JP4714044B2 - Color image signal encoding apparatus - Google Patents
Color image signal encoding apparatus Download PDFInfo
- Publication number
- JP4714044B2 JP4714044B2 JP2006062051A JP2006062051A JP4714044B2 JP 4714044 B2 JP4714044 B2 JP 4714044B2 JP 2006062051 A JP2006062051 A JP 2006062051A JP 2006062051 A JP2006062051 A JP 2006062051A JP 4714044 B2 JP4714044 B2 JP 4714044B2
- Authority
- JP
- Japan
- Prior art keywords
- color image
- luminance signal
- control information
- quantization
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Compression Of Band Width Or Redundancy In Fax (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Color Television Systems (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Description
本発明は適応量子化を用いたカラー画像符号化装置、特に色再現性に優れたカラー画像符号化装置に関するものである。 The present invention relates to a color image encoding apparatus using adaptive quantization, and more particularly to a color image encoding apparatus excellent in color reproducibility.
静止画や動画像等の画像データを小領域に分割し、人間の視覚特性を考慮して小領域毎に量子化値を適応的に与えることにより、画像を高品質に圧縮符号化する画像符号化方式が広く普及している。このような符号化方式として、静止画についてはJPEGが規定され、動画像についてはMPEG−1,2,4が国際標準として規定されている。 MPEG−2では、入力された画像は16×16画素のマクロブロックに分割され、輝度信号Yと2つの色差信号Pb,Prとの組に分割される。さらに、マクロブロックは、8×8画素の小ブロックに分割される。各小ブロックの画素データは直交変換され、直交変換係数が量子化される。直交変換係数の量子化値はマクロブロック毎に1つ与えられ、マクロブロックに含まれる複数の小ブロックの全てが同一の量子化値で符号化される。 Image code that compresses and encodes images with high quality by dividing image data such as still images and moving images into small regions and adaptively assigning quantization values to each small region in consideration of human visual characteristics The conversion method is widely used. As such an encoding method, JPEG is defined for still images and MPEG-1, 2, 4 is defined as an international standard for moving images. In MPEG-2, an input image is divided into 16 × 16 pixel macroblocks, and is divided into a set of a luminance signal Y and two color difference signals Pb and Pr. Furthermore, the macroblock is divided into small blocks of 8 × 8 pixels. The pixel data of each small block is orthogonally transformed and the orthogonal transformation coefficient is quantized. One quantized value of the orthogonal transform coefficient is given for each macroblock, and all of a plurality of small blocks included in the macroblock are encoded with the same quantized value.
TM5(Test Model 5, ISO/IEC JTC1/SC29/WG11/NO400, MPEG93/475)に代表されるように、従来のMPEG−2ビデオ準拠の動画像符号化方法では、圧縮符号化データの符号量は量子化値により制御されている。量子化値を大きくすれば発生符号量は減少し量子化値を小さくすれば発生符号化量は増大するので、目標符号量と過去の符号量とを比較し、平均符号量が目標符号量に近くなるように適応的に発生符号量が制御されている。量子化の際、量子化値を大きくすると画質は低下するが、人間の視覚特性を考慮し、劣化が検知され易い領域と検知されにくい領域とで量子化値を相違させることにより視覚的な画質の向上が図られている。例えば、人間の視覚特性として、周波数の高い領域は周波数の低い領域よりも劣化が検知されにくい特性がある。この視覚特性を考慮し、各マクロブロックの輝度成分の分散値(アクティビティ)を算出し、分散値の小さい領域は劣化が検知され易いため量子化値が小さくなるように制御し、分散値の大きい領域については劣化が検知されにくいため、量子化値が大きくなるように制御されている。 As represented by TM5 (Test Model 5, ISO / IEC JTC1 / SC29 / WG11 / NO400, MPEG93 / 475), the conventional MPEG-2 video-compliant moving image encoding method uses a code amount of compressed encoded data. Is controlled by the quantization value. If the quantization value is increased, the generated code amount decreases, and if the quantized value is decreased, the generated code amount increases, so the target code amount is compared with the past code amount, and the average code amount becomes the target code amount. The generated code amount is adaptively controlled so as to be close. When quantizing, if the quantized value is increased, the image quality deteriorates, but considering the human visual characteristics, the visual image quality is determined by making the quantized value different between the region where deterioration is easy to detect and the region where deterioration is difficult to detect. Improvements are being made. For example, as a human visual characteristic, there is a characteristic in which deterioration is less likely to be detected in a high frequency region than in a low frequency region. In consideration of this visual characteristic, the variance value (activity) of the luminance component of each macroblock is calculated. Since the degradation is easily detected in the region where the variance value is small, the quantization value is controlled to be small, and the variance value is large. Since it is difficult to detect deterioration in the region, the quantization value is controlled to be large.
