JP3403912B2 - Color converter - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color conversion device for performing various color conversions on a generated or input color image signal for display output, hard copy output, transmission or storage.

[0002]

2. Description of the Related Art Conventionally, a three-dimensional table lookup method using three-dimensional interpolation has been proposed as a method for easily and rapidly performing complicated and various color conversions in the fields of color printing and color hard copy. In these methods, the input color space is divided into a plurality of unit interpolation solid groups, the output values after color conversion at the vertex positions are stored in a table memory, and when a color signal is input, the input color is The unit interpolated solid included is selected, weighted addition is performed using the output values at a plurality of vertices of the interpolated solid, and interpolation is performed. As a result, it is possible to perform color conversion while using the table memory to perform free color conversion, relatively reducing the table capacity, and ensuring continuity over the entire color space. At present, cubic interpolation that divides the color space into a plurality of cubes, and triangular prism interpolation that further divides the cube into two triangular prisms (Japanese Patent Laid-Open No. 7-99).
586), a quadrangular pyramid interpolation (US 4,334,240) for further dividing the cube into three quadrangular pyramids, and the cube is 5
Tetrahedron interpolation for dividing into six or six tetrahedrons (Japanese Patent Publication No. 58-16180), Dis using the solid body center
henoid interpolation (eg James M. Kasson, et al .:
Performing color space conversions with three-dime
nsional linear interpolation, Journal of Electroni
c Imaging / Jul 1995 / Vol.4 (3) PP.226-249) and the like, and other methods such as cubic interpolation, quadrangular pyramid interpolation, and triangular prism interpolation are combined and switched (JP-A-7-95423).
Issue gazette).

By the way, the applicant of the present application has previously disclosed, in Japanese Patent Application Laid-Open No. 7-99587, an interpolated triangular prism. This is a method in which, in the three primary color input space, a diagonal triangular prism formed by two cubes that are diagonally adjacent to each other is used as a unit interpolation solid. This oblique triangular prism interpolation method has two major advantages.

The first advantage is that when the density space of cyan C, magenta M, and yellow Y is input for hard copy color correction, the minimum value of the three variables frequently used when black generation is performed is calculated. That is, the MIN calculation can be executed without error.

The second advantage is that when color reproduction is performed, it is easy to balance the visually important achromatic color (gray). In the input gray, the three primary colors do not completely match and are approximately equal to each other. Therefore, the gray is graphically distributed near the diagonal of the input space along the diagonal. In the oblique triangular prism interpolation, the interpolation solid itself is set so as to include the vicinity of the achromatic color, so the grid points used for delicate control of the output color during gray reproduction can be set more rationally than any other interpolation method. Adjustment becomes easy.

[0006]

However, in the oblique triangular prism interpolation, in order to completely cover the three primary color input spaces,
A table memory outside the range that is wider than the input space by one interpolation section is required. For example, red, green, blue (RGB), or cyan, magenta, yellow (C
In the three primary color input color coordinate system such as MY), it is considered that each axis originally has an input of 0 to 1. In the conventional interpolation method without skewing, the input space range is roughly divided and output colors on each grid point are accumulated to create a color conversion table. Is inclined with respect to the coordinate axis, color conversion output is calculated also in one section of the negative portion of 0 or less and one division portion of 1 or more of the variables of each axis, and tabulation is required. Therefore, the color conversion function has L
When a function whose domain is defined as 0 or more, such as OG and exponentiation, is used, the operation result is necessary for creating the table even in the negative input part that is not mathematically defined originally. There was a difficult problem. Also,
There is a problem that the memory amount increases because the capacity of the color conversion table for oblique-triangular prism interpolation is larger by 2 sections for each axis than the conventional color conversion table. Further, due to the increase in the number of color conversion tables, there is a problem in that there is no table compatibility between the oblique triangular prism interpolation and other interpolation methods, and the practicality is difficult. For such a problem, a method of switching and combining cubic interpolation, quadrangular pyramid interpolation, and triangular prism interpolation (Japanese Patent Laid-Open No. 7-95423) is already known, but the method disclosed in this publication is As will be described later, there is a problem in the method of interpolating the quadrangular pyramid, and in reality, interpolation continuity cannot be ensured.

