201225000 六、發明說明： 【發明所屬之技術領域】 本發明係關於一種用以實施圖像之邊緣強調處理之圖像 處理裝置。 . 【先前技術】 自先前以來提出有各種用以實施圖像之邊緣強調處理之 圖像處理裝置。 例如’於專利文獻1或非專利文獻1中，揭示有一種使用 Wavelet(小波）轉換實施圖像之邊緣強調處理之技術。 圖8係表示使用有先前之Wavelet轉換之圖像處理裝置 700之概略的方塊圖。 上述圖像處理裝置700中’如圖8所示，於Wavelet轉換 部701中對輸入圖像資料img_in進行waveiet轉換而生成4個 次頻帶（imgJLL、img_HL、img_LH、img—ΗΗ)。 次頻帶img_LL係對輸入之圖像資料img於水平方向及垂 直方向均使用低通濾波器而生成。 次頻帶img_HL係對輸入之圖像資料img於水平方向使用 高通濾波器、於垂直方向使用低通濾波器而生成。 • 次頻帶img_LH係對輸入之圖像資料img於水平方向使用 • 低通濾波器、於垂直方向使用高通濾波器而生成》 次頻帶係對輸入之圖像資料img於水平方向及垂 直方向均使用高通濾波器而生成。 所生成之4個次頻帶於後段之逆Wavelet轉換部703中進 行逆Wavelet轉換。此時，對所生成之4個次頻帶中次頻帶 159526.doc 201225000 img_LL以外之次頻帶於乘法部702中分別乘以強調係數 KHl、KLH、KHH之後，於逆Wavelet轉換部703中進行逆 Wavelet轉換。 如上所述，經逆Wavelet轉換之圖像成為進行了邊緣強 調之圖像。 [先前技術文獻] [專利文獻] [專利文獻1]曰本公開專利公報「日本專利特開2007-280202號公報（2007年10月25日公開）」 [非專利文獻] [非專利文獻 1 ] Adaptive image denoising and edge enhancement in scale-space using the wavelet transform，Pattern Recognition Letters 24(3) pp. 965-971, 2003 【發明内容】 [發明所欲解決之問題] 以下，一面參照圖9(a)〜圖9(d)，一面對圖8所示之圖像 處理裝置之邊緣強調處理進行說明。此處，於圖9(a)〜圖 9(d)中，橫軸表示圖像之像素位置，縱軸表示亮度值。 首先，將對圖9(a)所示之輸入之圖像資料（原信號）進行 Wavelet轉換而得之L次頻帶進行逆Wavelet轉換而得之結 果成為圖9(b)所示之曲線圖，且將對上述資料進行Wavelet 轉換而得之Η次頻帶進行逆Wavelet轉換而得之結果成為圖 9(c)所示之曲線圖。 此時，圖9(b)中表示將Η次頻帶設為零而輸入之狀態， 159526.doc 201225000 圖9(c)中表示將L次頻帶設為零而輸入之狀態。 其次，藉由將圖9(b)、圖9(c)所示之結果加以合成而獲 知·對原彳s號進行了邊緣修正之結果。此時之邊緣修正結果 如圖9(d)所示之曲線圖般。 * 且說’於先前之邊緣強調處理中，將經Wavelet轉換之 - 次頻帶中高頻率之Η次頻帶與強調係數相乘，並藉由逆201225000 VI. Description of the Invention: [Technical Field] The present invention relates to an image processing apparatus for performing edge emphasis processing of an image. [Prior Art] Various image processing apparatuses for performing edge enhancement processing of an image have been proposed from the past. For example, in Patent Document 1 or Non-Patent Document 1, there is disclosed a technique for performing edge emphasis processing of an image using Wavelet (wavelet) conversion. Fig. 8 is a block diagram showing an outline of an image processing apparatus 700 using a previous Wavelet conversion. In the image processing apparatus 700, as shown in Fig. 8, the Wavelet conversion unit 701 performs waveiet conversion on the input image data img_in to generate four sub-bands (imgJLL, img_HL, img_LH, img_ΗΗ). The sub-band img_LL is generated by using a low-pass filter for the input image data img in both the horizontal direction and the vertical direction. The sub-band img_HL is generated by using a high-pass filter in the horizontal direction and a low-pass filter in the vertical direction. • The sub-band img_LH is used for the input image data img in the horizontal direction. • The low-pass filter is generated by using the high-pass filter in the vertical direction.” The sub-band is used for the input image data img in both the horizontal and vertical directions. Generated by a high pass filter. The generated four sub-bands are subjected to inverse Wavelet conversion in the inverse Wavelet converter 703 of the subsequent stage. At this time, the sub-bands other than the sub-bands 159526.doc 201225000 img_LL in the generated four sub-bands are multiplied by the emphasis coefficients KH1, KLH, and KHH in the multiplication unit 702, respectively, and then inverse Wavelet is performed in the inverse Wavelet conversion unit 703. Conversion. As described above, the image converted by the inverse Wavelet becomes an image with an edge emphasis. [PRIOR ART DOCUMENT] [Patent Document 1] [Patent Document 1] Japanese Laid-Open Patent Publication No. 2007-280202 (published on Oct. 25, 2007) [Non-Patent Document] [Non-Patent Document 1] Adaptive image denoising and edge enhancement in scale-space using the wavelet transform, Pattern Recognition Letters 24(3) pp. 965-971, 2003 [Summary of the Invention] [Problems to be Solved by the Invention] Hereinafter, reference is made to FIG. 9(a). 9(d), an edge emphasis processing of the image processing apparatus shown in FIG. 8 will be described. Here, in Figs. 9(a) to 9(d), the horizontal axis represents the pixel position of the image, and the vertical axis represents the luminance value. First, the result of inverse Wavelet conversion of the L-th frequency band obtained by Wavelet conversion of the input image data (original signal) shown in FIG. 9(a) is a graph shown in FIG. 9(b). The result of inverse Wavelet conversion of the sub-band obtained by Wavelet conversion of the above data becomes the graph shown in FIG. 9(c). At this time, FIG. 9(b) shows a state in which the sub-band is set to zero and is input. 159526.doc 201225000 FIG. 9(c) shows a state in which the L-th sub-band is set to zero and is input. Next, by synthesizing the results shown in Figs. 9(b) and 9(c), it is known that the original s number has been subjected to edge correction. The edge correction result at this time is the same as the graph shown in Fig. 9(d). * and say that in the previous edge emphasis processing, the frequency band of the high frequency in the sub-band of the Wavelet conversion is multiplied by the emphasis coefficient, and by the inverse
Wavelet轉換而進行邊緣強調。因此，圖9(c)之η次頻帶之 信號成分中產生之振動得以強調，並如圖9(d)所示，於邊 緣修正結果中產生不自然之振動。亦即，產生如下問題： 先前之邊緣強調處理會導致不自然之邊緣強調結果。 本發明係鑒於上述課題而完成者，其目的在於提供一種 圖像處理裝置，其於進行Wavelet轉換之邊緣強調處理 中，可實現獲得邊緣修正結果中不產生振動之自然之邊緣 強調結果的邊緣強調處理。 [解決問題之技術手段] 為解決上述課題，本發明之圖像處理裝置係對輸入圖像 資料進行Wavelet轉換並進行圖像之邊緣強調處理者，其 特徵在於包含：Wavelet轉換部，其自輸入圖像資料生成 低頻側之次頻帶與高頻側之次頻帶；碼檢測部，其自藉由 ‘ Wavelet轉換部生成之低頻側之次頻帶而檢測表示水平、 垂直、傾斜方向之凹凸傾向的凹凸碼；絕對值運算部，其 求出上述高頻側之次頻帶之絕對值；及強調處理部，其使 藉由上述絕對值運算部求出之絕對值、與由上述碼檢測部 檢測之凹凸碼值相乘而進行圖像之邊緣之強調處理；且輪 159526.doc 201225000 出使藉由上述Wave丨et轉換部生成之低頻側之次頻帶、與 由上述強調處理部獲得之乘法值相加而得之圖像。 、 又，本發明之圖像處理方法係對輸入圖像資料進行 Wavelet轉換並進行圖像之邊緣強調處理者，其特徵在於 包含以下步驟：WaveIet轉換步驟，自輸入圖像資料生成 低頻側之次頻帶與高頻側之次頻帶；碼檢測步驟，自藉由Wavelet transforms for edge emphasis. Therefore, the vibration generated in the signal component of the n-th sub-band of Fig. 9(c) is emphasized, and as shown in Fig. 9(d), an unnatural vibration is generated in the edge correction result. That is, the following problems arise: The previous edge emphasis processing leads to an unnatural edge that emphasizes the result. The present invention has been made in view of the above problems, and an object thereof is to provide an image processing apparatus capable of realizing an edge emphasis of a natural edge enhancement result in which no vibration is generated in an edge correction result in edge enhancement processing of Wavelet conversion. deal with. [Means for Solving the Problems] In order to solve the above problems, the image processing device of the present invention performs Wavelet conversion on input image data and performs edge enhancement processing on an image, and is characterized in that it includes a Wavelet conversion unit that is self-input The image data generates a sub-band on the low-frequency side and a sub-band on the high-frequency side, and the code detecting unit detects the unevenness indicating the unevenness in the horizontal, vertical, and oblique directions from the sub-band on the low-frequency side generated by the Wavelet conversion unit. An absolute value calculation unit that obtains an absolute value of the sub-band on the high-frequency side, and an emphasis processing unit that obtains an absolute value obtained by the absolute value calculation unit and the unevenness detected by the code detection unit The code values are multiplied to perform edge processing of the image; and the wheel 159526.doc 201225000 outputs the sub-band of the low-frequency side generated by the Wave 丨et conversion unit, and the multiplication value obtained by the emphasis processing unit And get the image. Moreover, the image processing method of the present invention performs Wavelet conversion on the input image data and performs edge enhancement processing on the image, and is characterized in that the method includes the following steps: a WaveIet conversion step, generating a low frequency side from the input image data. Subband of frequency band and high frequency side; code detection step, by self-solving
