JP6190130B2 - Color signal conversion device, color signal restoration device, and program - Google Patents

Description

  The present invention relates to a color signal conversion device that converts color signals into luminance / color difference signals, a color signal restoration device that converts luminance / color difference signals into color signals, and a program.

Conventionally, when transmitting a color signal (RGB signal) as a video signal, conversion to a luminance / color difference signal (YCC signal) is performed. This conversion is performed by gamma correction, matrix calculation, and LPF (low-pass type). (For example, see Patent Document 1). FIG. 9 shows a conventional color signal converter for converting a color signal into a luminance / color difference signal. Here, the input color signal is expressed as r, g, and b in proportion to the light of each color of R, G, B, and the transmitted luminance / color difference signal is expressed as Y, C ₁ , C _2. doing. FIG. 10 shows a conventional color signal restoration apparatus that converts luminance / color difference signals into color signals. Since the output signal of the color signal restoration device 7 does not match the input signals r, g, and b of the color signal conversion device 6, they are represented as r ₁ , g ₁ , b ₁ .

The gamma correction unit 10 performs gamma conversion for converting a signal proportional to light intensity into a video signal. If the input is X and the output is Y, Y = ^Xγ . Here, γ is a constant, and usually γ = 0.45.

The matrix calculation unit 12 converts the input color signals R, G, and B into luminance / color difference signals Y, C ₁ , and C ₂ by 3 × 3 linear conversion expressed by Equation (1). Here, C _R , C _G , and C _B are constants indicating contributions from R, G, and B to luminance, and C _R + C _G + C _B = 1.0. For example, in the case of high vision, C _R = 0.2126, C _G = 0.7152, and C _B = 0.0722.

  The LPF unit 30 is a low-pass filter that allows only low frequency components to pass. The input is X (i, j), the output is Y (i, j), and the coefficient is Cnm (n = −N to N, m = -M to M), the LPF process is expressed by the following equation (2). Here, i and j are pixel positions, and N and M are the number of filter taps.

  The inverse matrix calculation unit 22 performs reverse conversion of the matrix calculation unit 12, and the conversion equation is expressed by equation (3).

The inverse gamma correction unit 40 performs the reverse conversion of the gamma correction unit 10, and the conversion formula is represented by X = Y ^{1 / γ} .

Japanese Patent No. 3253316

The problem of the conventional method will be described below. Although the image is two-dimensional, it will be described in one dimension for easy explanation. That is, input signals to the color signal converter are r (i), g (i), b (i), and output signals from the color signal converter are Y (i), C ₁ (2i), C ₂ ( 2i), i = 1, 2,.

  As a first problem, there is a problem that, depending on the input image, the output signal may be a negative signal that cannot be physically displayed even if it can be expressed in a mathematical expression. This point will be described by taking as an example a case where a fine stripe pattern of the signal R expressed by the equation (4) is input. Further, only matrix conversion and inverse matrix conversion are considered.

Here, G (i) = B (i) = 0 means that the signals G and B have the same value in all i. Signals Y, C ₁ and C ₂ after matrix conversion are expressed by equation (5).

When the luminance / color difference signals Y, C ₁ , C ₂ are subjected to inverse matrix conversion, the output signals R ₁ , G ₁ , B ₁ are expressed by equation (6).

As shown in the equation (6), there is no problem because the value of the signal R ₁ is positive, but the values of the even-numbered terms are negative for the signals G ₁ and B ₁ and cannot be physically reproduced.

As a second problem, there is a problem that the signal deteriorates when the conversion is repeatedly performed. Since LPF is used for signal processing, the signal deteriorates when the conversion of the signals r, g, b and the signals Y, C ₁ , C ₂ is repeated. When the input X of the LPF is a single frequency f expressed by the equation (7), the output Y of the LPF changes as shown in the equation (8) according to the frequency.

In an ideal LPF, F ₁ = F ₂ and there are no two items in the equation (8). However, the LPF that is actually used has a transition area indicated by two items because the number of taps cannot be increased. As an illustrative example, the frequency of F ₁ <f <F ₂ is taken, and the signals r, g, and b input to the color signal converter are given by equation (9), and gamma correction and inverse gamma correction are ignored for simplicity.

At this time, the signals Y, C ₁ , and C ₂ output from the color signal conversion device 6 are expressed by Equation (10).

When the signals Y, C ₁ and C ₂ in the equation (10) are input to the color signal restoration device 7, the signals r, g, and b output from the color signal restoration device 7 are represented by the equation (11). Since the value of the equation (11) is different from the r, g, and b values of the equation (9), they are expressed as r ₁ , g ₁ , b ₁ .