上述した輝度信号の分散値に基づいて量子化値を制御する方法は、人間の視覚特性に適合し、画質劣化を視覚的に目立ちにくくする利点が達成される。しかしながら、色再現性に十分な考慮が払われていないため、カラー画像の再現性に難点があり、特に彩度の高い領域の画質が低下する問題が生じていた。彩度の高い領域の画質を改善する方法として、R-Y色差信号から赤色信号成分を含む領域を検出し、赤色成分の高い領域のブロックの輝度信号の圧縮率を低くする方法が提案されているが(例えば、特開平8-34055号公報参照)、複雑な閾値処理を行って赤色成分の高い領域を特定しなければならず、符号化処理が複雑になる欠点があった。 The above-described method for controlling the quantization value based on the variance value of the luminance signal achieves the advantage of adapting to human visual characteristics and making the image quality degradation visually inconspicuous. However, since sufficient consideration has not been given to the color reproducibility, there is a problem in the reproducibility of the color image, and there has been a problem that the image quality in a particularly high-saturation region is deteriorated. As a method for improving the image quality in a high-saturation region, a method has been proposed in which a region including a red signal component is detected from the RY color difference signal and the compression ratio of the luminance signal of the block in the region having a high red component is lowered. (For example, refer to Japanese Patent Laid-Open No. 8-34055), there is a disadvantage that a complicated threshold process must be performed to identify a region having a high red component, which complicates the encoding process.
本発明の目的は、カラー再現性に優れたカラー画像符号化装置を実現することにある。
本発明の別の目的は、彩度の高い領域の画質低下を抑制したカラー画像符号化装置を提供することにある。
An object of the present invention is to realize a color image encoding apparatus excellent in color reproducibility.
Another object of the present invention is to provide a color image encoding apparatus that suppresses deterioration in image quality in a highly saturated area.
本発明によるカラー画像符号化装置は、輝度信号(Y)と2つの色差信号(Pb,Pr)とにより構成されるカラー画像信号を符号化するカラー画像符号化装置であって、
符号化すべきカラー画像信号を所定の画素数のブロックに分割する分割手段と、
分割されたカラー画像信号について、各ブロック毎に直交変換して変換係数を出力する直交変換手段と、
量子化制御情報に基づき、前記直交変換係数を量子化して量子化値を出力する量子化手段と、
量子化値を符号化して符号化データを出力する符号化手段と、
符号化データから発生符号量を計数する発生符号量計数手段と、
各ブロックの輝度信号及び2つの色差信号を受け取り、視覚特性に基づく制御情報を出力する視覚特性演算手段と、
前記発生符号量計数手段及び視覚特性演算手段からの出力情報に基づいて量子化制御情報を出力する量子化制御手段とを具えるカラー画像符号化装置において、
前記視覚特性演算手段は、各ブロックの輝度信号の分散値を算出する手段と、 輝度信号の直流成分を出力する手段と、輝度信号及び2つの色差信号から三原色RGBの直流成分を測定する手段と、前記輝度信号の直流成分と三原色RGBの直流成分との比率を演算する手段とを有し、
前記量子化制御手段は、輝度信号の分散値と発生符号量とに基づいて量子化制御情報(Q0)を発生する量子化制御情報発生手段と、前記輝度信号の直流成分と三原色の直流成分との比率を用いて量子化制御情報を補正する手段とを有し、補正された量子化制御情報(Q)を量子化手段に供給することを特徴とする。
A color image encoding device according to the present invention is a color image encoding device that encodes a color image signal composed of a luminance signal (Y) and two color difference signals (P b , P r ),
Dividing means for dividing the color image signal to be encoded into blocks of a predetermined number of pixels;
For the divided color image signal, orthogonal transform means for performing orthogonal transform for each block and outputting transform coefficients;
Based on quantization control information, quantization means for quantizing the orthogonal transform coefficient and outputting a quantization value;
Encoding means for encoding the quantized value and outputting encoded data;
Generated code amount counting means for counting the generated code amount from the encoded data;
Visual characteristic calculation means for receiving a luminance signal of each block and two color difference signals and outputting control information based on the visual characteristics;
In a color image encoding device comprising quantization control means for outputting quantization control information based on output information from the generated code amount counting means and visual characteristic calculation means,
The visual characteristic calculation means includes means for calculating a variance value of the luminance signal of each block, means for outputting a DC component of the luminance signal, means for measuring the DC component of the three primary colors RGB from the luminance signal and the two color difference signals, and And means for calculating the ratio of the direct current component of the luminance signal and the direct current component of the three primary colors RGB,
The quantization control means includes: quantization control information generation means for generating quantization control information (Q 0 ) based on a variance value of a luminance signal and a generated code amount; a direct current component of the luminance signal and a direct current component of three primary colors And a means for correcting the quantization control information using the ratio of the above and the corrected quantization control information (Q) is supplied to the quantization means.