WaVelet轉換步驟而生成之低頻側之次頻帶檢測表示水 平、垂直、傾斜方向之凹凸傾向的凹凸碼；絕對值運算步 驟’求出上述高頻側之次頻帶之絕對值；強調處理步驟， 使藉由上述絕對值運算步㈣&之絕對值、與±述碼檢測 步驟中檢測之凹凸碼相乘而進行圖像之邊緣之強調處理； 及輸出步驟，輸出使藉由上述Wavelet轉換步驟生成之低 頻側之次頻帶、與上述強調處理步驟中獲得之乘法值相加 而得之圖像。 根據上述構成，對藉由Wavelet轉換而生成之高頻側之 次頻帶求出絕對值，並使該絕對值、肖自藉由轉 換而生成之低頻側之次頻帶檢測之凹凸碼值相乘，藉此實 施向頻側之次頻帶中產生之振動得以抑制之邊緣的強調處 理。然後，使藉由上述Wavelet轉換而生成之低頻側之次 頻帶、與在振動得以抑制之狀態下實施強調處理後之高頻 側的次頻帶相加，藉此可獲得無不自然之振動之自然的邊 緣強調結果、即實施了线的邊緣強調處理之輸出圖像。 較佳為於將表示上述邊緣強調處理中之強調之程度的 係數設為強調係數時’上述強調處理部使藉由上述絕對值 159526.doc 201225000 運算部求出之絕對值、由上述碼檢測部檢測之凹凸碼值、 及上述強調係數相乘而進行圖像之邊緣之強調處理。 較佳為’上述圖像處理裝置進而包含乘法部，於將表示 上述邊緣強調處理中之強調之程度的係數設為強調係數 時，該乘法部使藉由Wavelet轉換部生成之高頻側之次頻 帶與上述強調係數相乘，且上述絕對值運算部求出藉由上 述乘法部而乘以強調係數後之高頻側之次頻帶的絕對值。 如上所述，使高頻側之次頻帶、與表示邊緣強調處理中 之強調之程度的係數即強調係數相乘，藉此可控制邊緣強 調之程度、即邊緣強調處理之效果。亦即，藉由變更上述 強調係數之值而可容易地控制邊緣強調處理之效果。 再者，乘以上述強調係數之時點可為次頻帶之絕對值運 算前後之任一時點。 上述圖像處理裝置包含對上述輸入圖像資料實施放大或 縮小處理之縮放處理部，上述Wavelet轉換部對藉由上述 縮放處理部而實施了放大縮小處理之圖像資料進行 Wavelet轉換。 根據上述構成，Wavelet轉換部對藉由上述縮放處理部 而實施了放大縮小處理之圖像資料進行轉換，藉 此可同時進行放大•縮小處理與邊緣強調處理。 通常’於已進行Wavelet轉換之情形時，輸入圖像資料 之尺寸成為-半之尺寸’故而於進行―·轉換之前， 必需以使輸入圖像資料之尺寸成為2倍尺寸之方式實施放 大處理。 159526.doc 201225000 因此，較佳為上述縮放處理部中實施將輸入圖像資料之 尺寸放大至2倍之放大處理。 又，為使輸出圖像之尺寸成為輸入圖像（原圖像）之尺寸 或成為作為目標之圖像之尺寸，亦可對原圖像之尺寸進行 放大或縮小處理。 上述同時處理於例如將SD(Standard Definhi〇n，標準解 析度）尺寸之圖像放大處理為FULL hd(full h㈣The sub-band detection on the low-frequency side generated by the WaVelet conversion step detects the concavo-convex code indicating the unevenness in the horizontal, vertical, and oblique directions; the absolute value calculation step 'determines the absolute value of the sub-band on the high-frequency side; emphasizes the processing steps, lends And performing an emphasis process on the edge of the image by multiplying the absolute value of the absolute value operation step (4) & and the embossing code detected in the code detection step; and outputting the output to output the low frequency generated by the Wavelet conversion step The sub-band of the side, the image obtained by adding the multiplication value obtained in the above-mentioned emphasis processing step. According to the above configuration, the absolute value is obtained for the sub-band on the high-frequency side generated by the Wavelet conversion, and the absolute value and the emboss code value of the sub-band detection on the low-frequency side generated by the conversion are multiplied. Thereby, emphasis processing for suppressing the edge generated by the vibration generated in the sub-band of the frequency side is performed. Then, the sub-band on the low-frequency side generated by the Wavelet conversion is added to the sub-band on the high-frequency side after the emphasis processing is performed in a state where the vibration is suppressed, whereby the natural vibration without the unnatural vibration can be obtained. The edge emphasizes the result, that is, the output image of the edge emphasis processing of the line is implemented. Preferably, when the coefficient indicating the degree of emphasis in the edge emphasis processing is the emphasis coefficient, the emphasis processing unit obtains the absolute value obtained by the calculation unit by the absolute value 159526.doc 201225000, and the code detecting unit The detected concavo-convex code value and the above-mentioned emphasis coefficient are multiplied to perform emphasis processing on the edge of the image. Preferably, the image processing apparatus further includes a multiplication unit that sets the coefficient indicating the degree of emphasis in the edge enhancement processing to an emphasis coefficient, and the multiplication unit causes the high frequency side generated by the Wavelet conversion unit. The frequency band is multiplied by the emphasis coefficient, and the absolute value calculation unit obtains an absolute value of the sub-band of the high-frequency side obtained by multiplying the emphasis coefficient by the multiplication unit. As described above, by multiplying the sub-band on the high-frequency side by the emphasis coefficient which is a coefficient indicating the degree of emphasis in the edge emphasis processing, the degree of edge emphasis, that is, the effect of the edge emphasis processing can be controlled. That is, the effect of the edge emphasis processing can be easily controlled by changing the value of the above-described emphasis coefficient. Furthermore, the time point multiplied by the above-mentioned emphasis coefficient may be any one of the time points before and after the absolute value of the sub-band. The image processing device includes a scaling processing unit that performs enlargement or reduction processing on the input image data, and the Wavelet conversion unit performs Wavelet conversion on the image data subjected to the enlargement and reduction processing by the scaling processing unit. According to the above configuration, the Wavelet conversion unit converts the image data subjected to the enlargement and reduction processing by the scaling processing unit, whereby the enlargement/reduction processing and the edge enhancement processing can be simultaneously performed. In general, when the Wavelet conversion has been performed, the size of the input image data is - half the size. Therefore, before the conversion is performed, it is necessary to perform the enlargement processing so that the size of the input image data becomes twice the size. 159526.doc 201225000 Therefore, it is preferable that the above-described scaling processing unit performs an enlargement process of enlarging the size of the input image data by a factor of two. Further, in order to make the size of the output image the size of the input image (original image) or the size of the target image, the size of the original image may be enlarged or reduced. The above simultaneous processing is, for example, amplifying an image of SD (Standard Definhi〇n, standard resolution) size to FULL hd (full h (4)
Definition全高解析度）尺寸之圖像並顯示之情形般因解像 度轉換而產生邊緣模糊之問題的情形時較為有效。 藉由使用本發明之圖像處理裝置，對於沉心格式之彩 色圖像資料、RGB格式之彩色圖像資料亦可實施自然的邊 緣強調處理。具體之構成如下所述。 為解决上述課題，本發明之圖像處理裝置係對格 式之彩色圖像資料騎Wavelet轉換並進行圖像之邊緣強 調處理者’其特徵在於包含：輸出對上述Y成分之圖像資 料進行了邊緣強調處理之Y成分之圖像的圖像處理部；輸 出對上述Cb成分之圖像資料進行了放大•縮小處理之⑽ 分=圖像的縮放處理部；及輸出對上述㈣分之圖像資料 、T 了放大縮小處理之Cr成分之圖像的縮放處理部；且 上述圖像處理部係、藉由上述圖像處理裝置而實現。 根據上述構成，可獲得對YCbCr格式之彩色圖像資料實 施了使圖像之邊緣無振動之自然的邊緣強調處理之輸出彩 色圖像。 為解决上述課題，本發明之圖像處理裝置係對格式 159526.doc 201225000 之彩色圖像資料進行WaVelet轉換並進行圖像之邊緣強調 處理者’其特徵在於包含：冑出對上述尺成分之圖像資料 進行了邊緣強域理之Μ分之圖像的圖像處理部；輸出 對上述G成分之圖像資料進行了邊緣強調處理之G成分之 圖像的圖像處理部；及輸出對上述Β成分之圖像資料進行 了邊緣強調處理之Β成分之圖像的圖像處理部；且上述全 部圖像處理部係藉由上述圖像處理裝置而實現。 根據上述構成，可獲得對RGB格式之彩色圖像資料實施 了使圖像之邊緣無振動之自然的邊緣強調處理之輸出彩色 圖像。 作為具備上述構成之圖像處理裝置之電子機器，若係需 要對輸入圖像實施邊緣處理之電子機器，則可適用任意之 電子機器可舉出例如電視接收機、電子書終端、行動終 端裝置等。 ' [發明之效果] 本發明之圖像處理裝置係對輸入圖冑資料進行Wavelet 轉換並進行圖像之邊緣強調處理者，其構成為包含： Wavelet轉換部’其自輸入圖像資料生成低頻側之次頻帶 與高頻侧之次頻帶；碼檢測部，其自藉由Wavelet轉換部 而生成之低頻侧之次頻帶檢測表示水平、垂直、傾斜方向 之凹凸傾向的凹凸碼；絕對值運算部，其求出上述高頻側 之人頻帶之絕對值，及強調處理部，其使藉由上述絕對值 運算部求出之絕對值、與由上述碼檢測部檢測之凹凸碼值 相乘而進行圖像之邊緣之強調處理；且輸出使藉由上述 159526.doc 201225000It is effective to use the case where the image of the full-resolution image is displayed and the edge blurring occurs due to the resolution conversion. By using the image processing apparatus of the present invention, natural edge enhancement processing can be performed for the color image data of the sinking format and the color image data of the RGB format. The specific constitution is as follows. In order to solve the above problems, the image processing apparatus of the present invention is characterized in that the color image data of the format is converted by Wavelet and the edge of the image is processed by the image processing device, which is characterized in that: the output of the image data of the Y component is edged. An image processing unit that emphasizes an image of the processed Y component; an image processing unit that enlarges and reduces the image data of the Cb component; (10) points = image scaling processing unit; and outputs image data of the above (four) points And a scaling processing unit for the image of the Cr component of the enlargement and reduction processing; and the image processing unit is realized by the image processing device. According to the above configuration, it is possible to obtain an output color image in which the color image data of the YCbCr format is subjected to natural edge enhancement processing for making the edges of the image free from vibration. In order to solve the above problems, the image processing apparatus of the present invention performs WaVelet conversion on the color image data of the format 159526.doc 201225000 and performs edge