Further, when the signal of the equation (11) is input to the color signal conversion device 6, the signals Y, C ₁ and C ₂ output from the color signal conversion device 6 become the equation (12).

When the conversion of the transmission signal is repeated in this way, the signal value is different and deteriorates. The difference between the input signals r, g, and b and the signals r ₁ , g ₁ , and b ₁ reproduced after transmission is considered to be natural because the transmission signal is subjected to LPF filtering on the color difference signal, but the conversion is repeated. It is not desirable that the signal deteriorates each time it is performed.

  An object of the present invention made in view of such circumstances is to improve signal clipping caused by conversion from a color signal to a luminance / color difference signal and conversion from a luminance / color difference signal to a color signal, and signal deterioration due to repeated conversion. Another object of the present invention is to provide a color signal conversion device, a color signal restoration device, and a program that can be used.

In order to solve the above problems, a color signal conversion apparatus according to the present invention is a color signal conversion apparatus for converting a color signal into a luminance / color difference signal, and linearly converts the input color signals r, g, and b. And a matrix operation unit for generating the luminance signal L, and averaging processing for each of the predetermined blocks for the luminance signal L and the color signals r and b to generate signals L _ave , r _ave , and b _ave . The averaging unit, the luminance signal L, and the signals L _ave , r _ave , b _ave are subjected to gamma conversion, and signals Γ (L), Γ (L _ave ), Γ (r _ave ), Γ (b _ave ), a difference signal between the signal Γ (r _ave ) and the signal Γ (L _ave ), and a difference signal between the signal Γ (b _ave ) and the signal Γ (L _ave ) Subsampling for subsampling processing Characterized in that it comprises a and.

In order to solve the above-described problems, a color signal restoration device according to the present invention is a color signal restoration device that converts a luminance / color difference signal into a color signal, performs linear conversion on the luminance / color difference signal, and converts the color signal into a color signal. An inverse matrix calculation unit to be generated, an interpolation unit that performs interpolation processing on the color signal to generate low frequency component signals R _low , G _low , and B _low of the color signal, and the signals R _low , G _low , B _low for performing linear transformation to generate a low frequency component of the luminance signal, a matrix calculation unit for generating a coefficient W indicating a ratio of the high frequency component to the low frequency component of the luminance signal, A multiplier for generating signals R _low · W, G _low · W, B _low · W by multiplying the low frequency components R _low , G _low , B _low of the color signal by a coefficient W, and the signals R _low · W, G _low · W, B performs averaging processing for each predetermined block against _ow · W, the signal _{R ave,} G _ave, an averaging unit for generating a _{B ave,} the signal in addition to the signal _{R low} · W to the signal _{R low} R signal _{R 1} obtained by subtracting _ave, the signal _{G low} to the signal _{G low} · W signals _{G 1} by subtracting the signal _{G ave} in addition, and the signal _{B low} · W were added the signal to the signal _{B low} A means for generating a signal B ₁ obtained by subtracting B _ave and an inverse gamma correction unit for performing inverse gamma conversion on the signals R ₁ , G ₁ , B ₁ are provided.

The color signal restoration apparatus according to the present invention is a color signal restoration apparatus that converts a luminance / color difference signal into a color signal, and performs an inverse matrix operation that performs linear conversion on the luminance / color difference signal to generate a color signal. A reverse matrix calculation unit that performs linear conversion on the luminance / color difference signal to generate a color signal, an inverse gamma correction unit that performs reverse gamma conversion on the luminance signal and the color signal, and the reverse gamma conversion An interpolation unit that performs interpolation processing on the color signals thus generated and generates low-frequency component signals R _low , G _low , B _low of the color signals, and linear conversion on the signals R _low , G _low , B _low It was carried out, and a matrix calculation unit for generating a low-frequency component of the inverse gamma converted luminance signal, and a coefficient generation unit for generating a coefficient W indicating a proportion of a high frequency component to low frequency component of the inverse gamma converted luminance signal, A multiplier for generating signals R _low · W, G _low · W, B _low · W by multiplying the low frequency components R _low , G _low , B _low of the color signal by a coefficient W; and the signals R _low · W, G _low · W and B _low · W are averaged for each predetermined block to generate signals R _ave , G _ave , and B _ave , and the signal R _low · W is _added to the signal R _low. signal _{r 1} obtained by subtracting the signal _{R ave} added, the signal _{B low} · the signal signal _{g 1} was added to _{G low} · W subtracting the signal _{G ave,} and the signal _{B low} to the signal _{G low} Means for generating a signal b ₁ by adding W and subtracting the signal B _ave .

