201025052 六、發明說明: 【發明所屬之技術領域】 本發明係關於一種影像分類技術’特別是一種具有自 動分類影像之功能的影像擷取裝置及用於該影像摘取裝置 之影像分類方法。 【先前技術】 隨著影像擷取裝置的普及,儲存的媒體的容量大幅增 加,但其成本卻日益降低。因此現在的使用者所使用的^ 存媒體能夠儲存大量的照片。但是當使用者拍了如此大量 的照片時,要整理照片時就顯得非常的耗費時間。在先前 ,術中,通常照片都依照其年、月、日之時間來排序。但 是當使用者在同一日中拍攝了大量的照片,仍 者在日後整理照片時的困擾。 在先前技術中,另外揭露了利用照片的偏好屬性進行 罄分類的方法。但是這種方法需要使用者在事先就將屬性種 類設定,事後再進行照片的_選。如此一來,仍然會造 使用者在使用上的不便。 方法來解決先前技術 因此’有需要發明一種新的分類 所具有的缺失。 【發明内容】 本發明之主要目的係在提供—種影像擁 像分類方法,係可自動分類該影軸取裝置賴取之及影像影 201025052 為達成上述之目的,本發明之影像擷取裝置勺 擷取模組、排列模組、偵測模組及分類模組其=影像 擷取模組用以擷取複數之影像,且複數之影像具’影像 擷取時間點;排列模組用以將複數之影像依數^ 後作順序排列:制模組於複數之影像中偵測出 及第二影像’其中第一影像之榻取時心 像;分類模組用以將包含第一影像之掏取時間點前之2 φ =像歸類為同-群組,而將包含第二影像之錄時間點^ 後之所有影像亦歸類為同一群組。 本發明亦揭示一種影像分類方法,包括以下步驟.八 才::該些影像各自之擷取時間點;依照影像之擷 : 先後順序對複數之影像進行排列;預設一分類時間間隔; 於複數之影像中找出所有相鄰之練時間點其時間間隔大 於分類時間間隔之第一影像及第二影像,其中,第一影像 汽取時間點係先於第二影像;以及將包含第—影像之榻 φ取時間點前之所有影像歸類為同一群組,而將包含第二影 像之擷取時間點之後之所有影像亦歸類為同一群組。’、 較佳地,偵測模組復供偵測影像擷取裝置之開關機時 破該分類模組所分類之同—群組之影像,係皆影像 置於一次開關機間之啟動週期所擷取者,而每一啟 =之首張影像可視為第二影像,且每-啟動週期之末 張影像可视為該帛一影像 古雖t於習知技術中當一影像擷取裝置攝取大量影像後 y刀類之因擾’本發明之影像擷取裝置及影像分類方 5 201025052 法係以時間間隔搭配啟動週期將所擷取的影像做自動分 類,基於一般人拍照的習慣可將影像作甚為合理的分類, 即使轉存到個人電腦後需要進一步整理,亦較過去影像均 存於單一資料夾之形式簡易許多。 再者,即使習知技術已有可按照片屬性分類者,然而 仍需要使用者自行在相機上作設定分類,而本發明之影像 擷取裝置及影像分類方法係以時間間隔搭配啟動週期將所 擷取的影像做自動分類,無須使用者在相機介面上費心的 進行影像分類。 【實施方式】 為讓本發明之上述和其他目的、特徵和優點能更明顯 易懂’下文特舉出本發明之具體實施例,並配合所附圖式, 作詳細說明如下。 請先參考圖1關於本發明影像掏取裝置之架構圖。 本發明之影像_取裝置1 〇具有影像分類之功能,能夠 自動地將所擷取的影像進行分類。影像擷取裝置10包括拙 •此電性連接之影像擷取模U、排列模二貞= 13、分類模組14、及儲存模組15。201025052 VI. Description of the Invention: [Technical Field] The present invention relates to an image classification technique, in particular, an image capture device having a function of automatically classifying images and an image classification method for the image pickup device. [Prior Art] With the popularity of image capture devices, the capacity of stored media has increased significantly, but its cost has been decreasing. Therefore, the memory used by current users can store a large number of photos. But when users take such a large number of photos, it takes a lot of time to organize the photos. In the past, during the operation, usually the photos were sorted according to their time of year, month and day. However, when the user took a large number of photos on the same day, they still have troubles in finishing the photos in the future. In the prior art, a method of performing 罄 classification using the preference attribute of a photo is additionally disclosed. However, this method requires the user to set the attribute type in advance, and then perform the _ selection of the photo afterwards. As a result, the user's inconvenience in use will still be created. The method solves the prior art and therefore there is a need to invent a new classification with the missing. SUMMARY OF THE INVENTION The main object of the present invention is to provide an image capturing method, which can automatically classify the image capturing device and image shadow 201025052. To achieve the above object, the image capturing device spoon of the present invention The capture module, the alignment module, the detection module and the classification module, the image capture module is used to capture a plurality of images, and the plurality of images have an 'image capture time point; the array module is used to The plurality of images are arranged in the order of the following: the module detects the second image 'in the image of the first image, and the classification module is used to capture the first image. 