enhancement processing on the image, which is characterized in that: the image of the scale component is extracted. An image processing unit that performs an image of a marginal strong-domain image, and an image processing unit that outputs an image of a G component that performs edge enhancement processing on the image data of the G component; The image processing unit that performs the image of the component of the edge enhancement processing on the image data of the Β component; and all of the image processing units described above are realized by the image processing device. According to the above configuration, it is possible to obtain an output color image in which the color image data of the RGB format is subjected to natural edge enhancement processing for making the edges of the image free from vibration. As an electronic device including the image processing device having the above-described configuration, if an electronic device that performs edge processing on an input image is required, any electronic device can be applied, for example, a television receiver, an e-book terminal, a mobile terminal device, or the like. . [Effect of the Invention] The image processing apparatus of the present invention performs Wavelet conversion on the input image data and performs edge enhancement processing on the image, and is configured to include: a Wavelet conversion unit that generates a low frequency side from the input image data The sub-band and the sub-band of the high-frequency side; the code detecting unit detects the concavo-convex code indicating the unevenness in the horizontal, vertical, and oblique directions from the sub-band on the low-frequency side generated by the Wavelet converter; the absolute value calculation unit The absolute value of the human frequency band on the high frequency side is obtained, and the emphasis processing unit performs multiplication by the absolute value obtained by the absolute value calculation unit and the concave and convex code value detected by the code detecting unit. Like the edge of the emphasis processing; and the output is made by the above 159526.doc 201225000
Wavelet轉換部生成之低頻側之次頻帶、與由上述強調處 理部獲得之乘法值相加而得的圖像。藉此，發揮如下效 果：可獲得實施了無不自然之振動之自然的邊緣強調處理 之輸出圖像。 【實施方式】 以下，對本發明之實施形態詳細地進行說明。 (實施形態1) 圖1係本貫施形態1之圖像處理裝置1 〇〇之概略構成方塊 圖。 如圖1所示’上述圖像處理裝置100為如下構成，即包含 比例放大部（縮放處理部）1 〇 1、Wavelet轉換部1 〇2、絕對值 運算部103、碼檢測部1 〇4、及強調處理部丨〇5。 即，圖像處理裝置1〇〇係對輸入圖像資料img一in進行 Wavelet轉換並進行圖像之邊緣強調處理的裝置，其包 含：比例放大部ιοί，其對輸入圖像資料img_in實施比例 放大處理；Wavelet轉換部102,其自藉由比例放大部ι〇ι 實把了比例放大處理之圖像img_up生成低頻側之次頻帶與 尚頻側之次頻帶；碼檢測部l04，其自藉由轉換部 102而生成之低頻側之次頻帶檢測表示水平、垂直、傾斜 方向之凹凸傾向的凹凸石馬；絕對值運算部1〇3，其求出藉 由Wavelet轉換部生成之向頻側之次頻帶之絕對值；及強 調處理部105，其使藉由上述絕對值運算部⑻求出之絕對 值、由上述碼檢測部刚檢測之凹凸丨、及預先設定之強 調係數相乘，且輸出使藉由上述Waveh轉換部生成之 159526.doc -10- 201225000 低頻側之次頻帶、與由上述乘法部獲得之乘法值於合成部 1 06中相加而得之輸出圖像資料img_out。 上述比例放大部101對輸入圖像資料img in實施比例放 大處理（放大或縮小處理）。 此處，上述比例放大部101可實現以下方法：方法（丨）， 將輸入圖像資料img_in放大至2倍，使輸出圖像資料 img 一 out之尺寸與輸入圖像資料img_in之尺寸相同；及方法 (2) ’將輸入圖像資料img_in放大至目標圖像尺寸之2倍， 以使輸出圖像資料img—out之尺寸與目標圖像之尺寸相 同。作為該比例放大部101中所用之放大方法，使用 Lanczos方法。然而，作為放大方法，並不限定於Lanczos 方法’亦可為其他放大方法。 再者’本實施形態1及下述各實施形態中，均對應用上 述（1)之方法’即’將輸入圖像資料丨111§—in放大至2倍以使 輸出圖像資料img—out之尺寸與輸入圖像資料img—in之尺寸 相同之方法的例進行說明。 藉由上述比例放大部1 〇丨進行了放大處理之輸入圖像資 料img_in係作為比例放大圖像img_Up而輸出至waveiet^ 換部102。 如圖2所示，上述Wavelet轉換部ι〇2對輸入之比例放大 圖像img一up作為原圖像，對原圖像img使用成對之低通濾 波益、高通濾波器而生成4個次頻帶（irng—LL、img_HL、 img_LH、img_HH)。此處，Wavelet轉換部 1〇2 中，使用 CDF9/7該種Wavelet轉換方式。但是，亦可為除此以外之 159526.doc 201225000The sub-band on the low-frequency side generated by the Wavelet conversion unit and the image obtained by adding the multiplication value obtained by the above-described emphasis processing unit. Thereby, an effect is obtained in which an output image of the natural edge enhancement processing in which no unnatural vibration is performed can be obtained. [Embodiment] Hereinafter, embodiments of the present invention will be described in detail. (Embodiment 1) Fig. 1 is a block diagram showing a schematic configuration of an image processing apparatus 1 according to a first embodiment. As shown in FIG. 1 , the image processing apparatus 100 includes a scale enlargement unit (scaling processing unit) 1 , a Wavelet conversion unit 1 , an absolute value calculation unit 103 , and a code detection unit 1 〇 4 . And emphasize the processing unit 丨〇5. That is, the image processing apparatus 1 is a device that performs Wavelet conversion on the input image data img_in and performs edge enhancement processing of the image, and includes a proportional enlargement unit ιοί, which scales the input image data img_in The Wavelet conversion unit 102 generates the sub-band on the low frequency side and the sub-band on the low frequency side from the image img_up of the proportional enlargement process by the proportional enlargement unit ι〇ι; the code detecting unit 104, by itself The sub-band on the low-frequency side generated by the conversion unit 102 detects the unevenness of the horizontal, vertical, and oblique directions, and the absolute value calculation unit 1〇3 obtains the frequency-by-frequency side generated by the Wavelet conversion unit. An absolute value of the frequency band; and an emphasis processing unit 105 that multiplies the absolute value obtained by the absolute value calculation unit (8), the unevenness detected by the code detecting unit, and a predetermined emphasis coefficient, and outputs the same The output image data img_ou obtained by adding the multiplication band of the 159526.doc -10- 201225000 low frequency side generated by the Waveh conversion unit and the multiplication value obtained by the above multiplication unit to the synthesis unit 106 t. The proportional enlargement unit 101 performs a proportional enlargement process (enlargement or reduction process) on the input image data img in. Here, the above-mentioned proportional amplification unit 101 can implement the following method: method (丨), the input image data img_in is enlarged to 2 times, so that the size of the output image data img_out is the same as the size of the input image data img_in; Method (2) 'Enlarge the input image data img_in to twice the size of the target image so that the size of the output image data img_out is the same as the size of the target image. As the amplification method used in the proportional amplification section 101, the Lanczos method is used. However, the amplification method is not limited to the Lanczos method' and may be other amplification methods. Furthermore, in the first embodiment and the following embodiments, the method of the above (1) is applied, that is, the input image data 丨111§-in is enlarged to 2 times so that the output image data is img-out. An example of a method in which the size is the same as the size of the input image data img_in will be described. The input image data img_in which has been enlarged by the above-described proportional enlargement unit 1 is output to the waveiet conversion unit 102 as the scaled up image img_Up. As shown in FIG. 2, the Wavelet conversion unit ι2 generates an up-converted image img-up as an original image, and generates a pair of low-pass filter and high-pass filters for the original image img. Band (irng-LL, img_HL, img_LH, img_HH). Here, in the Wavelet conversion unit 1〇2, the Wavelet conversion method of CDF9/7 is used. However, it can also be 159526.doc 201225000
Wavelet轉換方式，並不限定於CDF9/7。 次頻帶img_LL係對輸入之圖像資料img於水平方向及垂 直方向均使用低通濾波器而生成。 次頻帶img一HL係對輸入之圖像資料img於水平方向使用 高通渡波器、於垂直方向使用低通渡波器而生成。 次頻帶img_LH係對輸入之圖像資料img於水平方向使用 低通濾波器、於垂直方向使用高通濾波器而生成。 次頻帶img_HH係對輸入之圖像資料img於水平方向及垂 直方向均使用高通濾波器而生成。The Wavelet conversion method is not limited to CDF9/7. The sub-band img_LL is generated by using a low-pass filter for the input image data img in both the horizontal direction and the vertical direction. The sub-band img-HL system generates the image data img of the input using the high-pass waver in the horizontal direction and the low-pass waver in the vertical direction. The sub-band img_LH is generated by using a low-pass filter in the horizontal direction and a high-pass filter in the vertical direction. The sub-band img_HH is generated by using a high-pass filter for the input image data img in both the horizontal direction and the vertical direction.