Furthermore, in the color signal restoration device according to the present invention, the interpolation unit performs zero-order interpolation on the color signal to generate the signals R _low , G _low , and B _low .

Alternatively, in the color signal restoration apparatus according to the present invention, the interpolation unit performs a zero-order interpolation on the color signal, and generates a signal R _ip , G _ip , B _ip , An LPF unit for extracting low-frequency component signals R _lpf , G _lpf , and B _lpf of the signals R _ip , G _ip , and B _ip , and an average process for each predetermined block with respect to the signals R _lpf , G _lpf , and B _lpf And an averaging unit that generates signals (R _lpf ) _ave , (G _lpf ) _ave , and (B _lpf ) _ave , and the signals R _ipf , G _{lpf are included in the} signals R _ip , G _ip , B _ip , B _lpf is added to subtract the signals (R _lpf ) _ave , (G _lpf ) _ave , (B _lpf ) _ave to generate the signals R _low , G _low , B _low .

  In order to solve the above problems, a program according to the present invention causes a computer to function as the color signal conversion device.

  In order to solve the above problems, a program according to the present invention causes a computer to function as the color signal restoration device.

  According to the present invention, it is possible to improve signal clipping caused by conversion from a color signal to a luminance / color difference signal and conversion from a luminance / color difference signal to a color signal, and signal degradation due to repeated conversion.

1 is a block diagram illustrating a configuration of a color signal conversion apparatus according to a first embodiment of the present invention. It is a block diagram which shows the structure of the conversion process part in the color signal converter which concerns on the 1st Embodiment of this invention. 1 is a block diagram illustrating a configuration of a color signal restoration device according to a first embodiment of the present invention. It is a block diagram which shows the structure of the decompression | restoration process part in the color signal decompression | restoration apparatus which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the structure of the decompression | restoration process part in the color signal decompression | restoration apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the structure of the interpolation filter part in the color signal decompression | restoration apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the structure of the color signal converter which concerns on the 3rd Embodiment of this invention. It is a block diagram which shows the structure of the color signal decompression | restoration apparatus which concerns on the 3rd Embodiment of this invention. It is a block diagram which shows the structure of the conventional color signal converter. It is a block diagram which shows the structure of the conventional color signal decompression | restoration apparatus.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram showing a configuration of a color signal conversion apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the color signal conversion apparatus 1 includes a gamma correction unit 10 and a conversion processing unit 11. The gamma correction unit 10 is the same as the gamma correction unit 10 of the conventional color signal conversion device 6 shown in FIG. 9, and performs a process of converting a signal proportional to light intensity into a video signal. The conversion processing unit 11 performs color conversion processing using averaging processing without using LPF.

  FIG. 2 is a block diagram illustrating a configuration of the conversion processing unit 11 in the color signal conversion apparatus 1. The conversion processing unit 11 includes a matrix calculation unit 12, an averaging unit 13, and a subsampling unit 14.

For the following description, the number of pixels of the luminance signal Y output from the color signal converter 1 is I × J, and the number of pixels of the color difference signals C ₁ and C ₂ output from the color signal converter 1 is I ₁ × J. _Set to ₁ . Further, the ratio of I, I ₁ and J, J ₁ is represented by equation (13).

  The matrix calculation unit 12 is the same as the matrix calculation unit 12 of the conventional color signal conversion device 4 shown in FIG. 9, and performs the 3 × 3 linear conversion expressed by the equation (1) to generate the luminance / color difference signal. Generate and output to the averaging unit 13.

The averaging unit 13 averages the input signal for each block of P _H × P _V pixels, and outputs the averaged signal to the sub-sampling unit 14. The number of pixels of the input signal is I × J, the input image is Z (i, j) (where i = 1 to I, j = 1 to J), and the function for performing the averaging process is expressed as Ave (Z). To do. The number of pixels of the output signal of the averaging unit 13 is also I × J pixels, and when the output signal is expressed as Ave (Z) (i, j) (where i = 1 to I, j = 1 to J), Ave (Z) (i, j) is calculated by equation (14). Here, i and j in the equation (14) are expressed by the equation (15). I _H and j _V in the equation (15) are represented by the equation (16). int () is a function that converts a real number into an integer by truncation.

  Further, the equation (17) is established from the property of taking an average.