従来の符号化方法では、色成分による符号化劣化が考慮されていないため、彩度の高い画像領域の劣化が顕著になる欠点があった。また、同一の彩度の領域であっても、黄(R+G)、緑(G)、マゼンダ(R+B)、赤(R)、青(B)の順序で劣化が目立っていた。そこで、本発明では、輝度信号Yの寄与度だけでなく、三原色RGBの寄与度も考慮して量子化値を補正ないし制御する。すなわち、輝度信号の直流成分と三原色RGBの直流成分との比率を算出し、得られた比率のうち最も強く寄与している色の比率(N)を求め、求めた比率を用いて又は比率を変数として含む関数を用いて量子化制御情報を補正ないし補償する。この結果、彩度の高い画像領域について、その寄与度に応じて量子化値が制御されることになり、彩度の高い画像領域の符号化劣化を抑制することができる。 In the conventional encoding method, since the encoding deterioration due to the color component is not taken into consideration, there is a drawback that the deterioration of the image area with high saturation becomes remarkable. Even in the same saturation region, the deterioration was conspicuous in the order of yellow (R + G), green (G), magenta (R + B), red (R), and blue (B). Therefore, in the present invention, the quantization value is corrected or controlled in consideration of not only the contribution of the luminance signal Y but also the contributions of the three primary colors RGB. That is, the ratio of the direct current component of the luminance signal and the direct current component of the three primary colors RGB is calculated, the ratio (N) of the color that contributes most strongly among the obtained ratios is obtained, and the ratio or the ratio is calculated using the obtained ratio. Quantization control information is corrected or compensated using a function included as a variable. As a result, the quantization value is controlled according to the contribution degree of the image area with high saturation, and the encoding deterioration of the image area with high saturation can be suppressed.
本発明によるカラー画像信号符号化装置の好適実施例は、関数f(N)として、kを0<k≦1の値をとる係数とした場合に、f(N)=k(N−1)+1で表される関数を用いることを特徴とする。当該関数値を補正用のパラメータとして用いる場合、係数kの値を適切に選択することにより、色再現性に対する寄与率を適切に設定することができる。 In a preferred embodiment of the color image signal encoding apparatus according to the present invention, f (N) = k (N−1) where k is a coefficient having a value of 0 <k ≦ 1 as the function f (N). A function represented by +1 is used. When the function value is used as a correction parameter, the contribution rate to the color reproducibility can be appropriately set by appropriately selecting the value of the coefficient k.
本発明によるカラー画像信号符号化装置に別の好適実施例は、三原色RGBの直流成分を測定する手段は、以下の変換式、
R=Y+1.5748Pr
G=Y−0.1873Pb −0.4681Pr
B=Y+1.8556Pb
を用いて三原色の直流成分を算出することを特徴とする。
In another preferred embodiment of the color image signal encoding apparatus according to the present invention, means for measuring the direct current component of the three primary colors RGB is the following conversion formula:
R = Y + 1.5748P r
G = Y−0.1873P b −0.4681P r
B = Y + 1.8556P b
Is used to calculate the direct current component of the three primary colors.
本発明では、輝度信号に対する三原色RGBの寄与率を考慮して、量子化制御情報を補正しているので、彩度の高い画像領域の符号化劣化を適切に抑制することができ、カラー再現性に優れたカラー画像符号化装置を実現することできる。 In the present invention, since the quantization control information is corrected in consideration of the contribution ratio of the three primary colors RGB to the luminance signal, it is possible to appropriately suppress the coding deterioration of the image area with high saturation, and color reproducibility. It is possible to realize a color image encoding apparatus excellent in the above.
図1は本発明によるカラー画像符号化装置の一例を示す線図である。本例では、MPEG-2映像符号化(ISO/IEC13818-2)に適合した例について説明する。符号化すべきカラー画像信号は、輝度信号Yと2つの色差信号Pb,Prとで構成される。カラー画像信号は、分割手段1に入力し、16×16画素のマクロブロックに分割した後、さらにY信号及び2つの色差信号の各信号について8×8画素を単位とする信号に分割され、ブロック化された画像信号とする。 FIG. 1 is a diagram showing an example of a color image encoding apparatus according to the present invention. In this example, an example conforming to MPEG-2 video coding (ISO / IEC13818-2) will be described. Color image signal to be encoded, the luminance signal Y and two color difference signals P b, composed of the P r. The color image signal is input to the dividing means 1 and divided into 16 × 16 pixel macroblocks, and then divided into signals each having 8 × 8 pixels for each of the Y signal and the two color difference signals. It is assumed that the converted image signal.