此處，次頻帶img一LL係於水平方向、垂直方向之兩方 向上使用低通濾波器而生成，故而係尺寸縮小至一半之圖 像。次頻帶img_HL表示縱向之邊緣資訊，次頻帶img LH 表示橫向之邊緣資訊，次頻帶img一HH表示傾斜方向之邊 緣資訊。 上述Wavelet轉換部102將4個次頻帶中高頻側之次頻帶 (次頻帶img_HL、次頻帶img_LH、次頻帶img—HH)轉送至 絕對值運算部103。 上述絕對值運算部103求出所輸入之各次頻帶之絕對值 並轉送至後段之強調處理部1 〇5。 另一方面，上述Wavelet轉換部1〇2將作為低頻侧之次頻 帶之次頻帶img_LL轉送至碼檢測部104與合成部1〇6。 示之7抽頭濾波器自 具體而言，對作為輸 、傾斜之方向上分別 上述碼檢測部104中使用例如圖4所 所輸入之次頻帶img_LL進行碼檢測。 入圖像之次頻帶img_LL於水平、垂直 159526.doc 12 201225000 使用濾波器而檢測其值之碼(圖1所示之 . sign-di)。亦即，碼檢測物4中檢測之^=、slgn-vt、 次頻帶於水平方a #古士 J之碼係表示低頻側之 凸碼。此處“ D料方向之凹凸傾向的凹 匕處，於表不像素位置與亮度 (例如圓c,u、、丄 u度值之關係之曲線圖 (J如圖5(b))中，自下方觀察 出正作A踩 H線圓’於凹曲線部分輸 K止作為碼，於凸部分輸出 零。 ；K貞作為碼。於平坦部分輸出 上述碼檢測結果用於強調處理部丨〇5之運算中。 上述強調處理部105使自、絕對值運算部1〇；'料之高頻侧 之各次頻帶的絕對值、上述碼檢測部⑽之碼檢測結果、 及預先設定之強調係數相乘。 具體而言，使次頻帶之絕對值、碼sign_hz、及 強調係數kHL相乘，使次頻帶img_LH之絕對值、碼 gn~Vt及強調係數kLH相乘，使次頻帶img—HH之絕對 值碼sign-di、及強調係數kHH相乘。各相乘結果被轉送 至合成部106。 上述強調係數kHL ' kLH、kHH係用以決定邊緣強調處理之 效果之程度的係數。藉由變更該強調係數之值而可容易地 控制邊緣強調處理之效果。 又，強調係數kHL、kLH、kHH可預先設定，亦可根據圖像 内容而適當決定。此處，使用預先設定之強調係數。 上述合成部106使自Wavelet轉換部102轉送之次頻帶 img_LL與自強調處理部1〇5轉送之次頻帶img_HL、次頻帶 img_LH、及次頻帶img_HHi相乘結果相加，並作為輪出 159526.doc -13- 201225000 圖像資料img_out而輸出。 根據上述構成之圖像處理裝置1〇〇，對藉由Wavelet轉換 部102生成之高頻側之次頻帶求出絕對值，並使該絕對 值、與自藉由Wavelet轉換部102而生成之低頻側之次頻帶 檢測之凹凸碼值相乘，藉此可消除高頻側之次頻帶中產生 之振動。使該狀態之高週波側之次頻帶進而與強調係數相 乘’因此於消除振動之狀態下實施強調處理。然後，使藉 由上述Wavelet轉換部102生成之低頻側之次頻帶、與消除 振動之狀態下實施了強調處理之高頻側之次頻帶相加，藉 此獲得實施了無不自然感之自然的邊緣強調處理之輸出圖 像。 (實施形態2) 圖4係本實施形態2之圖像處理裝置200之概略構成方塊 圖。 如圖4所示’上述圖像處理裝置200為如下構成，即包含 比例放大部201、Wavelet轉換部202、絕對值運算部203、 碼檢測部204、強調處理部205、合成部206、及乘法部 207 〇 此處’上述圖像處理裝置200為與圖1所示之圖像處理事 置100大致相同之構成’但乘以強調係數之場所不同。圖 像處理裝置200中，於求出藉由Wavelet轉換部202而生成 之高頻側之次頻帶（次頻帶img_HL、次頻帶img_LH、次頻 帶img一HH)之絕對值之前，藉由乘法部207而乘以強調係 數。 159526.doc -14- 201225000 具體而言，圖像處理裝置200係對輸入圖像資料img_in 進行Wavelet轉換並進行圖像之邊緣強調處理之圖像處理 裝置’其包含：比例放大部2〇1 ’其對輸入圖像資料 imgjn進行放大；Wavelet轉換部2〇2，其自藉由比例放大 部201而實施了比例放大處理之圖像img_up生成低頻側之 次頻帶與高頻側之次頻帶；碼檢測部2〇4，其自藉由 Wavelet轉換部202而生成之低頻侧之次頻帶檢測表示水 平、垂直、傾斜方向之凹凸傾向之凹凸碼；乘法部2〇7 , 其使藉由Wavelet轉換部2〇2而生成之高頻側之次頻帶與強 調係數相乘；絕對值運算部203，其求出藉由上述乘法部 2 0 7而與強調係數相乘而得之尚頻側之次頻帶的絕對值； 及強調處理部205，其使藉由上述絕對值運算部2〇3求出之 絕對值、與由上述碼檢測部204檢測之凹凸碼值相乘；且 輸出使藉由上述Wavelet轉換部202生成之低頻側之次頻 帶、與由上述強調處理部205獲得之乘法值於合成部2〇6中 相加而得之輸出圖像資料img_〇ut。 於上述構成之圖像處理裝置200中，關於比例放大部 201、Wavelet轉換部202、絕對值運算部2〇3、碼檢測部 2〇4、及合成部206，具有與上述實施形態i之圖像處理裝 置100之比例放大部1〇1、Wavelet轉換部1〇2、絕對值運算 部103、碼檢測部1 〇4、及合成部i 06相同之功能。 本實施形態2之圖像處理裝置200中，將上述實施形態i 之圖像處理裝置100之強調處理部105中執行之處理於不同 之部分即乘法部207與強調處理部205中執行。 159526.doc 15 201225000 因此，本實施形態2與上述實施形態丨中，僅高頻側之次 頻帶與強調係數相乘之場所不同，故而輸出之輸出圖像資 料img—out之邊緣強調處理幾乎無不同，因此上述構成之 圖像處理裝置200發揮與上述實施形態1之圖像處理裝置 100相同之效果。 即’根據上述構成之圖像處理裝置200，使藉由Wavelet 轉換部202生成之高頻側之次頻帶、與強調係數相乘而實 施強調處理’其後求出絕對值’並使該絕對值、與自藉由 Wavelet轉換部202而生成之低頻側之次頻帶檢測之凹凸碼 相乘，藉此可消除高頻侧之次頻帶中產生之振動。然後， 使藉由上述Wavelet轉換部202生成之低頻側之次頻帶、與 振動消失之狀態下實施了強調處理之高頻側之次頻帶相 加’藉此獲得實施了無不自然感之自然的邊緣強調處理之 輸出圖像。 (實施形態1、2之效果） 如上所述，上述各實施形態中’以將原圖像放大至2倍 後進行邊緣強調為前提而進行以下之強調處理。其原因在 於，不進行逆Wavelet轉換而使圖像尺寸吻合。 即，將放大至2倍之圖像進行Wavelet轉換，並分解為4 個次頻帶LL、HL、LH、HH。此處所得之次頻帶與原圖像 之尺寸為相同。 此處，對於次頻帶LL設為無邊緣強調之輸出圖像。 使次頻帶HL之絕對值與自次頻帶LL檢測之凹凸碼 sign_hz相乘，藉此進行垂直邊緣強調。 159526.doc 16 201225000 使次頻帶LH之絕對值與自次頻帶LL檢測之凹凸碼 sign_vt相乘，藉此進行水平邊緣強調。 使次頻帶HH之絕對值與自次頻帶LL檢測之凹凸碼 sign一di相乘，藉此進行傾斜邊緣強調。 而且，若於次頻帶LL加上已強調之上述各邊緣，則可實 現消除振動之邊緣強調效果。 參照圖5(a)〜圖5(d)，對上述邊緣強調效果具體地進行說 明。 圖5(a)〜圖5(d)係表示圖像處理裝置1〇〇、圖像處理裝置 200之邊緣強調處理狀態之圖像的像素位置與亮度值之關 係之曲線圖》 上述構成之圖像處理裝置100中，對圖5(a)所示之作為原 圖像之輸入圖像資料imgjn藉由比例放大部ι〇ι而比例放 大至2倍之後，藉由Waveiet轉換部1〇2之換所生 成之低頻側之次頻帶（次帛帶img 一 LL)⑴、肖高頻側之次頻 帶（次頻帶img—HL、次頻帶img—LH、次頻帶img HH)(2)之 像素位置與亮度值的關係成為如圖5 (b)所示之曲線圖般。 該曲線圖中，顯示與高頻側之次頻帶⑺相乘之碼檢測 結果（3)。該碼檢測結果（3)係藉由碼檢測部1〇4檢測藉由Here, the sub-band img-LL is generated by using a low-pass filter in both the horizontal direction and the vertical direction, so that the image is reduced in size to half. The sub-band img_HL represents the longitudinal edge information, the sub-band img LH represents the lateral edge information, and the sub-band img-HH represents the edge information of the oblique direction. The Wavelet conversion unit 102 transfers the sub-bands (sub-band img_HL, sub-band img_LH, and sub-band img-HH) on the high-frequency side of the four sub-bands to the absolute value calculation unit 103. The absolute value calculation unit 103 obtains the absolute value of each of the input frequency bands and transfers it to the subsequent emphasis processing unit 1 〇5. On the other hand, the Wavelet conversion unit 1〇2 transfers the sub-band img_LL, which is the sub-band of the low-frequency side, to the code detecting unit 104 and the combining unit 1〇6. The 7-tap filter is specifically configured to perform code detection using the sub-band img_LL input, for example, in Fig. 4, in the direction of the input and the tilt, respectively. The sub-band img_LL of the image is horizontal and vertical. 159526.doc 12 201225000 The value of the signal is detected using a filter (sign.-sign shown in Figure 1). That is, the code detected in the code detecting object 4, the slgn-vt, and the sub-band in the horizontal direction a #古士 J are the convex codes on the low frequency side. Here, "the concave portion of the D-direction direction tends to be in the graph of the relationship between the pixel position and the brightness (for example, the relationship between the circle c, u, and 丄u degrees (J is shown in Fig. 5(b)). It is observed below that the A-line H-line circle 'in the concave curve portion is output as the code, and the convex portion outputs zero. ; K贞 as the code. The above-mentioned code detection result is outputted in the flat portion for emphasizing the processing unit 丨〇5 In the calculation, the emphasis processing unit 105 multiplies the absolute value of each frequency band of the high frequency side of the material, the code detection result of the code detecting unit (10), and the preset emphasis coefficient by the self- and absolute value calculation unit 1 Specifically, the absolute value of the sub-band, the code sign_hz, and the emphasis coefficient kHL are multiplied, and the absolute value of the sub-band img_LH, the code gn~Vt, and the emphasis coefficient kLH are multiplied to make the absolute value of the sub-band img-HH The code sign-di and the emphasis coefficient kHH are multiplied. The multiplication result is transferred to the synthesizing unit 106. The emphasis coefficients kHL 'kLH, kHH are coefficients for determining the degree of the effect of the edge emphasis processing. The value of the coefficient can easily control the effect of edge emphasis processing Further, the emphasis coefficients kHL, kLH, and kHH may be set in advance or may be appropriately determined depending on the image content. Here, the preset emphasis coefficient is used. The synthesizing unit 106 causes the sub-band img_LL and the self-transferred from the Wavelet conversion unit 102. It is emphasized that the sub-band img_HL, the sub-band img_LH, and the sub-band img_HHi multiplied by the processing unit 1〇5 are added, and are output as the round-out 159526.doc -13-201225000 image data img_out. The processing device 1A obtains an absolute value for the sub-band on the high-frequency