The subsampling unit 14 performs subsampling processing on the signal input from the averaging unit 13. The number of pixels of the input signal is I × J, the input image is Z (i, j) (where i = 1 to I, j = 1 to J), and the function for performing the sub-sampling process is SS (Z). write. The number of pixels of the output signal of the sub-sampling unit 14 is I ₁ × J ₁ , and the output signal is U (i ₁ , j ₁ ) (where i ₁ = 1 to I ₁ , j ₁ = ₁ to J ₁ ). In terms of notation, U (i ₁ , j ₁ ) is calculated by equation (18).

When the input video signal of the conversion processing unit 11 is R (i, j), G (i, j), B (i, j), the output signals Y (i, j), C _{1 of the} color signal conversion device ₁ (i ₁ , j ₁ ) and C ₂ (i ₁ , j ₁ ) are expressed by the equation (19).

  Since Ave (Y (i, j)) in the second and third terms is a value obtained from the first term, the color signal converting apparatus 1 is equivalently transmitting the value of the equation (20). become.

  Furthermore, from the equation (21), the color signal converter 1 is equivalent to transmitting the value of the equation (22).

  Next, a color signal restoration device that restores a color signal from the luminance / color difference signal input from the color signal conversion device 1 will be described.

  FIG. 3 is a block diagram showing the configuration of the color signal restoration apparatus 2 according to the first embodiment of the present invention. As shown in FIG. 3, the color signal restoration device 2 includes a restoration processing unit 20 and an inverse gamma correction unit 40. The restoration processing unit 20 restores the color signal from the luminance / color difference signal input from the color signal conversion apparatus 1. The inverse gamma correction unit 40 is the same as the inverse gamma correction unit 40 of the conventional color signal restoration device 7 shown in FIG. 10 and performs reverse conversion of the gamma correction unit 10.

  FIG. 4 is a block diagram illustrating a configuration of the restoration processing unit 20 in the color signal restoration device 2. As illustrated in FIG. 4, the restoration processing unit 20 includes a coefficient generation unit 21, a multiplication unit 27, an averaging unit 13, a subtraction unit 25, and an addition unit 28.

  The coefficient generation unit 21 generates a coefficient W indicating the ratio of the high frequency component to the low frequency component of the luminance signal. Specifically, the coefficient generation unit 21 includes an averaging unit 13, a subsampling unit 14, an inverse matrix calculation unit 22, an interpolation unit 23, a matrix calculation unit 24, a subtraction unit 25, and a division unit 26. With. The averaging unit 13 and the subsampling unit 14 perform the same processing as the averaging unit 13 and the subsampling unit 14 in the color signal conversion apparatus 1.

The inverse matrix calculation unit 22 performs the same processing as the inverse matrix calculation unit 22 of the conventional color signal restoration device 7, and performs SS (Ave) which is an input signal by 3 × 3 linear conversion expressed by the equation (3). (Y (i, j))), C ₁ (i ₁ , j ₁ ), C ₂ (i ₁ , j ₁ ), SS (Ave (R (i, j))), SS (Ave (G ( i, j))), SS (Ave (B (i, j))) are generated and output to the interpolation unit 23.

The interpolation unit 23 performs an interpolation process on the input signal from the inverse matrix calculation unit 22 to generate low frequency components R _low , G _low , and B _low of the color signal, and outputs them to the matrix calculation unit 24 and the multiplication unit 27. To do. The number of pixels of the input signal is I ₁ × J ₁ , the input image is U (i ₁ , j ₁ ) (where i ₁ = 1 to I ₁ , j ₁ = ₁ to J ₁ ), and interpolation processing is performed. The function to be performed is expressed as IP (U). The number of pixels of the output signal of the interpolation unit 23 is I × J, and when the output signal is expressed as Z (i, j) (where i = 1 to I, j = 1 to J), Z (i, j ) Is calculated by equation (23). Here, simple zero-order interpolation is used. Here, i _H and j _V in equation (22) are as shown in equation (16).

Since the input signals of the interpolation unit 23 are SS (Ave (R (i, j))), SS (Ave (G (i, j))), SS (Ave (B (i, j))), The output signal of the interpolation unit 23 is R _low = IP (SS (Ave (R (i, j)))), G _low = IP (SS (Ave (G (i, j))))), B _low = IP (SS (Ave (B (i, j)))).

The matrix calculation unit 24 performs linear transformation (matrix calculation) represented by the equation (24) on R _low , G _low , and B _low that are input signals to obtain IP (SS (Ave (Y (i, j)). ))) Is generated.