2 φ = before the time point = the image is classified as the same group, and all the images including the time point of the second image are also classified into the same group. The invention also discloses an image classification method, which comprises the following steps: eight:: the time points of each of the images; according to the image: sequentially arranging the images of the plurality; presetting a classification time interval; Finding a first image and a second image of all adjacent training time points whose time interval is greater than the classification time interval, wherein the first image capturing time point is prior to the second image; and the first image is included All the images before the time point of the couch φ are classified into the same group, and all the images after the capturing time point including the second image are also classified into the same group. Preferably, when the detecting module is re-used to detect the on/off of the image capturing device, the image of the same group classified by the classification module is broken, and the images are placed in a startup cycle between the switches. The first image of each of the starter images can be regarded as the second image, and the image at the end of each start-up period can be regarded as the image of the image, but it is taken by an image capturing device in the prior art. After a large number of images, the y-knife is disturbed. The image capturing device and image classification method of the present invention 5 201025052 The system automatically classifies the captured images with a time interval and a start-up cycle, and the image can be made based on the habit of taking pictures by ordinary people. A very reasonable classification, even if it needs to be further sorted after being transferred to a personal computer, is much easier than the past image is stored in a single folder. Moreover, even if the prior art has been classified by the photo attribute, the user still needs to set the classification on the camera by himself, and the image capturing device and the image classification method of the present invention are combined with the start-up period at time intervals. The captured images are automatically classified, and the user does not need to perform image classification on the camera interface. The above and other objects, features and advantages of the present invention will become more <RTIgt; <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; Please refer to FIG. 1 for the architecture diagram of the image capturing device of the present invention. The image capturing device 1 of the present invention has the function of image classification, and can automatically classify the captured images. The image capturing device 10 includes: the electronically connected image capturing module U, the array module 2, the sorting module 14, and the storage module 15.
影像擷取模組11用以擷取影像,當影像擷取模組U 擷取複數之影像後,㈣模組12用以將該複數之影像依掏 取時間點先後作順序排列。 摘測模組13於該複數之影像中偵測出所有相鄰之拇取 時間點其時間間隔大於-分類時間間隔之第一影像及第二 影像,其中第-影像之揭取時間點係先於第二影像。分類 201025052 模組14則用以脏 像歸類為同if包含第一影像之擷取時間點前之所有影 之所有影像亦歸二η含第^影像之類取時間點之後 之影像為-第_ ^為同一群組,意即同一群組中最先攝得 此外,偵測最後攝得之影像則為一第一影像。 間,而被該分類模λ供债測影像榻取裝置11之開關機時 像揭取裝置λ所分類巧―群組之影像,係皆景ί 期之末張影像可視為該第-ί像象’且每一啟動週 像,===:,模組一- :之:之意一^ 啟動z模中視影像棘裝置u之開關機次數及各 中前述第-、二影像出現次數而儲存有單一或複 數之群組。以圖2為例’儲存模組15可儲存有啟動週期 2〇a、啟動週期2〇b及啟動週期20c。而啟動週期漁有群 組2la及群組2ib,啟動週期20c有群組2lc id及 群組21e。每-啟動週期或是群組之首張影像可視為第二 影像32,且每-啟動週期或是群組之末張影像可視為第一 影像3卜而在影像擷取裝置10的同一開關機間之啟動週 期中,也可利用分類時間間隔劃分出不同之分類群组。因 此依不同之分類時間間隔之設定,在影像梅取裝置1〇的同 一啟動週期中可具有單-或是複數之分類群址。本發明並 7 201025052 不以單一群組為限。 接著請參考圖3’其係關於本發明影像分類方法之第一 實施例之流程圖。此處需應注意的是,以下雖以前述之影 像擷取裝置H)為例說明本發明之影像分類方法但本發明 之影像分類方法並不以使用在影像祿取裝置ι〇為限。 首先進行步驟301 :掏取複數之影像。 首先影像練裝置1G彻影像齡模組^擷取複數 之影像」其”取複數之影像具有複數之齡時間點。 接著進行步驟302 :分析出該些影像各自之擷取時間 i 模組13讀取複數之影像擷取時_之資料。