ブロック化された画像信号は、直交変換器2により各ブロック毎に直交変換され、その変換係数を量子化器3に供給する。尚、変換方式として、DCT(離散コサイン変換)を用いる。 The block image signal is orthogonally transformed for each block by the orthogonal transformer 2, and the transform coefficient is supplied to the quantizer 3. Note that DCT (discrete cosine transform) is used as a conversion method.
量子化器3は、量子化制御部4から供給される量子化制御情報(又は、量子化幅)に基づいて変換係数を量子化して量子化値を出力して可変長符号化器5に供給する。量子化値は各マクロブロックについて1つ与えられ、マクロブロックに含まれる複数の全てのブロックが同一の量子化値で符号化され、符号化器5から符号化データを出力する。 The quantizer 3 quantizes the transform coefficient based on the quantization control information (or quantization width) supplied from the quantization control unit 4, outputs a quantized value, and supplies the quantized value to the variable length encoder 5. To do. One quantization value is given for each macroblock, and all of a plurality of blocks included in the macroblock are encoded with the same quantization value, and encoded data is output from the encoder 5.
ブロック化された画像信号は視覚特性演算部6にも供給され、ブロック化された画像信号について人間の視覚特性に応じた演算処理が行われ、視覚特性に基づく制御情報を量子化制御部6に供給する。符号化器5から出力される符号化データは発生符号化量係数部7に供給され、発生符号化量が量子化制御部4に供給する。量子化制御部4は、視覚特性演算部6から供給される視覚特性制御情報と発生符号量係数部7から供給される発生符号量とを用いて演算処理し、量子化制御情報を量子化器3に供給する。 The blocked image signal is also supplied to the visual characteristic calculation unit 6, and calculation processing according to the human visual characteristic is performed on the blocked image signal, and control information based on the visual characteristic is sent to the quantization control unit 6. Supply. The encoded data output from the encoder 5 is supplied to the generated encoded amount coefficient unit 7, and the generated encoded amount is supplied to the quantization control unit 4. The quantization control unit 4 performs arithmetic processing using the visual characteristic control information supplied from the visual characteristic calculation unit 6 and the generated code amount supplied from the generated code amount coefficient unit 7, and the quantization control information is quantized. 3 is supplied.
図2は視覚特性演算部6の一例を示す線図である。輝度信号Yは、輝度信号分散値測定部10に供給され、輝度信号の分散値(アクティビティ)が測定され、輝度信号分散値情報として量子化制御部4に供給する。また、輝度信号は直流成分測定部11にも供給され、輝度信号の直流成分を検出する。尚、直流成分は、マクロブロックの平均値を表すものとする。 FIG. 2 is a diagram showing an example of the visual characteristic calculation unit 6. The luminance signal Y is supplied to the luminance signal variance value measuring unit 10, the variance value (activity) of the luminance signal is measured, and supplied to the quantization control unit 4 as luminance signal variance value information. The luminance signal is also supplied to the DC component measurement unit 11 to detect the DC component of the luminance signal. The DC component represents the average value of the macroblock.
Y信号及び2つの色差信号Pr,Pb は受像三原色RGBの直流成分(マクロブロックの平均値)を測定するRGB直流成分測定部12に供給する。RGB直流成分測定部12は、一例として以下の変換式に基づいて伝送三原色の直流成分[R],[G],[B]を求める。
Y =CrR+CgG+CbB
Pr=(B−Y)/{2(1−Cb )} (1)
Pb=(R−Y)/{2(1−Cr )}
ここで、Cr+Cg+Cb =1であり、ITU-R Rec.BT.709の場合、Cr =0.2126、Cg =0.7152、Cb =0.0722である。
上記(1)式より、三原色RGBは以下の式で表される。
R=Y+1.5748Pr
G=Y−0.1873Pb −0.4681Pr (2)
B=Y+1.8556Pb
RGB直流成分測定部12は、入力したY信号及び2つの色差信号Pr,Pb から式(2)を用いて三原色RGBの輝度値を求め、求めた輝度値からマクロブロックの平均値である直流成分[R],[G],[B]を求める。尚、上記変換式の数値は一例であり、各種画像符号化方式に対応した変換式の数値が用いられ、例えばDVDのNTSC方式の場合、ITU-R Rec. BT.601において Cr=0.299、Cg=0.587、Cb=0.114が用いられる。
The Y signal and the two color difference signals P r and P b are supplied to an RGB DC component measuring unit 12 that measures DC components (average values of macroblocks) of the image receiving three primary colors RGB. For example, the RGB direct current component measurement unit 12 obtains direct current components [R], [G], and [B] of transmission three primary colors based on the following conversion formula.
Y = C r R + C g G + C b B
P r = (BY) / {2 (1-C b )} (1)
P b = (R−Y) / {2 (1-C r )}
Here, C r + C g + C b = 1, and in the case of ITU-R Rec.BT.709, C r = 0.2126, C g = 0.7152, and C b = 0.0722.