  The subtraction unit 25 and the division unit 26 obtain the coefficient W (i, j), which is the output signal of the coefficient generation unit 21, using equation (25). Since IP (SS (Ave (Z))) = Ave (Z), equation (25) can be expressed by equation (26).

The multiplier 27 generates signals R _low · W, G _low · W, B _low · W obtained by multiplying the low frequency components R _low , G _low , and B _low of the color signal input from the interpolation unit 23 by a coefficient W. And output to the averaging unit 13.

The averaging unit 13 performs an averaging process on the input signals R _low · W, G _low · W, and B _low · W for each predetermined block, and generates signals R _ave , G _ave , B _ave , The result is output to the subtracting unit 25.

The subtraction section 25 and addition section 28, recovery processing section 20, (27) as shown in the expression signals _{R 1} obtained by subtracting the signal _{R ave} added signal _{R low} · W to the signal _{R low,} signal _{G low} signal _{G low} · W added signals _{G 1} minus the signal _{G ave,} the and the signal _{B low} by adding the signal _{B low} · W and outputs a signal _{B 1} obtained by subtracting the signal _{B ave} in the.

  It will be described below that the conventional problems can be solved by performing the arithmetic processing according to the present invention. For the first problem that the calculation result by the restoration processing unit 20 may be a negative value, the input will be described in one dimension as the value of equation (1). The average value of R, G, B and Y is given by equation (28).

  Since the average of Y (i, j) is given by equation (29), the coefficient W of equation (27) is given by equation (30).

  When these are applied to the equation (27), the equation (31) is obtained. At least in this example, the negative signal does not appear, and the signal deterioration is improved.

Next, a second problem that a signal deteriorates when conversion is repeatedly performed will be described. In order to solve the second problem, it is only necessary that the color conversion is performed again from the color restored equation (27) so that the value becomes equivalent to the equation (21). First, when the luminance Y ₁ is generated from the equation (27), the equation (32) is obtained.

  Here, equation (33) is established from equation (26).

  Therefore, equation (32) becomes equation (34).

  Here, since the equation (35) holds for all values from the definitions of average, subsample, and interpolation, the second item of the equation (34) is zero. Therefore, Expression (36) is established.

Next, an average is taken for R ₁ , G ₁ , and B _{1 in} equation (27). Expression (37) is established from Expression (17) and Expression (35). Similarly, equations (38) and (39) are also established.

Thus, according to the present invention, when the output signal of the color signal restoration device 2 is input to the color signal conversion device 1 and color conversion is performed again, the converted luminance signal Y ₁ is the same as the original luminance signal Y. Since the color signals R ₁ , G ₁ and B ₁ are stored with the average values of the original color signals R, G and B, deterioration due to repeated conversion can be improved.

(Second Embodiment)
Next, a second embodiment according to the present invention will be described. In the first embodiment, the interpolation unit 23 performs zero-order interpolation as shown in Equation (22). In this case, there is a possibility that a color block structure may be seen depending on the pattern. In order to prevent this, the color signal restoration apparatus of the second embodiment uses LPF and average together. Since the color signal conversion apparatus is the same as that of the first embodiment, only the color signal restoration apparatus 3 will be described. Further, the same components as those in the first embodiment are denoted by the same reference numerals, and description thereof is omitted.

  FIG. 5 is a block diagram showing the configuration of the color signal restoration apparatus 3 according to the second embodiment of the present invention. As illustrated in FIG. 5, the restoration processing unit 20 includes a coefficient generation unit 21 ′, a multiplication unit 27, an averaging unit 13, a subtraction unit 25, and an addition unit 28. The color signal restoration device 3 of the second embodiment is different from the color signal restoration device 2 of the first embodiment only in that an interpolation filter unit 29 is provided instead of the interpolation unit 23. Therefore, hereinafter, the interpolation filter unit 29 will be described.

  FIG. 6 is a block diagram illustrating a configuration of the interpolation filter unit 29. As illustrated in FIG. 6, the interpolation filter unit 29 includes an interpolation unit 23, an LPF unit 291, an averaging unit 13, a subtraction unit 25, and an addition unit 28. The averaging unit 13 and the interpolation unit 23 are the same as the averaging unit 13 and the interpolation unit 23 of the first embodiment.

The LPF unit 291 performs the calculation shown in the equation (40) on the signals R _ip , G _ip , B _ip input from the interpolation unit 23 and extracts the signals R _lpf , G _lpf , B _lpf .

  Here, P (n, m) is a constant string that determines the characteristics of the LPF. An example is shown in equation (41). N in the equation (40) is a constant, and N = 3 in the example of the equation (41).