The present invention solves such a conventional problem. In the color conversion apparatus, when the conventionally proposed oblique triangular prism interpolation is used, a color conversion table is not necessary for an input outside the color space. By doing so, it is possible to improve the practicality by eliminating the trouble of creating it, reducing the amount of memory used, and making it possible to share the color conversion table memory with other interpolation methods. The purpose is to implement a secure interpolation method.

[0008]

In order to achieve the above object, the color conversion apparatus of the present invention uses a color conversion table defined only in the input space when performing oblique triangular prism interpolation in the three primary color input spaces. So that interpolation can be performed. That is, the oblique triangular prism interpolation is used only in the color space excluding the periphery of the input space, and the tetrahedral interpolation method and the quadrangular pyramid interpolation method are used in the remaining one section space. As a result, the input color space is completely filled with the interpolated solid, and at the same time, the continuity of the interpolation with the oblique triangular prism interpolation is ensured.

Therefore, according to the present invention, by eliminating the need for the color conversion table for the input outside the color space,
It is possible to improve the practicality by eliminating the trouble of creating it, reducing the amount of memory used, and making it possible to share the color conversion table memory with other interpolation methods. The method can be realized.

[0010]

[0011]

[0012] The color conversion device according to claim 1 of the present invention,
The input color space is divided without any gap by combining the interpolated solids of the oblique triangular prism group including the vicinity of the achromatic color, the tetrahedral group of the peripheral part of the input color space and the oblique pyramid group of the peripheral part of the input color space, and the vertex of the interpolated solid A color conversion table memory that stores output values after color conversion at a position, an interpolation solid selection unit that selects the type of the interpolation solid from an input color signal, and an output at the vertex of the interpolation solid that includes the input color signal. A value, a weighting factor determined by the area in which the interpolated solid is arranged, and an interpolation unit that performs an interpolation operation using an offset value determined by the area in which the interpolated solid is arranged. While sharing a plane, the inside of the input color space is divided without gaps, the input color signal is subjected to arbitrary color conversion and output, and when an oblique triangular prism is used as an interpolated solid.
, The interpolated volume does not exist outside the input color space.
To fill the space, for inputs outside the input color space
The color conversion table is no longer needed, and other interpolation methods and color conversion
Table can be shared, and further, it has an effect of ensuring continuity of interpolation.

A color conversion device according to claim 2 of the present invention is
2. The triangle according to claim 1 , wherein an upper bottom surface and a lower bottom surface of the oblique triangular prism are parallel to a surface of the three coordinate axes forming the three primary color input spaces in which the variable of the first axis is constant, and form the upper and lower bottom surfaces. The two sides of each are parallel to the second and third axes, and the main axis of the oblique triangular prism is parallel to the diagonal axis expressing the achromatic color of the three primary color space, and as a result, the oblique triangular prism is the variable of the first axis. It has three planes as a boundary surface: a surface where the variables of the second axis are equal, a surface where the variables of the second axis and the variable of the third axis are equal, and a surface where the variables of the third axis and the variables of the first axis are equal. .

A color conversion device according to claim 3 of the present invention is
In the structure according to claim 1, the bottom surface composed of four points out of the five points forming the diagonal quadrangular pyramid matches the parallelogram that is the side surface of the triangular prism, and the remaining one vertex is the outermost contour of the color space. It is a point.

A color conversion apparatus according to a fourth aspect of the present invention is
In the construction of claim 1, or tetrahedron has been a 6 divided form a cube obtained by dividing the color space to share the diagonal of the cube, the tetrahedron aspect is the outermost side of the color space , Or touching another tetrahedron or pyramid.

With the above structure, the conventional oblique triangular prism interpolation is used mainly for most areas of the color space centering on the achromatic color, thereby eliminating the problem of jumping out in the periphery of the color space, which is a problem in the prior art. This has the effect of eliminating the need for a color conversion table for the input.

In addition, the tangent planes of the respective interpolated solids have a special arrangement in which they share common ridgelines, and also have the effect of maintaining continuity in the ridgelines sharing the interpolation method and the interpolation results on the plane. . Further, it has an effect of maintaining the interpolation performance of MIN calculation and the ease of color adjustment of achromatic gray, which are the advantages of the oblique triangular prism interpolation.