side generated by the Wavelet conversion unit 102, and detects the absolute value and the sub-band from the low-frequency side generated by the Wavelet conversion unit 102. By multiplying the concavo-convex code values, the vibration generated in the sub-band on the high-frequency side can be eliminated. The sub-band on the high-frequency side of the state is further multiplied by the emphasis coefficient. Therefore, the emphasis processing is performed in a state where the vibration is eliminated. The sub-band on the low-frequency side generated by the Wavelet conversion unit 102 is added to the sub-band of the high-frequency side subjected to the emphasis processing in the state in which the vibration is removed, thereby obtaining the implementation. (Embodiment 2) FIG. 4 is a block diagram showing a schematic configuration of an image processing apparatus 200 according to Embodiment 2. As shown in FIG. 4, the above image processing is performed. The apparatus 200 includes a scale enlargement unit 201, a Wavelet conversion unit 202, an absolute value calculation unit 203, a code detection unit 204, an emphasis processing unit 205, a synthesis unit 206, and a multiplication unit 207, where the above image processing is performed. The device 200 has substantially the same configuration as the image processing device 100 shown in FIG. 1 but differs in the place where the emphasis coefficient is multiplied. In the image processing device 200, the multiplication unit 207 is obtained before the absolute value of the sub-band (sub-band img_HL, sub-band img_LH, sub-band img-HH) on the high-frequency side generated by the Wavelet conversion unit 202 is obtained. Multiply by the emphasis factor. 159526.doc -14- 201225000 Specifically, the image processing apparatus 200 is an image processing apparatus that performs Wavelet conversion on the input image data img_in and performs edge enhancement processing of the image, which includes: the proportional enlargement unit 2〇1' The input image data imgjn is amplified; the Wavelet conversion unit 2〇2 generates the sub-band of the low-frequency side and the sub-band of the high-frequency side from the image img_up subjected to the proportional enlargement processing by the proportional amplification unit 201; The detecting unit 2〇4 detects the unevenness code indicating the unevenness in the horizontal, vertical, and oblique directions from the sub-band on the low-frequency side generated by the Wavelet conversion unit 202, and the multiplication unit 2〇7 by the Wavelet conversion unit The sub-band on the high-frequency side generated by the second generation is multiplied by the emphasis coefficient, and the absolute value calculation unit 203 obtains the sub-band on the frequency-frequency side obtained by multiplying the emphasis coefficient by the multiplication unit 207. And an emphasis processing unit 205 that multiplies the absolute value obtained by the absolute value calculation unit 2〇3 and the emboss code value detected by the code detection unit 204; and outputs the result by using the Wavelet The conversion unit 202 generates Times the low-frequency side band of multiplication obtained by the enhancement processing unit 205 of the value obtained by adding in the output image data synthesizing section 2〇6 img_〇ut. In the image processing device 200 having the above configuration, the proportional enlargement unit 201, the Wavelet conversion unit 202, the absolute value calculation unit 2〇3, the code detection unit 2〇4, and the synthesis unit 206 have the map of the above-described embodiment i. The functions of the proportional enlargement unit 1〇1, the Wavelet conversion unit 1〇2, the absolute value calculation unit 103, the code detection unit 1〇4, and the synthesis unit i06 of the processing device 100 are the same. In the image processing apparatus 200 of the second embodiment, the processing executed by the emphasis processing unit 105 of the image processing apparatus 100 according to the first embodiment is executed by the multiplication unit 207 and the emphasis processing unit 205 which are different parts. 159526.doc 15 201225000 Therefore, in the second embodiment and the above-described embodiment, only the sub-band on the high-frequency side is different from the embedding coefficient, and thus the edge-emphasis processing of the output image data img_out is almost eliminated. The image processing apparatus 200 having the above configuration is different from the image processing apparatus 100 according to the first embodiment described above. In other words, the image processing device 200 configured as described above performs the emphasis processing by multiplying the sub-band on the high-frequency side generated by the Wavelet conversion unit 202 by the emphasis coefficient, and then obtains the absolute value and then obtains the absolute value. The multiplier code of the sub-band detection on the low-frequency side generated by the Wavelet conversion unit 202 is multiplied, whereby the vibration generated in the sub-band on the high-frequency side can be eliminated. Then, the sub-band on the low-frequency side generated by the Wavelet conversion unit 202 is added to the sub-band of the high-frequency side subjected to the emphasis processing in the state where the vibration is lost, thereby obtaining a natural feeling of being unnatural. The edge emphasizes the processed output image. (Effects of the first and second embodiments) As described above, in the above-described respective embodiments, the following emphasis processing is performed on the premise that the original image is enlarged by two times and the edge is emphasized. The reason for this is that the image size is matched without inverse Wavelet conversion. That is, the image enlarged to twice is subjected to Wavelet conversion, and is decomposed into four sub-bands LL, HL, LH, and HH. The sub-band obtained here is the same as the size of the original image. Here, the sub-band LL is set as an output image without edge emphasis. The absolute value of the sub-band HL is multiplied by the astigmatism code sign_hz detected from the sub-band LL, thereby performing vertical edge emphasis. 159526.doc 16 201225000 The absolute value of the sub-band LH is multiplied by the astigmatism code sign_vt detected from the sub-band LL, thereby performing horizontal edge emphasis. The absolute value of the sub-band HH is multiplied by the astigmatism code sign-di detected from the sub-band LL, thereby performing oblique edge emphasis. Further, if the above-mentioned respective edges are emphasized in the sub-band LL, the edge-improving effect of eliminating vibration can be achieved. The edge emphasis effect described above will be specifically described with reference to Figs. 5(a) to 5(d). 5(a) to 5(d) are graphs showing the relationship between the pixel position and the luminance value of the image in the edge enhancement processing state of the image processing apparatus 1 and the image processing apparatus 200. In the image processing apparatus 100, the input image data imgjn as the original image shown in FIG. 5(a) is scaled up by a factor of 2 by the scale enlargement unit, and then the Waveiet conversion unit 1〇2 The sub-band of the low-frequency side (secondary band img-LL) (1) and the sub-band of the high-frequency side (sub-band img-HL, sub-band img-LH, sub-band img HH) (2) The relationship with the luminance value becomes as shown in the graph of Fig. 5 (b). In the graph, the code detection result (3) multiplied by the sub-band (7) on the high frequency side is displayed. The code detection result (3) is detected by the code detecting unit 1〇4.