If this LPF is used as an interpolation, the average value is not always maintained, so the problem cannot be satisfied. Therefore, in order to maintain the average value, the averaging unit 13 performs an averaging process on the signals R _lpf , G _lpf , and B _lpf input from the LPF unit 291 for each predetermined block, (R _lpf ) _ave , (G _lpf ) _ave , and (B _lpf ) _ave are generated.

The interpolation filter unit 29 adds the signals R _ipf , G _lpf , and B _lpf to the signals R _ip , G _ip , and B _ip by the subtracting unit 25 and the adding unit 28 to add signals (R _lpf ) _ave , (G _lpf ) _ave , (B _lpf ) _ave , R _ip , G _ip , B _ip low frequency component signal R _low = JP (SS (Ave (R (i, j)))), G _low = JP (SS (Ave ( G (i, j)))), B _low = JP (SS (Ave (B (i, j)))) is generated. A function that performs this processing is denoted as JP () in order to distinguish it from the zero-order interpolation function IP () of the first embodiment. The interpolation filter unit 29 generates a signal JP (U (i ₁ , j ₁ )) (i, j) based on the equation (42), where U (i ₁ , j ₁ ) is the input signal.

  The coefficient generation unit 21 ′ generates a coefficient W ′ according to the equation (43).

The restoration processing unit 20 generates the color signals R ₁ , G ₁ , B _{1 according} to the equation (44), similarly to the equation (27).

  As described above, in the second embodiment, it is possible to further improve the quality of the restored image by performing the LPF process after the 0th-order interpolation by the interpolation filter unit 29. Further, as in the first embodiment, it is possible to improve degradation due to signal clipping and conversion repetition.

(Third embodiment)
Next, a third embodiment according to the present invention will be described. In general, the luminance signal used in the video system is not accurate in terms of color engineering. In the definition of color engineering, a luminance L is directly generated from signals r, g, and b proportional to input light. In the present embodiment, a case where the average of each color component is transmitted in such a case will be described.

  FIG. 7 is a block diagram showing a configuration of a color signal conversion apparatus according to the third embodiment of the present invention. As shown in FIG. 7, the color signal conversion device 4 includes a matrix calculation unit 12 ′, an averaging unit 13, a gamma correction unit 10, a subtraction unit 25, a subsampling unit 14, and a constant multiplication unit 30. Prepare.

  The matrix calculation unit 12 ′ performs linear conversion represented by the equation (45) on the input signals r, g, and b to generate a luminance signal L.

The averaging unit 13 performs an averaging process on the luminance signal L and the signals r and b, and signals L _ave = Ave (L (i, j)), r _ave = Ave (r (i, j)), b _ave = Ave (b (i, j)) is generated and output to the gamma correction unit 10.

The gamma correction unit 10 performs gamma conversion on the luminance signal L and the signals L _ave , r _ave , and b _ave input from the averaging unit 13 to generate signals Γ (L), Γ (L _ave ), and Γ ( r _ave ), Γ (b _ave ). Here, Γ () is a function representing gamma conversion.

The sub-sampling unit 14 performs sub-sampling processing on the difference signal between the signal Γ (r _ave ) and the signal Γ (L _ave ) and the difference signal between the signal Γ (b _ave ) and the signal Γ (L _ave ).

Finally, the color signal conversion device 4 generates luminance / color difference signals Y, C ₁ and C ₂ based on the equation (46). The constant multipliers 30-1 and 30-2 perform arithmetic processing of 1/2 (1-C _B ) times and 1/2 (1-C _R ) times, respectively, as shown in the equation (46).

  This is equivalent to the transmission of equation (47) equivalently, as in the derivation from equation (19) to equation (22).

  FIG. 8 is a block diagram showing a configuration of a color signal restoration apparatus according to the third embodiment of the present invention. As illustrated in FIG. 8, the color signal restoration device 5 includes a coefficient generation unit 21 ″, a multiplication unit 27, an averaging unit 13, a subtraction unit 25, and an addition unit 28.

  The coefficient generation unit 21 ″ includes an averaging unit 13, a sub-sampling unit 14, an inverse matrix calculation unit 22, an inverse gamma correction unit 40, an interpolation unit 23, a matrix calculation unit 31, a subtraction unit 25, A division unit 26. The averaging unit 13 and the sub-sampling unit 14 perform the same processing as the averaging unit 13 and the sub-sampling unit 14 in the color signal conversion apparatus 1.