Embodiments of the present invention will be described in detail below with reference to the drawings. (Embodiment) FIG. 1 shows the configuration of a color conversion apparatus according to an embodiment of the present invention. In FIG. 1, 101 is an upper color signal of the input color signal, and 102 is a lower color signal of the input color signal. Reference numeral 103 is a tetrahedral area number determination unit, 104 is an interpolation solid selection unit, 105 is a tetrahedral area number Area, 106 is a color conversion table memory, 107 is a selector, 108 is a tetrahedral interpolation unit, 109 is a triangular prism interpolation unit, 110 Is an oblique quadrangular pyramid interpolation unit. Reference numeral 111 is a color conversion output value at each vertex of the tetrahedron, 112 is a color conversion output value at each vertex of the oblique triangular prism, and 113 is a color conversion output value at each vertex of the oblique quadrangular pyramid. 114 is a tetrahedral interpolation output, 115 is an oblique triangular prism interpolation output, and 116 is an oblique quadrangular pyramid interpolation output. Reference numeral 117 is interpolation stereoscopic selection information.

In this embodiment, the input color space is a space composed of three axes of 8 bits (RGB) for each axis, and the higher order signals (RH, GH, BH) which are the higher order 3 bits of each color.
The color conversion table is looked up at 101, and interpolation is performed by the low-order signal (RL, GL, BL) 102 which is the low-order 5 bits of each color. In the color conversion table memory 106, the input space is 9x which is a vertex obtained by dividing the input space into 8x8x8 cubes which are accessed by (RH, GH, BH).
Color conversion outputs at 9 × 9 = 729 grid points are stored. The color conversion output may be any function other than a function having discontinuity in the output. For example, since it is used for approximation of the color masking method in the main application of the present invention, a three-dimensional LUT constructed by using the color masking method using the minimum value function (MIN operation) disclosed in Japanese Patent Laid-Open No. 7-95423. Can be used as is as a color conversion table.

The high-order signal is one input to the interpolation solid selection unit 104. On the other hand, the lower signal is input to the tetrahedral area number determination unit 103, and it is determined which one of the six tetrahedrons the input color belongs to is further divided.

FIG. 2 shows six tetrahedron areas, Area = 0 to Area = 5, in the tetrahedron area determination, which are lower order signals (RL, GL, BL).
Is determined according to the flow chart shown in FIG. All of these tetrahedra are cubes of the same volume sharing the diagonal of the cube, and Area = 0 is set on the diagonal for convenience. The area numbers 0 to 5 as the result are the interpolation solid selection unit 104.
Entered in. The interpolated solid selection unit 104 uses the higher-order signal 10
An interpolation solid is selected from 1 and the tetrahedral area number 105. There are three types of interpolated solids: oblique triangular prism, tetrahedron, and oblique quadrangular pyramid.

FIG. 4 expresses the concept of dividing an RGB solid, which is an input color space, into these interpolated solids, that is, an oblique triangular prism, a tetrahedron, and an oblique quadrangular pyramid without any space. Here, the reason why such an interpolation solid is used will be described.

FIG. 25 is a trihedral view showing a color conversion table when the interpolation using only the oblique triangular prism in FIG. 4 is used. Here, for the sake of convenience, the number of divisions is divided into 4x4x4 axes. The three primary colors of R, G, B are 0 to 1.
The RGB space when changing all combinations up to (maximum value) is a cube formed by white circles 250 in FIG. 25, and the white circles are grid points. This RGB cube cannot be completely filled with the oblique triangular prism, and in order to cover the whole with the interpolation area, an extra point outside the RGB spatial solid range shown by the black dot 251 in FIG. 25 is required. However, as mentioned above, creating out-of-range tables is often difficult. Normally, RGB inputs are considered to represent 0 to 1, respectively. Therefore, for example, in the formula, the term of Log (R) or Pow
When a power calculation such as (R, 1/3) is included, the calculation result of the portion exceeding the original function domain must be stored in the table, which deteriorates the calculation accuracy. In consideration of this point, conventionally, when the oblique triangular prism interpolation is used, grid points are created by various extrapolation methods, but this also causes deterioration of calculation accuracy. In addition, there is a big problem that the capacity of the color conversion table increases with the increase of the grid points. When the number of divided cubes of each axis is N, the number of grid points outside the range is N = 2 (2N + 1) N (Equation 1) The number of grid points inside the range = (N + 1) (N + 1) (N + 1) .. (Equation 2), and when the number of divisions N = 4 in FIG. 25, the number of out-of-range lattice points is 72 and the number of in-range lattice points is 125, which is essentially only the color conversion table within the range. The capacity is increased by 57.6% as compared with the cubic interpolation method. This increase in table capacity is 37.3% when N = 8 and 21.4% when N = 16, and is still large even if the number of divisions increases.