From the above equation (1), the three primary colors RGB are expressed by the following equations.
R = Y + 1.5748P r
G = Y−0.1873P b −0.4681P r (2)
B = Y + 1.8556P b
The RGB direct current component measurement unit 12 obtains the luminance value of the three primary colors RGB from the input Y signal and the two color difference signals P r and P b using the equation (2), and is an average value of the macroblock from the obtained luminance value. DC components [R], [G], [B] are obtained. The numerical values of the conversion formula is one example, various image encoding method used is numeric conversion equation corresponding to, for example, the NTSC system of DVD, ITU-R Rec. BT.601 C r = 0.299 in, C g = 0.587 and C b = 0.114 are used.
求めたRGBの直流成分は、比率測定部13に供給する。この比率測定部13には輝度信号の直流成分[Y]も入力する。この比率測定部13においては、輝度信号の直流成分[Y]と三原色RGBの直流成分[R],[G],[B]との比率、[Y]/[R]、[Y]/[G]、及び[Y]/[B]をそれぞれ求める。この輝度信号に対する比率を求めることにより、ブロック中の彩度の高い画素が占める相対的な割合が求められる。すなわち、例えば、比率[Y]/[R]が一番小さい場合、当該ブロックは、彩度の高い赤の画素が占める割合が高いことを意味する。一方、比率[Y]/[R]の値が相対的に大きい場合、当該ブロックは、赤の画素が占める割合が小さいことを意味する。 The obtained RGB direct current components are supplied to the ratio measuring unit 13. The ratio measuring unit 13 also receives a DC component [Y] of the luminance signal. In this ratio measuring unit 13, the ratio between the DC component [Y] of the luminance signal and the DC components [R], [G], [B] of the three primary colors RGB, [Y] / [R], [Y] / [ G] and [Y] / [B] are obtained, respectively. By calculating the ratio with respect to the luminance signal, the relative ratio occupied by the highly saturated pixels in the block can be determined. That is, for example, when the ratio [Y] / [R] is the smallest, this means that the block occupies a high ratio of red pixels with high saturation. On the other hand, when the value of the ratio [Y] / [R] is relatively large, it means that the proportion of red pixels in the block is small.
求めた3つの比率[Y]/[R]、[Y]/[G]、及び[Y]/[B]は最小値検出部14に供給し、3つの比率のうち最も小さい比率の値をNとして検出する。すなわち、N=min{[Y]/[R],[Y]/[G],[Y]/[B]}を求める。 The obtained three ratios [Y] / [R], [Y] / [G], and [Y] / [B] are supplied to the minimum value detection unit 14, and the value of the smallest ratio among the three ratios is set. Detect as N. That is, N = min {[Y] / [R], [Y] / [G], [Y] / [B]} is obtained.
求めた最小値Nは、パラメータ発生部15に供給する。このパラメータ発生部15は、関数f(N)=k(N−1)+1 (0<k≦1)を有し、この関数に最小値Nを挿入して、パラメータf(N)を発生する。このパラメータは、輝度信号分散値と共に量子化制御部4に供給する。上記関数f(N)は、カラーの影響すなわち彩度による劣化の改善を図る場合、係数kの値を大きく設定し、彩度の高い画素が占める割合の高いブロックについて量子化値が小さくなるように作用する。一方、カラーに対する補正ないし補償作用を小さくする場合係数kの値を小さく設定する。 The obtained minimum value N is supplied to the parameter generator 15. The parameter generator 15 has a function f (N) = k (N−1) +1 (0 <k ≦ 1), and inserts a minimum value N into this function to generate a parameter f (N). . This parameter is supplied to the quantization control unit 4 together with the luminance signal dispersion value. In the function f (N), when the influence of color, that is, deterioration due to saturation is to be improved, the value of the coefficient k is set to be large so that the quantization value becomes small for a block having a high ratio of pixels having high saturation. Act on. On the other hand, when the correction or compensation action for the color is reduced, the value of the coefficient k is set small.