Wave丨et轉換部i 〇2之Waveiet轉換所生成之低頻側之次頻帶 ⑴之各方向（水平方肖、垂直方向、傾斜方向）之凹凸性而 獲得的結果。 如上所述，若將高頻側之次頻帶（2)與碼檢測結果（3)相 乘，則如圖5⑷所示，高頻側之次頻帶（2)之振動受到抑 159526.doc 201225000 制。 將以此方式抑制振動之高頻側之次頻帶（2)與上述低頻 側之次頻帶⑴相加後之結果成為如圖5⑷所*之曲線圖 般。該曲線圖係表示輸出圖像資料img一〇ut之像素位置與 亮度值之關係的曲線圖。 因此’自圖5(d)所示之曲線圖可去口，於輸出圖像資料 img_out中，幾乎看不到先前技術之如圖9(d)所示之振動。 由此可知，輸出圖像資料img_ouH^'實施無不自然感之自 然的邊緣強調處理後之輸出圖像。 至此，已對實施形態1之圖像處理裝置i 〇〇之效果進行了 說明’實施形態2之圖像處理裝置200亦發揮相同之效果。 如上所述，根據本實施形態i、2之圖像處理裝置1〇〇、 200 ’對Wavelet轉換後之高頻側之次頻帶以作為邊緣直接 強調之方式進行處理’故而並非如先前般強調高頻率之次 頻帶而進行Wavelet轉換並進行邊緣強調處理者。 因此’不會因逆Wavelet轉換所引起之不自然的振動而 導致不自然的邊緣強調結果。 上述各實施形態中’關於輸入圖像資料是否為彩色圖像 並未特別考慮，以下之實施形態3、4中，對輸入圖像資料 為彩色圖像之情形之邊緣強調處理進行說明。 (實施形態3) 本實施形態中，說明對YCbCr格式之彩色圖像資料進行 邊緣強調處理之圖像處理裝置。 圖6表示本實施形態之圖像處理裝置300之概略構成方塊 159526.doc •18· 201225000 圖。 此處，YCbCr格式之彩色圖像中’若僅對γ成分之圖像 實施邊緣強調處理，則成為實施全體圖像之邊緣強調處 理，故而圖像處理裝置300成為如下所述之構成。 即，上述圖像處理裝置300成為包含圖像處理部3〇1、比 例放大部302、及比例放大部303之構成。 上述圖像處理部301對Υ成分之輸入圖像資料img γ 貫施邊緣強调處理而獲得輸出圖像資料img γ 。該圖 像處理部301係於上述實施形態1之圖像處理裝置1〇〇或上 述實施形態2之圖像處理裝置200之任一者中實現。 上述比例放大部3 02對Cb成分之輸入圖像資料img_Cb_in 實施比例放大處理而獲得輸出圖像資料img_Cb_〇ut。 同樣地，上述比例放大部303對Cr成分之輸入圖像資料 img一Cr一in實施比例放大處理而獲得輸出圖像資料 img_Cr_out。 此處’當輸入圖像與輸出圖像為相同尺寸之情形時，無 需實施比例放大部302、303對於Cb成分之輸入圖像資料 img—Cb_in、及Cr成分之輸入圖像資料img—Cr」n的比例放 大處理。 根據上述構成之圖像處理裝置3〇〇，可獲得邊緣修正結 果中不產生振動之實施了自然的邊緣強調處理之¥(:13(^格 式之輸出彩色圖像。 (實施形態4) 本實施形態中，說明對RGB格式之彩色圖像資料進行邊 159526.doc •19- 201225000 緣強調處理之圖像處理裝置。 圖7表示本實施形態之圖像處理裝置400之概略構成方塊 圖。 上述圖像處理裝置400成為包含與R成分、（}成分、B成 分之各自對應之3個圖像處理部401之構成。該等3個圖像 處理部401設為具有同一功能者。 上述圖像處理部40 1例如對R成分之輸入圖像資料 img_R_in實施邊緣強調處理而獲得輸出圖像資料 img_R_〇ut。同樣地，對G成分之輸入圖像資料img_G_in實 施邊緣強調處理而獲得輸出圖像資料img_G_out，對b成分 之輸入圖像資料img_B_in實施邊緣強調處理而獲得輸出圖 像資料 img_B_out。 此處，上述圖像處理部401係於上述實施形態1之圖像處 理裝置100或上述實施形態2之圖像處理裝置2〇〇之任一者 中實現。 根據上述構成之圖像處理裝置4〇〇，可獲得邊緣修正結 果中不產生振動之實施了自然的邊緣強調處理之RGB格式 之輸出彩色圖像。 以上任一實施形態中，均對圖像處理裝置進行了說明， 但本發明不僅為裝置，以下所示之圖像處理方法亦包含於 申請專利範圍中。 即本發明之圖像處理方法係對輸入圖像資料進行 Wavelet轉換並進行圖像之邊緣強調處理者，其包含以下 步驟：wavelet轉換步驟’藉_avelet轉換部1〇2而自輸入 159526.d〇c •20- 201225000 圖像資料生成低頻側之次頻帶與高頻側之次頻帶；碼檢測 步驟藉由碼檢測部104而自藉由Wavelet轉換步驟生成之 低頻側之次頻帶檢測表示水平、垂直、傾斜方向之凹凸傾 向的凹凸碼；絕對值運算步驟，藉由絕對值運算部1〇3而 求出上述高頻側之次頻帶之絕對值；強調處理步驟，藉由 強調處理部1〇5而使藉由上述絕對值運算步驟求出之絕對 值、與上述碼檢測步驟中檢測之凹凸碼相乘並進行圖像之 邊緣之強調處理，及輸出步驟，輸出藉由上述合成部1〇6 而使藉由上述Wavelet轉換步驟生成之低頻側之次頻帶、 與上述強調處理步驟中獲得之乘法值相加而得的圖像。 P便為以上構成之圖像處理方法亦發揮如下效果，即與 上述圖像處理裝置同樣地，可獲得實施了無不自然之振動 之自然的邊緣強調處理之輸出圖像。 又，作為具備上述構成之圖像處理裝置之電子機器，只 要係需要對輸入时實施邊緣處理之電子機$，則可使用 任意之電子機器，例如可舉出電視接收機、電子書終端、 行動終端裝置等。 又本發明之圖像處理裝置100_執行之處理亦可藉由 以下所示之控制程式而作為軟體實現。 上述控制程式係使電腦執行以下步驟者：Wavelet轉換 步驟，自輸人圖像資料生成低頻側之次頻帶與高頻側之次 頻帶’碼檢測步驟，自藉由上述仏…以轉換步驟而生成 之低頻側之次頻帶檢測表示水平、垂直、傾斜方向之凹凸 傾向的凹凸碼’絕對值運算步驟，求出上述高頻側之次頻 I59526.doc -21- 201225000 帶之絕對值，強調處理步驟，使藉由上述絕對值運算步驟 求出之絕對值、肖述碼檢測步驟中檢測之凹凸竭相乘而 進行圓像之邊緣之強調處理；及輸出步驟，輸出使藉由上 述Wavelet轉換步驟生成之低頻側之次頻帶、與上述強調 處理步驟中獲得之乘法值相加而得的圖像。 又，上述控制程式亦可記錄於電腦可讀取之記錄媒體中 並供給至電腦而執行。 根據以上所述，本發明之圖像處理裝置1〇〇可藉由硬體 邏輯而構成，亦可如下所述使用cpu並藉由軟體而實現。 該情形時，圖像處理裝置1〇〇包含：執行實現各處理（比 例放大處理、Wavelet轉換處理等之各步驟）之控制程式之 命令的CPU(central processing unit，中央處理單元）；儲存 有上述控制程式之R〇M(read only memory，唯讀記憶 體），展開上述控制程式之RAM(rand〇m access ， 隨機存取記憶體）；及儲存有上述控制程式及各種資料之 記憶體等之記憶裝置（記錄媒體）等。而且，本發明之目的 亦可藉由以下方法達成：將電腦可讀取地記錄有實現上述 功能之軟體即上述控制程式之程式碼（執行形式程式、中 間碼程式、原始程式）的記錄媒體供給至電腦，由該電腦 (或 CPU(central processing unit，中央處理單元）、Mpu (microprocessor unit，微處理單元））讀出記錄於記錄媒體 中之程式碼並執行。 作為上述記錄媒體，可使用例如磁帶或卡式磁帶等之帶 系列；包含軟性（fl〇ppy，註冊商標）磁碟/硬性磁碟等之磁 159526.doc -22· 201225000 碟或 CD-ROM(compact disk read only memory，唯讀光碟 記憶體）/MO(magneto optical，磁光碟）/MD(magnetic disk，磁盤）/DVD(digital versatile disc，數位多功光碟）/ CD-R(compact disk-recordable，可記錄光碟）等之光碟的碟 系列；IC(integrated circuit積體電路）卡（包含記憶體卡）/光 卡等之卡系列；或mask ROM(光罩唯讀記憶體）/EPROM (erasable programmable read only memory，可抹除可程式 化唯讀記憶體）/EEPROM(electrically erasable programmable read only memory，電子可擦可程序化唯讀記憶體）/flash ROM(快閃唯讀記憶體）等之半導體記憶體系列等。 又，亦可將圖像處理裝置100與通訊網路可連接地構 成，並經由通訊網路而供給上述程式碼。作為該通訊網 路，並無特別限定，可利用例如網際網路、内部網路、外 部網路、LAN(local area network，局域網路）、ISDN (integrated service digital network，整合服務數位網路）、 VAN(value-added network，增值網路）、CATV(community antenna television，有線電視）通訊網、虛擬專用網（virtual private network)、電話線路網、移動通訊網、衛星通訊網 等。又，作為構成通訊網路之傳送媒體，並無特別限定， 可利用例如 IEEE(Institute of Electrical and Electronics Engineers，電機電子工程師學會）1394、USB(universal serial bus，通用串列匯流排）、電力線傳輸、電規電視線 路、電話線、ADSL(asymmetric digital subscriber line，標 準非對稱數位用戶迴路）線路等之有線，亦可利用如 159526.doc -23- 201225000The result obtained by the unevenness of each direction (horizontal square, vertical direction, and oblique direction) of the sub-band (1) on the low-frequency side generated by the Waveiet conversion unit i 〇2. As described above, when the sub-band (2) on the high-frequency side is multiplied by the code detection result (3), as shown in Fig. 5 (4), the vibration of the sub-band (2) on the high-frequency side is suppressed by 159526.doc 201225000. . The result of adding the sub-band (2) on the high-frequency side of the vibration to the sub-band (1) on the low-frequency side in this manner is as shown in the graph of Fig. 5 (4). The graph is a graph showing the relationship between the pixel position of the output image data img and the luminance value. Therefore, the graph shown in Fig. 5(d) can be removed, and in the output image data img_out, the vibration shown in Fig. 9(d) of the prior art is hardly seen. From this, it can be seen that the output image data img_ouH^' implements an output image after the natural edge enhancement processing without an unnatural feeling. Thus, the effect of the image processing apparatus i of the first embodiment has been described. The image processing apparatus 200 of the second embodiment also exhibits the same effects. As described above, the image processing apparatuses 1A and 200' according to the first and second embodiments of the present embodiment deal with the sub-band of the high-frequency side after the Wavelet conversion as the direct emphasis of the edge. Therefore, the emphasis is not as high as before. The Wavelet conversion is performed on the sub-band of the frequency and the edge emphasis processor is performed. Therefore, the unnatural edge is not emphasized by the unnatural vibration caused by the inverse Wavelet transformation. In the above embodiments, whether or not the input image data is a color image is not particularly considered. In the following embodiments 3 and 4, the edge emphasis processing in the case where the input image data is