For these reasons, there is a demand for an oblique triangular prism interpolation method that does not require use of grid points outside the range, and the present invention solves this as follows. The advantage of the oblique triangle is mainly inside the color space, such as adjustment in the vicinity of the achromatic color, and the disadvantage of popping out of the range of the oblique triangle appears around the color space. For this reason, the use is restricted so that the oblique triangular prism interpolation is used only in the range that does not overflow the color space.

FIG. 5 shows an oblique triangular prism interpolation portion in the present invention. In this way, since the oblique triangular prism interpolation is limited to the inside of the color space, a portion not included in the interpolation section remains around the color space. Therefore, this space is filled with a method using another interpolation solid. At this time, it is important that condition 1) fill all the RGB spaces without gaps, and condition 2) do not cause discontinuity in the interpolation result at the connecting portion of different interpolated solids.

FIG. 4 shows a part of how the RGB space is divided into an oblique triangular prism, an oblique quadrangular pyramid and a tetrahedron based on the above first condition, and not only the space is filled but also the continuous space is continuously filled. The feature is that planes and edges are always shared for the sake of sex, and the details of the dividing method will be described below. This is determined by the higher order signals (RH, GH, BH) and the tetrahedral area number.

First, regarding the tetrahedral interpolation use portion filled with the tetrahedron, when viewed from the positive direction of G as shown in FIG. 6, each G constant division plane surrounds the periphery of the RB plane. It can be seen that there are corners corresponding to only one region, corners corresponding to two regions, and corners corresponding to three regions in the cube. For example, on a side such as (RH, BH) = (1,0), only the tetrahedron having the tetrahedron region number = 0 corresponds to the tetrahedral interpolation part, but (RH, BH) = (0,0) In the corner cube,
Two tetrahedrons with tetrahedron region numbers = 0 and 5 correspond to the tetrahedron interpolation part, and in the cube of the corner of (RH, BH) = (3,0), tetrahedron region numbers = 0,1, It can be seen that the three tetrahedra of 2 correspond to the tetrahedral interpolation portion.

Next, regarding the diagonal quadrangular pyramid interpolation portion filled with the diagonal quadrangular pyramids, when viewed from the positive direction of G as shown in FIG. It is distributed so that it surrounds. The type of diagonal quadrangular pyramid is (1-5)
Type, (3-1)) type, (2-4) type, (4-0) type 4
They are classified into types, and in FIG. 7, type (1-5) and type (3-1)
A part of the mold, a part of the (2-4) mold and a part of the (4-0) mold are illustrated as overlapping from the top. These meanings are shown in FIGS. 8 and 9. The quadrangular pyramid (1-5) type of FIG. 8 is a tetrahedron with Area = 1 in a cube with smaller R in two cubes adjoining in the R direction and a tetrahedron with Area = 5 in a cube with larger R. It means an oblique quadrangular pyramid formed by combining with the body. Hereinafter, these tetrahedra are referred to as 150 and 151, respectively. Similarly, a tetragonal pyramid (2-4) type is a tetrahedron with Area = 2 in a cube with a smaller R in two cubes adjoining in the R direction and a tetrahedron with Area = 4 in a cube with a larger R. It means an oblique quadrangular pyramid made by combining and. Hereinafter, these tetrahedra are referred to as 240 and 241 respectively. Similarly, the tetragonal pyramid (3-1) type is a tetrahedron with Area = 3 in a cube with a smaller B in two cubes adjacent to each other in the B direction and a tetrahedron with an Area = 1 in a cube with a larger B. It means an oblique quadrangular pyramid made by combining and. Hereinafter, these tetrahedra are referred to as 310 and 311 respectively. Similarly, a tetragonal pyramid (4-0) type is a tetrahedron of Area = 0 in a cube with a smaller B in the two cubes adjacent to each other in the B direction and a tetrahedron of Area = 0 in a cube with a larger B. It means an oblique quadrangular pyramid made by combining and. Hereinafter, these tetrahedra are referred to as 400 and 401, respectively.