図3は、量子化制御部4の一例を示す線図である。視覚特性演算部から供給される輝度信号分散値及び発生符号量係数部から供給される発生符号量は量子化制御情報発生部20に供給する。この量子化制御情報発生部20は、前のブロックの符号量と輝度信号の分散値とに基づいて量子化制御情報Q0を発生する。すなわち、目標符号量と過去の符号量とを比較して平均符号量が目標符号量に近くなるように制御すると共に、輝度信号の分散値が大きいブロックについては量子化値が大きくなるように制御し輝度信号の分散値が小さいブロックについては量子化値が小さくなるように制御する。例えば、過去のマクロブロックの分散値の平均と当該マクロブロックの分散値の平均との比を目標符号量から求まる量子化値に乗算して量子化制御情報を発生する。尚、量子化制御情報発生部20は、MPEG-2に準拠した既存の量子化制御情報発生手段を用いることができる。 FIG. 3 is a diagram illustrating an example of the quantization control unit 4. The luminance signal variance value supplied from the visual characteristic calculation unit and the generated code amount supplied from the generated code amount coefficient unit are supplied to the quantization control information generating unit 20. The quantization control information generation unit 20 generates quantization control information Q 0 based on the code amount of the previous block and the variance value of the luminance signal. In other words, the target code amount is compared with the past code amount and the average code amount is controlled to be close to the target code amount, and the quantization value is controlled for the block having a large variance value of the luminance signal. For a block with a small variance value of the luminance signal, the quantization value is controlled to be small. For example, the quantization control information is generated by multiplying the quantization value obtained from the target code amount by the ratio of the average of the dispersion values of the past macroblocks and the average of the dispersion values of the macroblocks. The quantization control information generation unit 20 can use existing quantization control information generation means compliant with MPEG-2.
量子化制御情報Q0 及びパラメータf(N)は、補正部21に供給する。この補正部21は、例えば乗算器で構成することができ、量子化制御情報Q0 にパラメータf(N)を乗算して、パラメータf(N)によりカラー補正された量子化制御情報Qを出力することができる。すなわち、従来の符号化装置により得られる量子化制御情報Q0に、例えば彩度の高い画素が占める割合を表す情報であるパラメータf(N)を乗算することにより、彩度の高い画素が占める割合の高いブロックないし領域について、その割合に応じて量子化値が補正ないし補償されるため、彩度の高い領域の劣化を改善することができる。例えば、彩度の高い画素が占める割合の高いブロックについては、量子化制御情報Q0 に対して量子化値が小さくなるように関数f(N)を用いて量子化制御情報を補正することにより、彩度の高いブロックの画質の劣化を改善することができる。 The quantization control information Q 0 and the parameter f (N) are supplied to the correction unit 21. The correction unit 21 can be configured by a multiplier, for example, and multiplies the quantization control information Q 0 by the parameter f (N), and outputs the quantization control information Q color-corrected by the parameter f (N). can do. That is, by multiplying the quantization control information Q 0 obtained by the conventional encoding apparatus by, for example, the parameter f (N) that is information indicating the ratio of pixels with high saturation, pixels with high saturation occupy. Since the quantized value is corrected or compensated according to the ratio or the block or area having a high ratio, it is possible to improve the deterioration of the highly saturated area. For example, for a block having a high proportion of pixels with high saturation, the quantization control information is corrected using the function f (N) so that the quantization value becomes smaller than the quantization control information Q 0 . Therefore, it is possible to improve the deterioration of the image quality of blocks with high saturation.
本発明は上述した実施例だけに限定されず、種々の変更や変形が可能である。例えば、上述した実施例では、量子化制御情報にパラメータf(N)を乗算する構成を採用したが、量子化制御情報にパラメータf(N)を加算することにより量子化制御情報を補正することも可能である。また、補正用のパラメータの生成方法は一例であり、輝度信号の直流成分と三原色RGBの直流成分との比率を用いて種々の補正用のパラメータを生成することも可能である。 The present invention is not limited to the above-described embodiments, and various changes and modifications can be made. For example, in the above-described embodiment, the configuration in which the quantization control information is multiplied by the parameter f (N) is adopted. However, the quantization control information is corrected by adding the parameter f (N) to the quantization control information. Is also possible. Further, the method for generating correction parameters is merely an example, and various correction parameters can be generated using the ratio between the DC component of the luminance signal and the DC components of the three primary colors RGB.
さらに、上述した実施例では、輝度信号の直流成分と三原色RGBの直流成分との比率として、{[Y]/[R],[Y]/[G],[Y]/[B]}を用いたが、{[R]/[Y],[G]/[Y],[B]/[Y]}を用いることもでき、この場合これらの比率のうち最大値をNとして用いる。 Furthermore, in the above-described embodiment, {[Y] / [R], [Y] / [G], [Y] / [B]} is used as the ratio of the direct current component of the luminance signal and the direct current component of the three primary colors RGB. Although used, {[R] / [Y], [G] / [Y], [B] / [Y]} can also be used. In this case, the maximum value of these ratios is used as N.
さらに、上述した実施例では、1つのマクロブロックについて輝度信号と三原色RGBとの比率を算出したが、マクロブロックを4分割又は16分割、あるそれ以上に細分割した領域ないしブロックを用いて比率を算出することも可能である。 Furthermore, in the above-described embodiment, the ratio between the luminance signal and the three primary colors RGB is calculated for one macroblock. However, the ratio is determined by using a region or block obtained by subdividing the macroblock into four or sixteen, or more than that. It is also possible to calculate.
