201028017 六、發明說明： 【發明所屬之技術領域】 本發明是有關於一種晝面頻率提高之轉換方法 (frame rate up conversion ; FRUC)，且特別是有關於一 種用於轉換提高晝面頻率而使用之移動補償插補方法。 【先前技術】 在一般晝面頻率提高的轉換方法（frame rate up 參 conversion ; FRUC)中，若兩原始晝面的資訊已知，則 可從兩原始晝面之間插補加入一新的晝面。具體地來 說，一旦對兩連續晝面的運動向量進行分析和估算，便 可因此獲得連續晝面相關的運動資訊，然後新的畫面便 可根據相關的運動資訊或運動向量插補加入於連續晝 面之間。 對目前已知的原始資料來說，動晝影像可能以每秒 24個晝面（24Hz)的頻率記錄於其中，而原始影像資料 ® 則可能以每秒30個晝面（30Hz)的頻率記錄於其中。為 了在大螢幕上觀賞具24Hz或30Hz的影像資料，而不 影響其顯示的品質，習知技術係利用晝面頻率提高的轉 換方法，將24/30HZ的晝面轉換為60Hz的晝面。然而， 若是欲進一步改善影像的顯示品質，則必須將60Hz的 畫面頻率轉換至更高才行，此時則有必要利用晝面頻率 提高的轉換方法，將60Hz的晝面頻率轉換至如120Hz 的晝面頻率。 201028017 【發明内容】 本發明的目的是在提供一種轉換畫面頻率之方 法，藉以提升晝面頻率’並因此改善影像的顯示品質。201028017 VI. Description of the Invention: [Technical Field] The present invention relates to a frame rate up conversion (FRUC), and in particular to a method for converting and increasing the frequency of a facet The motion compensation interpolation method. [Prior Art] In the general frame frequency up conversion method (FRUC), if the information of the two original faces is known, a new one can be inserted between the two original faces. surface. Specifically, once the motion vectors of two consecutive planes are analyzed and estimated, continuous motion information related to the motion can be obtained, and then the new picture can be added to the continuous motion according to the relevant motion information or motion vector interpolation. Between the faces. For currently known raw materials, moving images may be recorded at 24 frames per second (24 Hz), while raw image data may be recorded at 30 frames per second (30 Hz). In it. In order to view 24 Hz or 30 Hz image data on a large screen without affecting the quality of its display, the conventional technique converts the 24/30 Hz surface into a 60 Hz surface using a kneading frequency-enhanced conversion method. However, if you want to further improve the display quality of the image, you must convert the 60Hz picture frequency to a higher level. In this case, it is necessary to convert the 60Hz facet frequency to 120Hz, for example, by using the kneading frequency increase conversion method. Face frequency. 201028017 SUMMARY OF THE INVENTION An object of the present invention is to provide a method of converting a picture frequency, thereby increasing the facet frequency and thereby improving the display quality of the image.
依照本發明一實施例，提出一種轉換視頻信號之晝 面頻率之方法，包含：接收一折疊影像序列，此折疊影 像序列係存在於或轉換自視頻信號之一連續時間之晝 面序列」其中折疊影像序列包含複數個互異之原始畫 面，且每一個原始晝面存在有相對應數量之重複晝面； 修改上述原始晝面;對與修改後之上述原始晝面相關之 至少一運動向量進行估算；以及根據運動向量插補新畫 面於修改後之原始晝面之間。 依照本發明另一實施例，提出一種轉換視頻信號之 面頻率之方法’包含.接收存在於或轉換自視頻信號 =一晝面序列；對畫面序列進行運動估算；根據運動估 f之結果判別晝面序列是否為一折疊影像序列，其中折 ς影像序列包含複數個互異之原始晝面，且每一個原始 ^在有相對應數量之重複畫面；當畫面序列為折疊 =像序列時，修改原始晝面；對與修改後之原始畫面相 =至少一運動向量進行估算;以及根據運動向量插補 新晝面於修改後之原始畫面之間。 根據本發明之技㈣容L述轉換畫面頻率之 /可將具有高晝面頻率的畫面序列轉換為具有更 ^面頻率的新畫面序列，藉以改善動晝影像的顯示品 201028017 【實施方式】 下列實施例係揭露一種提高視頻信號之畫面頻率 的轉換方法（frame rate up conversion ; FRUC)。在下述 轉換方法中，當兩原始晝面的資訊已知時，則可從兩原 始晝面之間插補加入新的畫面。第1圖係緣示依照本發 明實施例之一種在連續晝面之間插補新畫面的方式。如 第1圖所示，當兩連續畫面A和B的運動向量進行估 算時，可因此獲得連續晝面A和B的相關運動資訊， ❹ 然後新的晝面C可根據晝面A和B的相關運動資訊插 補加入。 第2圖係繪示依照本發明實施例之一種轉換視頻 信號之晝面頻率之方法的流程圖。首先，接收一折叠 (pulldown)影像序列（步驟202) ’其中此折疊影像序列係 存在於或轉換自視頻信號中之一連續時間的書面序 列’並包括複數個互異的原始畫面，且每一個原^晝面 均存在有相對應數量且重複的晝面。接著，調整修改折 ❹ 疊影像序列中的原始畫面（步驟204)。之後，對與修改 後的原始晝面相關之至少一運動向量進行估算（步驟 2〇6)。然後，根據運動向量的估算結果’將新畫面插補 於修改後的原始畫面之間（步驟208)。 下列係以例子來說明利用上述轉換方法將6〇Hz金 面頻率轉換至ΠΟΗζ晝面頻率的情形。第％圖至第& 圖係依照本發明實施例繪示藉由如第2圖所示之 方法將原始影像序列轉換為新影像序列的過程。如第 3a圖所不’首先接收存在於或轉換自連續晝面序列， 201028017 且具有60Hz晝面頻率的3: 2折疊影像序列i i i 7 7 j 3 • 13 13…，其中數字1、7、13…等係例示性地代表序列 中晝面的位置（或順序）。接著，對折疊影像序列中的原 始畫面進打修改。以本實施例而言，係對折疊影像序列 中原始晝面之重複晝面（即卜7、13等）進行移除的動 作（如第3b圖所示）。 在原始晝面中的重複晝面移除之後，接著對原始畫 面中剩餘晝面之兩相鄰晝面（即2和7、7和13、等） • ㈣關的運動向量進行估算。以本實施例而言，係根據 運動向量在原始晝面中剩餘的兩相鄰晝面之間插補連 續時間的新畫面。舉例來說，如第3c圖所示，晝面23 4 5 6係插補於晝面1和7之間，而晝面789HUH2 則是插補於畫面7和13之間，依此類推。然後，便可 因而產生具晝面頻率120Hz的新晝面序列1 2 3 4 5 6 7 8 9 10 11 12 13...