藉此 始:、:13即可判定出複數之影像之擷取時間點。其中在 像中建立掏取時間點資料之方式係為現有之影像 法在綠取影像時皆會建立的相關資訊 ,故在此不 益私接步驟303:依照複數之擷取賴點先後順序對 複數之影像進行排列。 像録模、组11齡複數之影像後,係利用排列模 1 且甘細 <數之影像依掏取時間點先後順序進行排列,即得 知其擷取順序。 時間:::步驟304:於複數之影像中找出所有相鄰之擷取 .、、間間隔大於分類時間間隔之第一影像及第二影 像。 的擷=3:,组13進-步依照排序,分別找出各影像間 取時間點疋否相隔超過一預先設定之分類時間間隔。 8 201025052 需注意的是,上琉认、 定,可為三十分鐘、十;^類時Λ間隔亦可由使用者設 限。如此_來,即可分二鐘,本發明並不以此為 的第-影像31及第二二出間點大於分類 最後進行步驟305:將 之所有影像歸類為同一群:U像之擷取時間點前 間點之後之所有影像歸類為另一同—群之_取時 門二m ’當第一影像31及第二影像32之梅取賠 b間隔時,可將第-影像b及第= 32 Q刀為在影像操取裝 一t/像 影像。因此分類模組以^ 前所擷取之影像分成相同的二::1及第-:象31之 組13内。再將第二# ^組巾,雜存於儲存模 模組U内的另一二32中及之後操㈣ 著分類之方法並不以上述之流程為限。接 ❿ 本發明影像分類方法之第二實施例之步 首先進行步驟401 :擷取複數之影像。 ,首先影像擷取裝置10利用影像擷取模組11擷取複數之 影像。由於此步驟與步驟301相同,故在此不再贅述。 再進行步驟402 :紀錄影像擷取裝置之開關機時間點。 偵測模組13偵測出在影像擷取時,影像擷取裝置1〇的 ^機到關機之時間點及其時間長短,並將其分別記錄於複 數之影像之儲存資料中。 後 9 201025052 接著進行步驟403 :分析出該些影像各自之擷取時間 點,以及步驟4〇4 :依照複數之擷取時間點先後順序對複數 之影像進行排列。 由於步驟403及步驟404和步驟302及步驟3〇3相似,故 在此不再贅述。 接著進行步驟405:於複數之影像中找出所有相鄰之擷 取時間點其時間間隔大於分類時間間隔之第一影像及第二 序夕像。 取時7步依’序,分別找出各影像間的掘 :„相隔超過一預先設定之分類時間間隔。如此 像第可取時間點大於分類時間間隔的第一影 視將㈣關_關之料之首張影像 ❹ 接著摘測模組13係讀取各個影像中 建立的開關機時間點。如此—來便 ^所 中所擷取。當不=二=個開關機間之啟動週期 也可判定為不同===啟動週期的資料時, 間隔。因此,第1傻 、、曰去之間隔大於分類時間 矛影像31即可視為在影俊描俶壯取 J -次啟動週期中之末張影像’第二影像裝置10的 榻取裝置1G的—次啟動週期中之首張影像可視為在影像 201025052 間點之後之所有影像歸類為另;二=7影像之擁取時 刀類模、卫14係根據上述的步驟,將複數之聲傻推/ Γ3Γ:4Τ分類模組14係將第-影像 的啟動週期-二: =影像错存到職㈣内的另一啟:二 述之分類之方法並不以上 驟次序亦可加以改變:成本發明之目的,上述之步 体、束’影像擷取裝置10在進行擷取影像時,即可 大、ΐ二C類。因此使用者在曰後檢視所摘取之 大量的影像時,即可方便地進行進一步的整理。 相較於習知技術中當一影像操 3以:類之困擾,本發明之影像•裝置及二= «隔搭配啟動將所練的影像做自動分 甚為合理的分類, 存於單-資料夾=易要;多一步整理,亦較過去影像均 再者’即使習知技術已有可按照片屬性分類者,然而 仍需要使用者自行在相機上作設定分類,而本發明之影像 擷取裝置及影像分類方法係以時間間隔搭配啟動週期將所 操取的影像做自動分類,無須使用者在相機介面上費心的 11 201025052 進行影像分類。 綜上所陳,本發明無論就目的、手段及功效,在在均 顯示其迥異於習知技術之特徵,懇請貴審查委員明察, 早曰賜准專利,俾嘉惠社會,實感德便。惟應注意的是, 上述諸多實施例僅係為了便於說明而舉例而已,本發明所 主張之權利範圍自應以申請專利範圍所述為準,而非僅限 於上述實施例。 【圖式簡單說明】 圖1係本發明影像擷取裝置之架構圖。 圖2係本發明影像擷取裝置之儲存模組之一實施例之示意 圖。 圖3係本發明影像分類方法之第一實施方式之步驟流程圖。 圖4係本發明影像分類方法之第二實施方式之步驟流程圖。 【主要元件符號說明】 影像擷取裝置10 ®影像掘取模組11 排列模組12 偵測模組13 分類模組14 儲存模組15 啟動週期20a、20b、20c 群組21a、21b、21c、21d、21e 第一影像31 12 201025052 第二影像32The image capture module 11 is configured to capture images. After the image capture module U captures the plurality of images, the module 12 is configured to sequentially arrange the plurality of images according to the time points. The extracting module 13 detects, in the image of the plurality of images, the first image and the second image whose time interval is greater than the time of the classification time, wherein the first image is extracted first. In the second image. Classification 201025052 The module 14 is used to classify all the images of the image before the time point of the capture of the first image, and the image after the time point is taken as the second image. _ ^ is the same group, meaning that the first group in the same group is taken. In addition, the last image captured is detected as a first image. And the image of the group is classified by the classification mode λ for the image-taking device 11 of the image-receiving device 11, and the image of the group is regarded as the first image. Like 'and each start-up image, ===:, module one-:: meaning one ^ start the z-mode view of the image-spinning device u and the number of occurrences of the above-mentioned first and second images There are single or plural groups. Taking FIG. 2 as an example, the storage module 15 can store a startup cycle 2〇a, a startup cycle 2〇b, and a startup cycle 20c. The start cycle has a group 2la and a group 2ib, and the start cycle 20c has a group 2lc id and a group 21e. The first image of each start-up period or group can be regarded as the second image 32, and the last image of each start-up