸入至影像搜尋器，以對於 影像的相關影像圖片與文字資訊進行進一步的查詢，以在 11 201042473 J 1ZUJ> LWi.UUC/n ，用者旅遊或是拍狀後，便於使崎豐富其麵遊記或 是拍攝記錄。另外，使用者亦可以於拍攝影像的同時，利 用拍攝影像的可攜式裝置内建的地理定位系統，將影像與 地理定位資訊結合，利用安裝於可攜式裝置上的影像搜尋 器，立時進行所拍攝之影像相關資料搜尋，可提供使用者 即時的影像與文字資訊，例如所拍攝景物或文物的旅遊導 覽，歷史紀錄。藉由所查詢影像的相關地理定位資訊縮小 搜尋範圍，提高以圖像搜尋圖像或是以圖像搜尋文字的簡 易程度與搜勞結果的正確性，且使用者可以在觀賞實際景 物或文物的當下，立即獲得相關的旅遊資訊，便於使用者 立即規劃行程以及提升使用者的旅遊深度。 义雖“、丨本發明已以貫施例揭露如上，然其並非用以限定 本^明，任何所屬技術領域中具有通常知識者，在不脫離 ^發明之精神和範圍内，當可作些許之更動與潤飾，故本 發明之保護範圍當視後附之申請專利範圍所界定者為準。 【圖式簡單說明】 圖1疋本發明之一實施例的一種資料搜尋方法流程簡 圖。 圖2是本發明之一實施例的一種影像搜尋器簡圖。 -立曰3八疋依照本發明之—實施例的一種影像輸入方法 不思圖。 一 Q 33疋依照本發明之另—實施例的一種影像輸入方 法示意圖。 、 12 201042473 -J x u v vr 1.. doc/n 圖4是依照本發明之一實施例的泰姬瑪哈陵之位置二 維條碼圖。 ® 發明之_實施例的由候選資料組成的候選資 料庫不意圖。 ' 的示本發明之一實施例的對比查詢影像與候選影像 圖7是本發明之-實施例的顯示搜尋結果示意簡圖。 ® 【主要元件符號說明】 S101〜S107 :方法流程步驟 200:影像搜尋器 202 :查詢模組 204 :搜尋模組 206 :網際網路 208 .影像對比模組 210 :顯示模組 〇 300:使用者介面 302、302a :影像輸入襴位 304 :定位資訊輸入攔位 306 :按鍵 308 :查詢影像 308a、508、51〇 :物件輪廓 310、318 :地理定位資訊 312 :實際景物 13 201042473 J 丄厶UJLWi.UULyn 314 :可攜式裝置 316 :衛星 320 :影像摘要 322 :查詢資訊 500 :候選資料庫 502:候選資料 504、504a〜504卜504m :候選影像 506、506h、506k、506c、506m :文字資料It refers to Fig. 1, Fig. 2 and Fig. 6, and then, in the image matching module 2 of the step _ 兮 兮 200 200, there is a candidate image - at least the information of the deducted data is included. The method of the second image 308 and the candidate image (four) (including at least the candidate image 50 as the 5041) includes: at least: image features of: J images 504a to 5〇41, 5〇41) 1 includes the candidate image 5〇4a shown in FIG. 6 to the image collision including the image secret, the image _, the shadow 10 201042473 ^izu^iwi.doc/n the outline of the image, the image outline is, for example, in an image or a picture Object outline. Referring to FIG. 6, the query image 308, the candidate images 5〇4g and 5〇4k are taken as an example, and the object contours 508 in the porch-passing candidate image pass of the Taj Mahal of the shirt image 308+ are almost overlapped. On the contrary, the query in the image 3〇8, the Mahal's wheel #通 and the candidate image in the g of the object wheel waste yuan king do not pay. Therefore, after the contrast ratio S105 is performed by the image matching module 208, the candidate data with the candidate image 5 is filtered and filtered into the search result. Conversely, after the image comparison, the candidate image 504g differs greatly from the look-up image 3G8, so the candidate material 502 having the candidate image 5〇4g is kicked out of the search result. Figure 7 is a schematic diagram showing the results of display search according to an embodiment of the present invention. Referring to FIG. 1 , FIG. 2 and FIG. 7 , in step §1〇7, a text data 5G6 of each candidate material in the search result is displayed by the display module 210 of the image searcher 200 (as shown in FIG. 7 at least The text data is over, 黯, 5〇6c and 506m) and at least - the candidate image 5〇4 (including at least 〇, as shown in Fig. 7, images 5〇4h, 5G4k, 5〇4c and 504m). However, the manner in which the search results are displayed in the present invention is not limited to the manner in which the candidate images are matched with the text data. In another embodiment, the manner in which the display module 21 displays the search result includes displaying only one of the candidate materials in the search result. In the present invention, the user can input the image and the geolocation information related to the image to the image searcher after the image is captured, so as to further query the related image and text information of the image, at 11 201042473 J 1ZUJ> LWi.UUC/n, after the user travels or pats the picture, it is convenient for the Kawasaki to enrich his face or record. In addition, the user can also use the built-in geolocation system of the portable device to capture the image, and combine the image with the geolocation information, and use the image searcher installed on the portable device to immediately perform the image. The image-related data search can provide users with instant video and text information, such as tourist guides and historical records of the captured scenes or cultural relics. The search range is narrowed by the related geolocation information of the image to be searched, and the ease of searching for images by images or searching for texts and the accuracy of the search results are improved, and the user can watch the actual scenes or cultural relics. At the moment, you can get relevant travel information immediately, so that users can immediately plan their itinerary and improve the user's travel depth. Although the present invention has been disclosed in the above embodiments, it is not intended to limit the scope of the invention, and any person having ordinary knowledge in the technical field can make some changes without departing from the spirit and scope of the invention. The scope of protection of the present invention is defined by the scope of the appended claims. [FIG. 1] FIG. 1 is a schematic flow chart of a data searching method according to an embodiment of the present invention. 2 is a schematic diagram of an image searcher according to an embodiment of the present invention. - An image input method according to an embodiment of the present invention is not considered. A Q 33疋 according to another embodiment of the present invention A schematic diagram of an image input method. 12 201042473 -J xuv vr 1. doc/n FIG. 4 is a two-dimensional bar code diagram of the position of Taj Mahal in accordance with an embodiment of the present invention. A candidate database composed of candidate materials is not intended. 'Comparative query image and candidate image showing an embodiment of the present invention. FIG. 7 is a schematic diagram showing the display search result of the embodiment of the present invention. S101~S107: Method Flow Step 200: Image Finder 202: Query Module 204: Search Module 206: Internet 208. Image Comparison Module 210: Display Module 〇300: User Interface 302, 302a: Image Input Clamp 304: Positioning Information Input Block 306: Button 308: Query Image 308a, 508, 51〇: Object Profile 310, 318: Geolocation Information 312: Actual Scene 13 201042473 J 丄厶UJLWi.UULyn 314: Portable Device 316: satellite 320: image summary 322: query information 500: candidate database 502: candidate data 504, 504a~504 504m: candidate images 506, 506h, 506k, 506c, 506m: text data
1414