(如第 3d 圖所示）。 第4a圖至第4d圖係依照本發明另一實施例繪示藉 粵 ώ如第2圖所示之轉換方法將原始影像序列轉換為新 影像序列的過程。同樣地，如第4a圖所示，首先接收 存在於或轉換自連續畫面序列，且具有6他晝面頻率 的3 : 2折疊影像序列M丨77 13 13 13 。接著，對 f疊影像序列中的原始晝面進行修改。以本實施例而 :制其方法是對互異原始晝面中之每—者進行選擇性地 複1，使付互異原始晝面中之每—者均存在有相同數量 =複晝I舉例來m丨和13兩者原先均存在 有二個重複晝面’而晝面7原先存在有二個重複畫面， 201028017 = 進如行第複：，而與晝面1和13-樣具有相 里耵旦面（如第4b圖所示）。 晝面= = ; =改之後，接著對與修改後之 蚓啕立呉畫面（即1和7、7和13、蓉U日關 =動向量進行估算。以本實施例而言，其方法是在估 异、動向量之後，在修改後之原始畫面中相鄰兩互異晝 面之間’根據運動向量插補數個新畫面。在此值得注音 :是，新晝面插補的數量係小於原始晝面中；“ 數量。舉例來說，如第4c圖所示，畫面= 相鄰畫面1和7之Η二去 你播補於 _ 7之間，而晝面10 10係插補於相鄰畫面 和之間，依此類推；其中晝面4、10、…等的數量 丄7、13、...等的數量。然後，便可因而 產生具畫面頻率120ΗΖ的新3:2折叠影像序歹 7 10 10 13 13 13...(如第 4d 圖所示）。 在此值得注意的是，H述實施例仙3 : ❹ 序列來實現其僅是用來說明本發明，而非用以 像序列此，2:2折疊影像序列或其他折疊影 像序列亦可以相_方式制於本發明中。 頻广本發日以—實__—種轉換視 ，號之畫面頻率之方法的流程圖。首先，接收存在於 或轉換自視頻信號的晝面序列(步驟502)。接著，對蚩 面t列進行運動估算(步驟綱)。之後，根據運動估； =。果㈣所純之畫面序収否為折疊影像 驟506)，其中折疊影像序列包括複數個 （ 面，且每-個原始畫面均存在有相對應數量且 201028017 面，例如第3a圖所示的3:2折疊影像序列工ΐΐ77ΐ3 =3 ...。若是畫面序縣—般序列而非折叠影像 的話，則畫面序列會經由一般的運動補 :(步二5〇8)。反之，若是晝面序列為折疊影像序列的 缺播11 的原來晝面進行調整修改（步驟51〇)。 …後，對與修改後的原始畫面相關之至少一 :于估算(步驟5…而後’根據運動向量估算的二J進 插補新晝面於修改後的原始畫面之間（步驟514)。在此 值得注意的是，本實施例中的步驟51〇、512和Η# 刀別類似於第2圖中所示之步驟2〇4、2〇6和2〇8。’、 由上述本發明之實施例可知，應用前述轉換畫面頻 拖ίΓί可將具有高畫面頻率（如：6GHZ)的畫面序列轉 、、為/、有更南畫面頻率（如：12〇Hz)的新畫面序 以改善影像的顯示品質。 —雖然本發明已以實施例揭露如上，然其並非用以限 φ =本發明，任何具有本發明所屬技術領域之通常知識 者，在不脫離本發明之精神和範圍内，當可作各種之更 動與潤飾，因此本發明之保護範时視後附之申請專利 範圍所界定者為準。 【圖式簡單說明】 第1圖係繪示依照本發明實施例之一種在連續畫 面之間插補新晝面的方式。 、 _第2圖係繪示依照本發明實施例之一種轉換視頻 信號之畫面頻率之方法的流程圖。 8 201028017 第3a圖至第3d圖係依照本發明實施例緣示藉由如 第2圖所示之轉換方法將原始影像序列轉換為新影像 序列的過程。 第4a圖至第4d圖係依照本發明另一實施例繪示藉 由如第2圖所示之轉換方法將原始影像序列轉換為新 影像序列的過程。 第5圖係依照本發明另一實施例繪示一種轉換視 頻信號之晝面頻率之方法的流程圖。 ❹ 【主要元件符號說明】 202〜208、502〜514 :步驟According to an embodiment of the invention, a method for converting a facet frequency of a video signal is provided, comprising: receiving a folded image sequence, the folded image sequence being present or converted from a sequence of one consecutive time of the video signal" The image sequence includes a plurality of different original images, and each original surface has a corresponding number of repeated faces; modifying the original face; estimating at least one motion vector related to the modified original facet And interpolating the new picture between the modified original faces according to the motion vector. According to another embodiment of the present invention, a method for converting a surface frequency of a video signal is provided, which includes: receiving or converting from a video signal = a facet sequence; performing motion estimation on the picture sequence; and determining the result according to the motion estimation f. Whether the face sequence is a folded image sequence, wherein the folded image sequence comprises a plurality of mutually different original faces, and each original has a corresponding number of repeated pictures; when the picture sequence is a fold = image sequence, the original is modified昼; estimating the original picture phase with the modified at least one motion vector; and interpolating the new picture between the modified original pictures according to the motion vector. According to the technique of the present invention, the picture sequence having the high face frequency can be converted into a new picture sequence having a higher face frequency, thereby improving the display of the moving picture 201028017. [Embodiment] The embodiment discloses a frame rate up conversion (FRUC) for increasing the picture frequency of a video signal. In the following conversion method, when the information of the two original faces is known, a new picture can be added by interpolation between the two original faces. Figure 1 is a diagram showing the manner in which a new picture is interpolated between successive faces in accordance with an embodiment of the present invention. As shown in Fig. 1, when