1 分割手段
2 直交変換器
3 量子化器
4 量子化制御部
5 符号化器
6 視覚特性演算部
7 発生符号量係数部
10 輝度信号分散値測定部
11 直流成分測定部
12 RGB直流成分測定部
13 比率測定部
14 最小値検出部
15 パラメータ発生部
20 量子化制御情報発生部
21 補正部
DESCRIPTION OF SYMBOLS 1 Dividing means 2 Orthogonal transformer 3 Quantizer 4 Quantization control part 5 Encoder 6 Visual characteristic calculation part 7 Generated code amount coefficient part 10 Luminance signal dispersion value measurement part 11 DC component measurement part 12 RGB DC component measurement part 13 Ratio measurement unit 14 Minimum value detection unit 15 Parameter generation unit 20 Quantization control information generation unit 21 Correction unit
Claims (3)
符号化すべきカラー画像信号を所定の画素数のブロックに分割する分割手段と、
分割されたカラー画像信号について、各ブロック毎に直交変換して変換係数を出力する直交変換手段と、
量子化制御情報に基づき、前記直交変換係数を量子化して量子化値を出力する量子化手段と、
量子化値を符号化して符号化データを出力する符号化手段と、
符号化データから発生符号量を計数する発生符号量計数手段と、
各ブロックの輝度信号及び2つの色差信号を受け取り、視覚特性に基づく制御情報を出力する視覚特性演算手段と、
前記発生符号量計数手段及び視覚特性演算手段からの出力情報に基づいて量子化制御情報を出力する量子化制御手段とを具えるカラー画像符号化装置において、
前記視覚特性演算手段は、各ブロックの輝度信号の分散値を算出する手段と、 輝度信号の直流成分を出力する手段と、輝度信号及び2つの色差信号から三原色RGBの直流成分を測定する手段と、前記輝度信号の直流成分と三原色RGBの直流成分との比率を演算する手段とを有し、
前記量子化制御手段は、輝度信号の分散値と発生符号量とに基づいて量子化制御情報(Q0)を発生する量子化制御情報発生手段と、前記輝度信号の直流成分と三原色の直流成分との比率を用いて量子化制御情報を補正する手段とを有し、補正された量子化制御情報(Q)を量子化手段に供給することを特徴とするカラー画像信号符号化装置。 A color image encoding apparatus that encodes a color image signal composed of a luminance signal (Y) and two color difference signals (P b , P r ),
Dividing means for dividing the color image signal to be encoded into blocks of a predetermined number of pixels;
For the divided color image signal, orthogonal transform means for performing orthogonal transform for each block and outputting transform coefficients;
Based on quantization control information, quantization means for quantizing the orthogonal transform coefficient and outputting a quantization value;
Encoding means for encoding the quantized value and outputting encoded data;
Generated code amount counting means for counting the generated code amount from the encoded data;
Visual characteristic calculation means for receiving a luminance signal of each block and two color difference signals and outputting control information based on the visual characteristics;
In a color image encoding device comprising quantization control means for outputting quantization control information based on output information from the generated code amount counting means and visual characteristic calculation means,
The visual characteristic calculation means includes means for calculating a variance value of the luminance signal of each block, means for outputting a DC component of the luminance signal, means for measuring the DC component of the three primary colors RGB from the luminance signal and the two color difference signals, and And means for calculating the ratio of the direct current component of the luminance signal and the direct current component of the three primary colors RGB,
The quantization control means includes: quantization control information generation means for generating quantization control information (Q 0 ) based on a variance value of a luminance signal and a generated code amount; a direct current component of the luminance signal and a direct current component of three primary colors And a means for correcting the quantization control information by using the ratio of the color image signal, and supplying the corrected quantization control information (Q) to the quantization means.
前記量子化制御手段は、前記関数f(N)を用いて前記量子化制御情報(Q0)を補正することを特徴とする請求項1に記載のカラー画像信号符号化装置。 The visual characteristic calculation means includes means for obtaining a minimum value or a maximum value (N) of a ratio of a direct current component of the luminance signal and a direct current component of the three primary colors, and a function f including the minimum value or the maximum value (N) as a variable. And (N) generating means,
The color image signal encoding apparatus according to claim 1, wherein the quantization control unit corrects the quantization control information (Q 0 ) using the function f (N).