period or group can be regarded as the first image 3 and the same switch machine of the image capturing device 10 During the start-up period, different classification groups can also be divided by the classification time interval. Therefore, according to the setting of different classification time intervals, there may be a single- or complex classification group address in the same startup period of the image capturing device. The present invention and 7 201025052 are not limited to a single group. Next, please refer to FIG. 3' for a flowchart of a first embodiment of the image classification method of the present invention. It should be noted here that the image classification method of the present invention is described below by taking the image capturing device H) as described above, but the image classification method of the present invention is not limited to the image capturing device. First, step 301 is performed: capturing a plurality of images. First, the image training device 1G image age module ^ captures the plural image "the" image of the complex number has a complex age time point. Then proceeds to step 302: analyzes the capture time of each of the images i module 13 reads Take the data of the plural image at the time of taking the _. From this::: 13 can determine the time point of the image of the complex number. The way to find the time point data in the image is the existing image method. Green will take the relevant information that will be established when the image is taken, so it is not beneficial to the private step 303: the plural images are arranged according to the order of the plurals. After recording, the image of the 11-year-old plural is used. The order of the ordering time is determined by using the modulo 1 and the images of the cassicles are counted in order of time. Time::: Step 304: Find all adjacent captures in the image of the complex number . , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Time interval 8 201025052 It is intended that the upper 琉 、 、 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , The image 31 and the second and second outgoing points are larger than the classification. Finally, step 305 is performed: all the images are classified into the same group: all the images after the time point before the U image are classified as another one - the group_ When the first image 31 and the second image 32 are in the interval of b, the first image b and the third 32 knife can be used to load a t/image image in the image manipulation. The group is divided into the same two::1 and the first: the group 13 of the group 31. The second #^ group towel is stored in the other two of the storage module U. The method of classification is not limited to the above process. Next, the second embodiment of the image classification method of the present invention first performs step 401: capturing a plurality of images. First, the image capturing device 10 The image capture module 11 captures the image of the plurality. Since this step is the same as step 301, it will not be described here. Step 402: Record The switching time of the image capturing device is detected. The detecting module 13 detects the time point of the image capturing device 1 to the shutdown time and the length of time when the image capturing is performed, and records the time in the plural In the stored data of the image. After 9 201025052, proceed to step 403: analyze the time points of each of the images, and step 4:4: arrange the images of the plurality according to the order of the points of the plural. Step 403 and step 404 are