the motion vectors of the two consecutive pictures A and B are estimated, the relevant motion information of the continuous faces A and B can be obtained, and then the new face C can be based on the faces A and B. Interpolation of relevant sports information. Figure 2 is a flow chart showing a method of converting the face frequency of a video signal in accordance with an embodiment of the present invention. First, a pulldown image sequence is received (step 202) 'where the folded image sequence is present or converted from a written sequence of one of the video signals for a continuous time' and includes a plurality of mutually different original images, and each There are corresponding numbers and repeated faces on the original surface. Next, the original picture in the folded image sequence is modified (step 204). Thereafter, at least one motion vector associated with the modified original facet is estimated (step 2〇6). Then, a new picture is interpolated between the modified original pictures based on the estimation result of the motion vector (step 208). The following is an example to illustrate the case of converting the 6 Hz gold frequency to the face frequency using the above conversion method. The first to the <RTIgt;&> diagrams illustrate the process of converting an original image sequence into a new image sequence by the method as shown in Figure 2, in accordance with an embodiment of the present invention. As shown in Figure 3a, the 3:2 folded image sequence iii 7 7 j 3 • 13 13..., which is present in or converted from the continuous facet sequence, 201028017 and having a 60 Hz facet frequency, is received first, where the numbers 1, 7, 13 The equations exemplarily represent the position (or order) of the faces in the sequence. Next, the original picture in the folded image sequence is modified. In the present embodiment, the action of removing the repeated faces of the original facet (i.e., Bu, 7, etc.) in the folded image sequence (as shown in Fig. 3b). After the repeated facets in the original facet are removed, the motion vectors of the two adjacent faces (ie 2, 7, 7, and 13, etc.) • (4) of the remaining faces in the original picture are then estimated. In the present embodiment, a new picture of the continuous time is interpolated between the two adjacent pupils remaining in the original pupil based on the motion vector. For example, as shown in Fig. 3c, the facet 23 4 5 6 is interpolated between the faces 1 and 7, while the face 789HUH2 is interpolated between the screens 7 and 13, and so on. Then, a new face sequence 1 2 3 4 5 6 7 8 9 10 11 12 13 ... having a face frequency of 120 Hz can be generated (as shown in Fig. 3d). 4a to 4d are diagrams showing a process of converting an original image sequence into a new image sequence by a conversion method as shown in Fig. 2 according to another embodiment of the present invention. Similarly, as shown in Fig. 4a, a 3:2 folded image sequence M丨77 13 13 13 existing or converted from a continuous picture sequence and having a 6-plane frequency is received first. Next, the original facets in the f-stack image sequence are modified. In this embodiment, the method is to selectively repeat each of the mutually different original faces, so that each of the different original faces has the same number = 昼 I There are two repeated faces in both m丨 and 13, and there are two repeated pictures in the face 7, 201028017 = enter the line first: and the face 1 and 13 have the same Once in the face (as shown in Figure 4b). After the correction = = ; = after the change, then the estimated and the modified picture (ie 1 and 7, 7, and 13, Rong U day off = motion vector is estimated. In this embodiment, the method is After estimating the difference and the motion vector, in the modified original picture, between the adjacent two different planes, 'interpolate several new pictures according to the motion vector. It is worthy of phonetic: Yes, the number of new face interpolation Less than the original face; "Quantity. For example, as shown in Figure 4c, the picture = adjacent screens 1 and 7 goes to you between _ 7 and the 10 10 lines are interpolated Between adjacent pictures and between, and so on; the number of 昼4, 10, ..., etc. 