The inverse matrix calculation unit 22 performs the same processing as the inverse matrix calculation unit 22 of the conventional color signal restoration device 7, and performs SS (Ave) which is an input signal by 3 × 3 linear conversion expressed by the equation (3). (Y (i, j))), C ₁ (i ₁ , j ₁ ), C ₂ (i ₁ , j ₁ ), SS (Ave (R (i, j))), SS (Ave (G ( i, j))), SS (Ave (B (i, j))) is generated and output to the inverse gamma correction unit 40.

The inverse gamma correction unit 40 performs inverse gamma conversion on the luminance signal Y and the color signal input from the inverse matrix calculation unit 22 to obtain L (i, j), Γ ⁻¹ (D _R (i ₁ , j _{^{1), Γ -1 (D G}} (i 1, j 1), to produce a _{^{Γ -1 (D B (i 1}} , j 1). here, Γ ^-1 () is a function representing the inverse gamma conversion is there.

The interpolation unit 23 performs an interpolation process on the input signal from the inverse matrix calculation unit 22, and the low frequency component R _low = IP (Γ ⁻¹ (D _R (i ₁ , j ₁ ))), G of the color signal. _low = IP (Γ ⁻¹ (D _G (i ₁ , j ₁ ))), B _low = IP (Γ ⁻¹ (D _B (i ₁ , j ₁ ))) is generated, and the matrix calculation unit 31 and multiplication are performed. To the unit 27.

The matrix calculation unit 31 performs linear conversion on the signals R _low , G _low , and B _low input from the interpolation unit 23 to generate a low frequency component of the luminance signal L.

  The coefficient generation unit 21 ″ generates a coefficient W ″ represented by the equation (48).

The multiplier 27 generates signals R _low · W, G _low · W, B _low · W obtained by multiplying the low frequency components R _low , G _low , and B _low of the color signal input from the interpolation unit 23 by a coefficient W. And output to the averaging unit 13.

The averaging unit 13 performs an averaging process on the input signals R _low · W, G _low · W, and B _low · W for each predetermined block, and generates signals R _ave , G _ave , B _ave , The result is output to the subtracting unit 25.

The subtraction section 25 and addition section 28, a color signal restoration unit 5, (49) as shown in the expression signals _{r 1} obtained by subtracting the signal _{R ave} added signal _{R low} · W to the signal _{R low,} signal G signal _{G low} · W added signal _{g 1} minus the signal _{G ave} been, and generates a signal _{b 1} obtained by subtracting the signal _{B ave} added signal _{B low} · W to the signal _{B low} to _low.

  As described above, according to the third embodiment, similarly to the first embodiment, it is possible to improve deterioration due to signal clipping and conversion. Further, in the present embodiment, the interpolation process is performed by the interpolation unit 23, but an interpolation filter unit 29 may be provided instead of the interpolation unit 23 as in the second embodiment. In this case, the image quality of the restored image can be improved.

  It should be noted that a computer can be suitably used for causing the color signal converters 1 and 4 to function as described above, and such a computer is a program describing processing contents for realizing the functions of the color signal converters 1 and 4. Is stored in the storage unit of the computer, and this program is read and executed by the CPU of the computer.

  In addition, a computer can be suitably used to function as the color signal restoration devices 2, 3, and 5 described above, and such a computer performs processing contents for realizing the functions of the color signal restoration devices 2, 3, and 5. Is stored in the storage unit of the computer, and the program is read and executed by the CPU of the computer.

  Although the above embodiments have been described as representative examples, it will be apparent to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications and changes can be made without departing from the scope of the claims.

  Thus, the present invention is useful for applications that convert color signals into luminance / color difference signals and applications that convert luminance / color difference signals into color signals.

1, 4 color signal conversion device 2, 3, 5 color signal restoration device 10 gamma correction unit 11 conversion processing unit 12, 12 ′ matrix operation unit 13 averaging unit 14 subsampling unit 20 restoration processing unit 21, 21 ′, 21 ″ Coefficient generation unit 22 Inverse matrix operation unit 23 Interpolation unit 24, 31 Matrix operation unit 25 Subtraction unit 26 Division unit 27 Multiplication unit 28 Addition unit 29 Interpolation filter unit 30 Constant multiplication unit 40 Inverse gamma correction unit 291 LPF unit

Claims (7)