3. The color image signal encoding apparatus according to claim 2, wherein when the function f (N) is a coefficient having a value of 0 <k ≦ 1, f (N) = k (N−1). A color image signal encoding apparatus using a function represented by +1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2006062051A JP4714044B2 (en) | 2006-03-08 | 2006-03-08 | Color image signal encoding apparatus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2006062051A JP4714044B2 (en) | 2006-03-08 | 2006-03-08 | Color image signal encoding apparatus |
Publications (2)
Publication Number | Publication Date |
---|---|
JP2007243515A JP2007243515A (en) | 2007-09-20 |
JP4714044B2 true JP4714044B2 (en) | 2011-06-29 |
Family
ID=38588612
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP2006062051A Expired - Fee Related JP4714044B2 (en) | 2006-03-08 | 2006-03-08 | Color image signal encoding apparatus |
Country Status (1)
Country | Link |
---|---|
JP (1) | JP4714044B2 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10298888B2 (en) | 2015-08-25 | 2019-05-21 | Kabushiki Kaisha Toshiba | Adaptive analyzer of sensitivity difference of same color pixels and adaptive corrector of sensitivity difference of same color pixels |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5488168B2 (en) * | 2010-04-27 | 2014-05-14 | パナソニック株式会社 | Image encoding device |
JP6066583B2 (en) * | 2012-05-01 | 2017-01-25 | キヤノン株式会社 | Moving picture coding apparatus and moving picture coding method |
JP7065594B2 (en) | 2017-12-08 | 2022-05-12 | キヤノン株式会社 | Image coding device and its control method, and program |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0923423A (en) * | 1995-05-02 | 1997-01-21 | Sony Corp | Signal encoding method and device and signal recording medium |
JPH10164581A (en) * | 1996-12-03 | 1998-06-19 | Sony Corp | Method and device for coding image signal and signal-recording medium |
JPH10341436A (en) * | 1997-06-06 | 1998-12-22 | Matsushita Electric Ind Co Ltd | Highly efficient encoding device |
JP2004015340A (en) * | 2002-06-05 | 2004-01-15 | Kddi R & D Laboratories Inc | Quantization control method for encoding moving picture |
-
2006
- 2006-03-08 JP JP2006062051A patent/JP4714044B2/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0923423A (en) * | 1995-05-02 | 1997-01-21 | Sony Corp | Signal encoding method and device and signal recording medium |
JPH10164581A (en) * | 1996-12-03 | 1998-06-19 | Sony Corp | Method and device for coding image signal and signal-recording medium |
JPH10341436A (en) * | 1997-06-06 | 1998-12-22 | Matsushita Electric Ind Co Ltd | Highly efficient encoding device |
JP2004015340A (en) * | 2002-06-05 | 2004-01-15 | Kddi R & D Laboratories Inc | Quantization control method for encoding moving picture |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10298888B2 (en) | 2015-08-25 | 2019-05-21 | Kabushiki Kaisha Toshiba | Adaptive analyzer of sensitivity difference of same color pixels and adaptive corrector of sensitivity difference of same color pixels |
Also Published As
Publication number | Publication date |
---|---|
JP2007243515A (en) | 2007-09-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8204105B2 (en) | Transcoder and coded image conversion method | |
US10136033B2 (en) | Techniques for advanced chroma processing | |
US6430222B1 (en) | Moving picture coding apparatus | |
US6122400A (en) | Compression encoder bit allocation utilizing colormetric-adaptive weighting as in flesh-tone weighting | |
US7974340B2 (en) | Adaptive B-picture quantization control | |
US11336904B2 (en) | Video coding using a saliency map | |
WO2008020687A1 (en) | Image encoding/decoding method and apparatus | |
US20190116383A1 (en) | Techniques for video compression | |
KR20170137076A (en) | Quantization parameter (qp) calculation for display stream compression (dsc) based on complexity measure | |
US7676107B2 (en) | Method and system for video classification | |
JP2018101867A (en) | Image processing device, image processing method and program | |
AU4875899A (en) | Encoding apparatus and method | |
JP2018101866A (en) | Image processing device, image processing method and program | |
JP2006005650A (en) | Device and method for image encoding | |
JP4714044B2 (en) | Color image signal encoding apparatus | |
US20060256858A1 (en) | Method and system for rate control in a video encoder | |
JP3429429B2 (en) | Visual error processing method and image coding apparatus using this method | |
JP2007081474A (en) | Image coding apparatus and method | |
JP3437901B2 (en) | Video encoding method and apparatus | |
TW201711466A (en) | Video encoding device, video decoding device, video encoding method, video decoding method, and program | |
US20060008004A1 (en) | Video encoder | |
KR20150015460A (en) | Dynamic quantisation method for video encoding | |
JP2003289542A (en) | Method, equipment, and program for coding image information, method, equipment, and program for decoding image information, image information coding/ decoding method, and image information coding transmission system | |
US20150010073A1 (en) | Dynamic quantization method for encoding data streams | |
US8929433B2 (en) | Systems, methods, and apparatus for improving display of compressed video data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20080606 |
|
A01 | Written decision to grant a patent or to grant a registration (utility model) |
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20110301 |
|
A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20110325 |
|
R250 | Receipt of annual fees |
Free format text: JAPANESE INTERMEDIATE CODE: R250 |
|
R250 | Receipt of annual fees |
Free format text: JAPANESE INTERMEDIATE CODE: R250 |
|
LAPS | Cancellation because of no payment of annual fees |