丄 7, 13, ..., etc. Then, a new 3:2 fold with a picture frequency of 120 便可 can be generated. Image sequence 10 7 10 10 13 13 13... (as shown in Fig. 4d). It is worth noting here that the embodiment of Example 3: ❹ sequence is implemented to illustrate the invention, not In order to use the sequence, a 2:2 folded image sequence or other folded image sequence can also be made in the present invention. The frequency of the present day is as follows - the actual __ type conversion view Flowchart of the method of picture frequency. First, the face sequence existing or converted from the video signal is received (step 502). Then, motion estimation (step step) is performed on the t-th column of the facet. ; =. Fruit (4) The pure picture sequence is folded as image 506), wherein the folded image sequence includes a plurality of faces, and each of the original pictures has a corresponding number and 201028017 faces, for example, Figure 3a The 3:2 folded image sequence shows 77ΐ3 =3 .... If the sequence is a sequence rather than a folded image, the sequence of pictures will be compensated by the general motion: (Step 2:5-8). The facet sequence is adjusted and modified for the original facet of the folded image sequence (step 51〇). After that, at least one is related to the modified original picture: after estimation (step 5... then 'based on the motion vector The estimated two-input interpolation is between the modified original pictures (step 514). It is worth noting that the steps 51〇, 512, and Η# in this embodiment are similar to the second figure. Steps 2〇4, 2〇6 and 2〇8 shown in It can be seen from the above embodiments of the present invention that the application of the aforementioned converted picture frequency can be used to rotate a picture sequence having a high picture frequency (eg, 6 GHz), and /, to have a more south picture frequency (eg, 12 〇 Hz). New screen order to improve the display quality of the image. - Although the present invention has been disclosed above by way of example, it is not intended to limit the invention, and any one of ordinary skill in the art to which the invention pertains may not deviate from the present invention. In the spirit and scope of the invention, the various modifications and refinements may be made, and therefore the scope of the invention is defined by the scope of the appended claims. [FIG. 1] A method of interpolating a new face between successive pictures in an embodiment of the invention. _ Figure 2 is a flow chart showing a method of converting the picture frequency of a video signal in accordance with an embodiment of the present invention. 8 201028017 Figures 3a through 3d illustrate the process of converting an original image sequence into a new image sequence by a conversion method as shown in Figure 2, in accordance with an embodiment of the present invention. 4a through 4d illustrate a process of converting an original image sequence into a new image sequence by a conversion method as shown in Fig. 2, in accordance with another embodiment of the present invention. Figure 5 is a flow chart showing a method of converting the face frequency of a video signal in accordance with another embodiment of the present invention. ❹ [Main component symbol description] 202~208, 502~514: Step