similar to step 302 and step 3〇3, and therefore are not described herein again. Step 405 is performed to find all adjacent extraction time points in the complex image, and the time interval is greater than the classification time interval. The first image and the second sequence image. Take the 7 steps in order to find out the difference between the images: „separated by more than a preset classification time interval. So the first film with the desired time point greater than the classification time interval will be (4) off_off The first image ❹ Next, the measurement module 13 reads the time point of the switch established in each image. Thus, it can be retrieved. When not = two = the start cycle between the switches can be determined as Different === start period data, interval. Therefore, the first silly, and the interval between the squats and the smashing time is greater than the classification time spear image 31 can be regarded as the last image in the J-time start cycle. The first image in the first activation cycle of the couching device 1G of the second image device 10 can be regarded as all the images after the image 201025052 is classified as another; the second image of the image capture mode is the other. According to the above steps, the plural sounds are pushed / Γ3Γ: 4Τ classification module 14 is the first-initial period of the first image - two: = image is stored in the other (4): the method of classification It may not be changed in the order of the above: for the purpose of the invention, The step body and the beam image capturing device 10 can be used to capture images, and can be used in large and small class C. Therefore, when the user views a large number of images taken after the scene, the user can conveniently carry out further steps. Compared with the prior art, when an image manipulation 3 is troubled by the class, the image device and the device of the present invention and the second = «separate and start to automatically divide the image to be properly classified, and exist in the classification. Single-folder=easily; more than one step, and more than the past images, even if the prior art has been classified by photo attribute, it still requires the user to make a set classification on the camera, and the present invention The image capturing device and the image classification method automatically classify the captured images with the time interval and the start-up cycle, without the user having to troubleshoot the image on the camera interface 11 201025052. In summary, the present invention is aimed at The means, the functions and the effects are all different in the characteristics of the prior art, so please ask the reviewing committee to inspect the patents, and to give the company a good faith. It should be noted that the above-mentioned The embodiments are intended to be illustrative only, and the scope of the claims is intended to be limited by the scope of the claims, and not limited to the above embodiments. Figure 2 is a schematic diagram of an embodiment of a storage module of the image capturing device of the present invention. Figure 3 is a flow chart showing the steps of the first embodiment of the image classification method of the present invention. Flow chart of the second embodiment of the classification method. [Description of main component symbols] Image capture device 10 ® image capture module 11 alignment module 12 detection module 13 classification module 14 storage module 15 startup cycle 20a 20b, 20c group 21a, 21b, 21c, 21d, 21e first image 31 12 201025052 second image 32
1313