The parts other than the tetrahedron and the oblique quadrangular pyramid described above become the parts of the oblique triangular prism. Regarding the oblique triangular prism, types of the oblique triangular prism will be described with reference to FIGS. 10 and 11. FIG. 10 shows four kinds of origin positions (0) to (3) which form four kinds of oblique triangular prisms, and FIG. 11 shows the type of oblique triangular prisms. The origin position number (0) is an oblique triangular prism when the point a, which is the origin, is the same as the cube determined from the upper signal. Origin position number (1) is -B when (0)
, The origin position number (2) is (0) in the -R direction, and the origin position number (3) is (0) in the -R direction and the -B direction. Thus, in addition to the origin position of the oblique triangular prism, as shown in FIG. 11, the type 0 of the oblique triangular prism abc-efg on the front side at each position is shown.
And is classified into type 1 of adc-ehg. As shown in (Table 1), the types PR of the oblique triangular column formed by combining the origin position number and the type are (3-1), (2-0), (0
There are 6 types of (0), (0-1), (1-1), and (3-0). Figure 12 shows that when one cube is specified,
In this cube, the above-described six types of portions belonging to the oblique triangular prism are expressed, and each is a tetrahedron.

[Table 1]

The above-described selection of the interpolated solid is executed according to the sectional view as shown in FIG. In FIG. 13, a triangle 130 indicates a tetrahedral interpolation portion, and a parallelogram 131
Indicates an oblique quadrangular pyramid interpolation portion. For example, (RH, BH) =
For a cube that is (0,0), Area = 0 and Area
= 5 tetrahedron is tetrahedral interpolation, and Area = 1 is (R
H, BH) = (0, 1) is combined with Area = 5 to form an oblique pyramid interpolation part (1-5), and similarly, a tetrahedron with Area = 4 is (RH, BH) = (0, 1). Area = 0 to form an oblique quadrangular pyramid interpolation unit (4-0).
The interpolated solid selection unit 104 selects an interpolated solid according to the flowcharts of FIGS. 14 to 20 in FIG. 13 on the premise of the above graphic processing. These flow charts select three types of T (tetrahedral interpolation), PY (oblique quadrangular pyramid interpolation), and PR (oblique triangular prism interpolation) and specify the respective types according to the Area number and the numerical values of the upper signals RH and BH. It is a thing. For example, if Area = 1, the process jumps to FIG. 16, the tetrahedral interpolation T (1) (indicating Area = 1), and the triangular prism P.
The R (2-0) type, the oblique quadrangular pyramid PY (150) type, or the (311) type is selected. Thus, T,
Interpolation solid selection information 117, which is a selection information of interpolation solids such as PR and PY and type information, is passed through the selector 107 to a tetrahedral interpolation unit 108, an oblique triangular prism interpolation unit 109,
An oblique quadrangular pyramid interpolation unit 110 and a color conversion table memory 106
Sent to.

As described above, after the interpolation solid selection unit 104 obtains the selection information of the interpolation solid, the selector 107 selects one from the three interpolation units, and the corresponding interpolation units 108,
The actual interpolation calculation is performed at only one of 109 and 110. Two irrelevant interpolated outputs are either ignored or no interpolation operation is performed originally.