a color image will be described. (Embodiment 3) In the present embodiment, an image processing apparatus that performs edge enhancement processing on color image data of the YCbCr format will be described. Fig. 6 is a view showing a schematic configuration of the image processing apparatus 300 of the present embodiment 159526.doc • 18· 201225000. Here, in the color image of the YCbCr format, if the edge enhancement processing is performed on only the image of the γ component, the edge enhancement processing for performing the entire image is performed. Therefore, the image processing apparatus 300 has the following configuration. In other words, the image processing device 300 includes the image processing unit 3〇1, the proportional enlargement unit 302, and the scale enlargement unit 303. The image processing unit 301 performs edge enhancement processing on the input image data img γ of the Υ component to obtain the output image data img γ . The image processing unit 301 is realized by any of the image processing device 1 of the first embodiment or the image processing device 200 of the second embodiment. The proportional amplification unit 312 performs a scaling process on the input image data img_Cb_in of the Cb component to obtain an output image data img_Cb_〇ut. Similarly, the proportional amplification unit 303 performs a scaling process on the input image data img_Cr_in of the Cr component to obtain an output image data img_Cr_out. Here, when the input image and the output image are the same size, it is not necessary to implement the input image data img_Cb_in for the Cb component and the input image data img-Cr for the C component. Proportional amplification of n. According to the image processing apparatus 3 configured as described above, it is possible to obtain a natural color edge enhancement processing that does not generate vibration in the edge correction result (: 13 (output format color image of the ^ format) (Embodiment 4) This embodiment In the embodiment, an image processing apparatus that performs edge 159526.doc •19-201225000 edge emphasis processing on color image data of the RGB format will be described. Fig. 7 is a block diagram showing a schematic configuration of the image processing apparatus 400 of the present embodiment. The image processing device 400 includes three image processing units 401 corresponding to each of the R component, the (} component, and the B component. The three image processing units 401 have the same function. The portion 40 1 performs edge enhancement processing on the input image data img_R_in of the R component to obtain the output image data img_R_〇ut. Similarly, the edge image processing is performed on the input image data img_G_in of the G component to obtain the output image data. img_G_out performs edge enhancement processing on the input image data img_B_in of the b component to obtain an output image data img_B_out. Here, the image processing unit 401 is based on the above The image processing device 100 of the first embodiment or the image processing device 2 of the second embodiment is realized. According to the image processing device 4 configured as described above, vibration is not generated in the edge correction result. The output color image of the RGB format of the natural edge enhancement processing is implemented. In any of the above embodiments, the image processing apparatus has been described. However, the present invention is not only an apparatus, but also an image processing method described below. It is included in the scope of the patent application. That is, the image processing method of the present invention performs Wavelet conversion on the input image data and performs edge enhancement processing on the image, and includes the following steps: wavelet conversion step 'borrowing_avelet conversion unit 1〇 2, since the input 159526.d〇c • 20-201225000 image data generates the sub-band of the low frequency side and the sub-band of the high frequency side; the code detection step is generated by the code detecting unit 104 from the low frequency side generated by the Wavelet conversion step The sub-band detection is a concavo-convex code indicating the tendency of the unevenness in the horizontal, vertical, and oblique directions; in the absolute value calculation step, the above-described high is obtained by the absolute value calculation unit 1〇3 An absolute value of the sub-band of the frequency side; an emphasis processing step of multiplying the absolute value obtained by the absolute value calculation step by the emphasis processing unit 1〇5 and multiplying the embossed code detected in the code detecting step An emphasis processing of the edge of the image, and an output step of outputting the sub-band of the low-frequency side generated by the Wavelet conversion step by the combining unit 1〇6, and adding the multiplication value obtained in the emphasis processing step In the image processing method of the above configuration, P has an effect that an output image subjected to natural edge enhancement processing without unnatural vibration can be obtained similarly to the above-described image processing apparatus. Further, as an electronic device including the image processing device having the above configuration, any electronic device can be used as long as it is necessary to perform edge processing on the input, and examples thereof include a television receiver, an e-book terminal, and an action. Terminal device, etc. Further, the processing executed by the image processing apparatus 100_ of the present invention can be realized as software by a control program shown below. The above control program causes the computer to perform the following steps: a Wavelet conversion step, generating a sub-band and a high-frequency side sub-band 'code detection step from the input image data from the input image data, by generating the above-mentioned steps... The sub-band detection on the low-frequency side indicates the embossing code 'absolute value calculation step indicating the unevenness in the horizontal, vertical, and oblique directions, and the absolute value of the sub-frequency I59526.doc -21 - 201225000 band on the high-frequency side is obtained, and the processing steps are emphasized. And performing an emphasis process on the edge of the circular image by multiplying the absolute value obtained by the absolute value calculation step and the unevenness detected in the schematic code detecting step; and outputting the output to be generated by the Wavelet conversion step The sub-band on the low frequency side and the image obtained by adding the multiplication value obtained in the above-described emphasis processing step. Further, the above control program can be recorded on a computer-readable recording medium and supplied to a computer for execution. As described above, the image processing apparatus 1 of the