A color signal conversion device for converting a color signal into a luminance / color difference signal,
A matrix calculation unit that performs linear conversion on the input color signals r, g, and b to generate a luminance signal L;
An averaging unit that performs an averaging process for each predetermined block on the luminance signal L and the color signals r and b, and generates signals L _ave , r _ave , and b _ave ;
The luminance signal L and the signals L _ave , r _ave , and b _ave are subjected to gamma conversion to generate signals Γ (L), Γ (L _ave ), Γ (r _ave ), and Γ (b _ave ). A gamma correction unit;
A sub-sampling unit that performs a sub-sampling process on the difference signal between the signal Γ (r _ave ) and the signal Γ (L _ave ) and the difference signal between the signal Γ (b _ave ) and the signal Γ (L _ave ) When,
A color signal conversion device comprising: A color signal restoration device that converts luminance / color difference signals into color signals,
An inverse matrix calculation unit that performs linear conversion on the luminance / color difference signal and generates a color signal;
An interpolation unit that performs an interpolation process on the color signal to generate low-frequency component signals R _low , G _low , and B _low of the color signal;
A matrix calculation unit that performs linear conversion on the signals R _low , G _low , and B _low to generate a low-frequency component of a luminance signal;
A coefficient generator that generates a coefficient W indicating a ratio of a high-frequency component to a low-frequency component of the luminance signal;
A multiplier for generating signals R _low · W, G _low · W, B _low · W by multiplying the low frequency components R _low , G _low , B _low of the color signal by a coefficient W;
An averaging unit that performs an averaging process for each predetermined block on the signals R _low · W, G _low · W, and B _low · W to generate signals R _ave , G _ave , and B _ave ,
The signal the to _{R low} signal _{R low} · signals _{R 1} to W added by subtracting the signal _{R ave,} signals _G 1 obtained by subtracting the signal _{G ave} adding the signal _{G low} · W to the signal _{G low,} It means for generating a signal _{B 1} obtained by subtracting the signal _{B ave} and adding the signal _{B low} · W to the signal _{B low,}
An inverse gamma correction unit that performs inverse gamma conversion on the signals R ₁ , G ₁ , and B ₁ ;
A color signal restoration apparatus comprising: A color signal restoration device that converts luminance / color difference signals into color signals,
An inverse matrix calculation unit that performs linear conversion on the luminance / color difference signal and generates a color signal;
An inverse gamma correction unit that performs inverse gamma conversion on the luminance signal and the color signal;
An interpolation unit that performs interpolation processing on the color signal that has undergone inverse gamma conversion to generate low-frequency component signals R _low , G _low , and B _low of the color signal;
A matrix calculation unit that performs linear conversion on the signals R _low , G _low , and B _low and generates a low frequency component of the luminance signal that has been subjected to the inverse gamma conversion;
A coefficient generation unit that generates a coefficient W indicating a ratio of a high-frequency component to a low-frequency component of the luminance signal subjected to the inverse gamma conversion;
A multiplier for generating signals R _low · W, G _low · W, B _low · W by multiplying the low frequency components R _low , G _low , B _low of the color signal by a coefficient W;
An averaging unit that performs an averaging process for each predetermined block on the signals R _low · W, G _low · W, and B _low · W to generate signals R _ave , G _ave , and B _ave ,
The signal the to _{R low} signal _R signal _{r 1} which _low · W added subtracting the signal _{R ave,} signal _g 1 obtained by subtracting the signal _{G ave} adding the signal _{G low} · W to the signal _{G low,} It means for generating a signal _{b 1} which is obtained by subtracting the signal _{B ave} and adding the signal _{B low} · W to the signal _{B low,}
A color signal restoration apparatus comprising: The interpolator performs zero-order interpolation to the color signal, the signal _{R low,} G _low, and generates a _{B low,} the color signal restoring apparatus according to claim 2 or 3 . The interpolation part is
A zero-order interpolation unit that performs zero-order interpolation on the color signal to generate signals R _ip , G _ip , and B _ip ;
An LPF unit for extracting low-frequency component signals R _lpf , G _lpf , B _lpf of the signals R _ip , G _ip , B _ip ;
An averaging unit that performs an averaging process for each of the predetermined blocks on the signals R _lpf , G _lpf , and B _lpf to generate signals (R _lpf ) _ave , (G _lpf ) _ave , and (B _lpf ) _ave , Prepared,
The signals R _ipf , G _lpf , B _lpf are added to the signals R _ip , G _ip , B _ip to subtract the signals (R _lpf ) _ave , (G _lpf ) _ave , (B _lpf ) _ave , and the signal R _low , _{G low,} and generates a _{B low,} the color signal restoring apparatus according to claim 2 or 3. A program for causing a computer to function as the color signal conversion apparatus according to claim 1 . A program for causing a computer to function as the color signal restoration device according to any one of claims 2 to 5 .

Images