First, the calculation of the tetrahedral interpolation unit 108 will be described with reference to FIG. The tetrahedral interpolation unit 108 receives the color conversion table output value 111 consisting of four points, the lower-order signal 102, and the interpolation solid selection information 118 as inputs, and outputs the final interpolation result 114. When the tetrahedral area number Area to which the input point P belongs is determined, the output value at P is the output value (a) at the four vertices a, b, c, d of the tetrahedron,
(B), (c), (d) and weighting factor MAX, MII
It can be calculated as follows using D and MIN. (P) = (1-MAX). (A) + (MAX-MID). (B) + (MID-MIN). (C) + MIN. (D) (Equation 3) Weighting factor MAX Means the maximum value among RL, GL, and BL, MID means the intermediate value, and MIN means the minimum value, which can be determined from the lower signal 102 and the tetrahedral area number Area 105. In the example shown in FIG. 21, Area = 2,
As can be seen from FIG. 3, MAX = RL MID = BL
MIN = GL. In the configuration of 4 vertices, each point is a grid point address of the color conversion table memory as shown in Expression 4 and the offset value (DR, DG, D) is used as the upper signal (RH, GH, BH).
B) is added, and this offset value is determined as shown in (Table 2). (RH + DR, GH + DG, BH + DB) (Equation 4)

[Table 2] Therefore, the upper signal 10 is stored in the color conversion table memory.
T (0) to T (5) determined by 1 and 118 are input, output values 111 at four points a, b, c, d are output using (Table 2) and Equation 2, and then tetrahedron The interpolation unit 108 executes the interpolation calculation of Expression 3.

Next, the calculation of the oblique triangular prism interpolating unit 109 is shown in FIG.
Will be explained. The output values of the color conversion table from points a to h are represented by (a) to (h). A straight line drawn from the input point P in parallel with the line segment ae and the lower bottom surface abc
If the intersection points with d and the upper bottom surface efgh are m and n, respectively, the output (m) and (n) at the points m and n are as follows: the oblique triangular prism has a type number = 0, that is, PR = (* -0), * means At any time
Using the weighting factors wr, wg, and wb, (m) = (a) + wr · {(b) − (a)} + wb · {(c) − (a)} (n) = (c) + wr · { (F) − (c)} + wb · {(g) − (f)} (P) = (m) + wg · {(n) − (m)} (Equation 5) Type = 1, that is, When PR = (* − 1) and * are arbitrary, (m) = (a) + wr · {(c) − (d)} + wb · {(d) − (a)} (n) = (c) + Wr * {(g)-(h)} + wb * {(h)-(e)} (P) = (m) + wg * {(n)-(m)} ... (Equation 6) It Here, the weighting factors wr, wg, and wb are calculated as shown in (Table 3). Also, the grid point a
From h to (h), the offsets of the respective vertex positions of (DR, DG, DB) in (Table 4) are added as in Equation 4 to obtain 6 points, which are output values 112 at 6 points.
09 is input. The asterisks in Table 4 indicate two points that are not used due to different types, such as the point d when type = 0.

[Table 3]

[Table 4]

Next, the calculation of the oblique quadrangular pyramid interpolation unit 110 will be described in detail. FIG. 23 shows a general pyramid.
The method of interpolation is shown. The bases a, b, c, d,
To obtain an interpolated value at the input point P with respect to a quadrangular pyramid formed by four points and five points of f which are vertices, a straight line fP and a base ab are obtained.
It passes through the auxiliary points Q and P, which are the intersections with cd, and the base abc
A ′, b ′, c ′, d ′, which are the intersections of the plane parallel to d and the respective ridge lines fa, fb, fc, fd, are used. Hereinafter, the abbreviated method in which the output value at the point x is represented by (x) is used. Line segment fQ
The ratio k between the line segment and the line segment fP is k = fP / fQ (Equation 7) The values at four points on the ridge line are (a ′) = (f) + k · {(a) − (f)} (B ′) = (f) + k · {(b) − (f)} (c ′) = (f) + k · {(c) − (f)} (d ′) = (f) + k · {( d)-(f)} (Equation 8) On the plane a'b'c'd ', draw a parallel line from P to the side a'd', and set the intersection with the side a'b 'to the point R', also P
A line segment parallel to the side a'b ', and the intersection point with the side a'd' is the point S ', l = a'R' / a'b 'm = a'S' / a'd ' ... (Equation 9) The ratio is calculated. This ratio l. m can also be obtained on the plane abcd. In that case, the straight line fR ′ and the side ab
Using the intersection point R with and the intersection point S between the straight line S ′ and the side ad, l = aR / ab m = aS / ad (Equation 10). If the above values are used, the interpolation value (P) at the point P
Is (P) = (1-1). (1-m). (A ') + l. (1-m). (B') + lm. (C ') + (1-l). m · (d ′) ... (Equation 11) As a feature, on each ridge line, it results in linear interpolation of the two end points of the ridge line, and side surface triangle interpolation and plane quadrilateral interpolation are other interpolation solids such as tetrahedral interpolation side surface and oblique triangular prism side surface. Since it matches the interpolation of, continuity is maintained. Even in the case of using the same quadrangular pyramid interpolation, when the formula (39) of Japanese Patent Laid-Open No. 7-85423 is used, the continuity cannot be maintained, and even if the space can be filled without a gap, the interpolation result is There is a drawback that jumps occur.