present invention can be constructed by hardware logic, and can also be realized by software using cpu as follows. In this case, the image processing apparatus 1A includes a CPU (central processing unit) that executes a command to implement a control program for realizing each of the processes (proportional enlargement processing, Wavelet conversion processing, and the like); Control program R〇M (read only memory), RAM (rand〇m access, random access memory) for developing the above control program; and memory for storing the above control program and various data. Memory device (recording medium), etc. Furthermore, the object of the present invention can also be achieved by a computer readable record of a recording medium in which the software for realizing the above function, that is, the program code (execution format program, intermediate code program, and original program) of the software for realizing the above functions is recorded. To the computer, the computer (or CPU (central processing unit), MPU (microprocessor unit)) reads the code recorded in the recording medium and executes it. As the recording medium, a tape series such as a magnetic tape or a cassette tape can be used, and a magnetic 159526.doc -22·201225000 disc or CD-ROM including a soft (fl〇ppy, registered trademark) disk/hard disk can be used. Compact disk read only memory, /MO (magneto optical), MD (magnetic disk) / DVD (digital versatile disc) / CD-R (compact disk-recordable , discs that can record discs, etc.; IC (integrated circuit) cards (including memory cards) / optical cards, etc.; or mask ROM (mask read-only memory) / EPROM (erasable) Programmable read only memory, eraseable programmable read only memory / EEPROM (electrically erasable programmable read only memory) / flash ROM (flash read only memory) Semiconductor memory series, etc. Alternatively, the image processing apparatus 100 may be connected to a communication network and supplied with the above code via a communication network. The communication network is not particularly limited, and may be, for example, an internet, an internal network, an external network, a LAN (local area network), an ISDN (integrated service digital network), and a VAN. (value-added network), CATV (community antenna television) communication network, virtual private network (virtual private network), telephone line network, mobile communication network, satellite communication network, etc. Further, the transmission medium constituting the communication network is not particularly limited, and for example, IEEE (Institute of Electrical and Electronics Engineers) 1394, USB (universal serial bus), power line transmission, and the like can be used. Cables for electric TV lines, telephone lines, ADSL (asymmetric digital subscriber line) lines, etc. can also be used, such as 159526.doc -23- 201225000
IrDA(infrared data associati〇n，红夕卜線數據協定）或遙控之 類的紅外線、Bluetooth(藍牙，註冊商標）、8〇2 i丨無線、 HDR(high dynamic range，高動態範圍）、行動電話網衛 星線路、數位i也面波網等之無冑。再I，本發明亦可以上 述程式碼以電子傳送而具體化之載人搬送波中之電腦資料 信號的形態下亦可實現。 本發明並不限定於上述各實施形態，於申請專利範圍内 可進行各種變更，對於將不同實施形態中分別揭示之技術 手段加以適當組合所獲得之實施形態亦包含於本發明之技 術範圍内。 [產業上之可利用性] 本發明可利用於需要對輸出圖像進行邊緣強調處理之圖 像顯示裝置，尤其是使用有液晶面板之液晶顯示裝置。 【圖式簡單說明】 圖1係本發明之實施形態1之圖像處理裝置之概略構成方 塊圖。 圖2係說明Wavelet轉換之概要之圖。 圖3係表示圖1所示之圖像處理裝置之碼檢測部中使用之 濾波器例之圖。 圖4係本發明之實施形態2之圖像處理裝置之概略構成方 塊圖。 圖5(a)〜(d)係表示圖1或圖4所示之圖像處理裝置之邊緣 強調處理狀態之圖像的像素位置與亮度值之關係之曲線 圖。 159526.doc 24- 201225000 圖6係本發明之實施形態3之圖像處理裝置之概略構成方 塊圖。 圖7係本發明之實施形態4之圖像處理裝置之概略構成方 塊圖。 圖8係先前之圖像處理裝置之概略構成方塊圖。 圖9(a)〜(d)係表示圖8所示之圖像處理裝置之邊緣強調處 理狀態之圖像的像素位置與亮度值之關係之曲線圖。 【主要元件符號說明】 100 圖像處理裝置 101 比例放大部 102 Wavelet轉換部 103 絕對值運算部 104 碼檢測部 105 強調處理部 106 合成部 200 圖像處理裝置 201 比例放大部 202 Wavelet轉換部 203 絕對值運算部 204 碼檢測部 205 強調處理部 206 合成部 207 乘法部 300 圖像處理裝置 159526.doc 25- 201225000 301 圖像處理部 302 比例放大部 303 比例放大部 400 圖像處理裝置 401 圖像處理部 HH 次頻帶 HL 次頻帶 LH 次頻帶 LL 次頻帶 kHH 強調係數 kHL 強調係數 kLH 強調係數 159526.doc -26-IrDA (infrared data associati〇n, red ray line data protocol) or infrared such as remote control, Bluetooth (registered trademark), 8 〇 2 i 丨 wireless, HDR (high dynamic range), mobile phone The network satellite line and the digital i are also innocent. Further, the present invention can also be realized in the form of a computer data signal in a manned transmission wave in which the above code is embodied by electronic transmission. The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention. The embodiments obtained by appropriately combining the technical means disclosed in the different embodiments are also included in the technical scope of the present invention. [Industrial Applicability] The present invention is applicable to an image display device which requires edge enhancement processing on an output image, and in particular, a liquid crystal display device using a liquid crystal panel. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing a schematic configuration of an image processing apparatus according to a first embodiment of the present invention. Figure 2 is a diagram illustrating the outline of a Wavelet conversion. Fig. 3 is a view showing an example of a filter used in the code detecting unit of the image processing apparatus shown in Fig. 1; Fig. 4 is a block diagram showing a schematic configuration of an image processing apparatus according to a second embodiment of the present invention. Figs. 5(a) to 5(d) are graphs showing the relationship between the pixel position and the luminance value of the image in the edge enhancement processing state of the image processing apparatus shown in Fig. 1 or Fig. 4. 159526.doc 24-201225000 Fig. 6 is a block diagram showing a schematic configuration of an image processing apparatus according to a third embodiment of the present invention. Fig. 7 is a block diagram showing a schematic configuration of an image processing apparatus according to a fourth embodiment of the present invention. Fig. 8 is a block diagram showing a schematic configuration of a conventional image processing apparatus. Figs. 9(a) to 9(d) are graphs showing the relationship between the pixel position and the luminance value of the image in the edge emphasis processing state of the image processing apparatus shown in Fig. 8. [Description of Main Element Symbols] 100 Image Processing Apparatus 101 Proportional Enlarger Unit 102 Wavelet Conversion Unit 103 Absolute Value Calculation Unit 104 Code Detection Unit 105 Emphasizing Processing Unit 106 Synthesis Unit 200 Image Processing Apparatus 201 Proportional Enlargement Unit 202 Wavelet Conversion Unit 203 Absolute Value calculation unit 204 code detection unit 205 emphasis processing unit 206 synthesis unit 207 multiplication unit 300 image processing device 159526.doc 25-201225000 301 image processing unit 302 scale enlargement unit 303 scale enlargement unit 400 image processing device 401 image processing Part HH sub-band HL sub-band LH sub-band LL sub-band kHH emphasis coefficient kHL emphasis coefficient kLH emphasis coefficient 159526.doc -26-