FIG. 24 shows a case where the above general quadrangular pyramid interpolation is applied to the oblique quadrangular pyramid (1-5) type in the present invention. The meanings of points a, b, c, d, f and points P and Q are shown in FIG.
3 and the interpolation formula is the same as formula 11. Here, the method of obtaining the three weighting factors k, l, and m is calculated in each case as shown in (Table 5).

[Table 5] However, m and l cannot be calculated at the vertices on the oblique quadrangular pyramid where k = 0. Therefore, interpolation cannot be performed at the vertex position, and the color conversion table itself is output. Also,
The composition points for each type are in (Table 6) (DR, D
G, DB) offsets at the respective vertex positions are added as in Expression 4, and five points are selected.

[Table 6] Eventually, five points a, b, c, d, and f are selected from (Table 6) to become the grid point output 113, and k from these (Table 5) is changed to y in Expression 8 (a ′), (b ′). , (C '),
(D ') is obtained, and then the interpolated output value is also obtained from l and m of (Table 5) and Equation 11.

The capacity of the color conversion table memory and the number of bits of the color signal line described in the embodiment of the present invention are completely arbitrary, and other values may be used.

[0037]

As described above, according to the present invention, the conventional diagonal triangular prism interpolation is mainly used for most of the area of the color space centering on the achromatic color, so that the color space surrounding the conventional color space has been a problem. The color conversion table is not necessary for input outside the color space, and the effort of creating it can be saved, because the color conversion table outside the input space is eliminated. Can be reduced and the color conversion table memory can be shared with other interpolation methods, and the practicality thereof can be significantly improved.

Further, the tangent planes of the interpolated solids are specially arranged so as to share a common ridge line, and the interpolation method is carried out so as to maintain continuity in the ridge line and the interpolation result on the plane. be able to. Further, the interpolation performance of the MIN calculation and the ease of color adjustment of achromatic gray, which are the advantages of the oblique triangular prism interpolation, can be maintained as they are. Therefore, when creating a color conversion table, problems with points outside the input range do not occur, and color conversion is performed with the advantage of oblique triangular prism interpolation while maintaining compatibility with the conventional interpolation method and color conversion table. be able to.

[Brief description of drawings]

FIG. 1 is a block diagram of a color conversion device according to an embodiment of the present invention.

FIG. 2 is a diagram showing division into tetrahedral regions.

[Fig. 3] Flow chart of tetrahedral region determination

FIG. 4 is a conceptual diagram of dividing a color space into a tetrahedron, an oblique quadrangular pyramid, and an oblique triangular prism.

FIG. 5: Region in color space using oblique triangular prism interpolation

FIG. 6 region in color space using tetrahedral interpolation

FIG. 7: Region in color space using oblique quadrangular pyramid interpolation

FIG. 8 Classification of oblique quadrangular pyramids (1-5) (2-4)

FIG. 9 Classification of oblique quadrangular pyramids (3-1) (4-0)

[Figure 10] Classification by the origin position of the oblique triangular prism

FIG. 11 Classification of oblique triangular prisms PR (0) -PR (3) Relationship with tetrahedral regions

[Fig. 12] Classification types of oblique prisms 0 and 1

[Fig. 13] Selection diagram of interpolation solid

FIG. 14 is a flowchart for selecting an interpolation solid part 1

FIG. 15 is a flowchart for selecting an interpolation solid part 2

FIG. 16 is a flowchart for selecting an interpolation solid part 3

FIG. 17 is a flowchart for selecting an interpolation solid part 4

FIG. 18 is a flowchart for selecting an interpolation solid part 5

FIG. 19 is a flowchart for selecting an interpolation solid part 6

FIG. 20 is a flowchart for selecting an interpolation solid, part 7

FIG. 21 is a diagram of a tetrahedral interpolation method.

FIG. 22 is a diagram of an oblique triangular prism interpolation method.

FIG. 23 is a diagram of a pyramid interpolation method.

FIG. 24 is a diagram of an oblique quadrangular pyramid interpolation method.

FIG. 25 is a diagram showing a region outside the color space in conventional diagonal triangular prism interpolation.

[Explanation of symbols]

101 Upper color signal of input color signal 102 Lower-level signal of input color signal 103 Tetrahedral area number determination unit 104 Interpolation solid selection section 105 tetrahedron area number Area 106 color conversion table memory 107 selector 108 Tetrahedral interpolation unit 109 Oblique triangular prism interpolator 110 oblique quadrangular pyramid interpolation unit 111 Output value of color conversion at each vertex of tetrahedron 112 Color conversion output value at each vertex of the oblique triangle 113 Color conversion output value at each vertex of the diagonal quadrangular pyramid 114 tetrahedral interpolation output 115 Oblique triangular prism interpolation output 116 Oblique quadrangular pyramid interpolation output 117 Interpolation solid selection information

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI H04N 9/64 H04N 1/46 Z (56) Reference JP-A-8-223432 (JP, A) JP-A-7-99587 ( JP, A) JP 5-328114 (JP, A) JP 7-131668 (JP, A) JP 8-98046 (JP, A) JP 9-69961 (JP, A) JP Flat 9-307777 (JP, A) (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04N 1/60 B41J 2/525 G06T 1/00 510 G09G 5/06 H04N 9/64

Claims (4)

(57) [Claims] 1. An input color space is divided without a gap by a combination of an interpolated solid of an oblique triangular prism group including an achromatic color neighborhood, a tetrahedron group in an input color space peripheral portion and an oblique tetragonal pyramid group in an input color space peripheral portion, A color conversion table memory that stores output values after color conversion at the vertex positions of the interpolation solid, an interpolation solid selection unit that selects the type of the interpolation solid from an input color signal, and the interpolation solid that includes the input color signal. Of the output values at the vertices, the weighting factor determined by the area where the interpolated solid is arranged, and the offset value determined by the offset value determined by the area where the interpolated solid is arranged. The color conversion device that divides the inside of the input color space without any gaps while sharing the ridge line and all planes, performs arbitrary color conversion on the input color signal, and outputs the color signal. 2. The upper bottom surface and the lower bottom surface of the oblique triangular prism are parallel to the surface of the three coordinate axes forming the three primary color input space where the variable of the first axis is constant, and the upper and lower bottom surfaces of the triangle form the upper and lower bottom surfaces. Each side is parallel to the second axis and the third axis, and the main axis of the oblique triangular prism is parallel to the diagonal axis representing the achromatic color of the three primary color space, and as a result, the oblique triangular prism has the variable of the first axis and the second axis.
A boundary surface having three planes, that is, a surface where the variables of the axis are equal, a surface where the variables of the second axis and the variable of the third axis are equal, and a surface where the variables of the third axis and the variable of the first axis are equal. 1. The color conversion device described in 1 . 3. A bottom surface composed of four points out of five points forming an oblique quadrangular pyramid matches a parallelogram that is a side surface of an oblique triangular prism, and the remaining one vertex is one outermost point of the color space. The color conversion device according to claim 1 . 4. A tetrahedron has been a 6 divided form a cube obtained by dividing the color space to share the diagonal of the <br/> cube, the tetrahedral sides in the outermost side of the color space The color conversion device according to claim 1 , wherein the color conversion device is present, or is in contact with another tetrahedron or an oblique tetragon.