JPH01161993A - Coding system for color picture information - Google Patents

Abstract

PURPOSE:To save information quantity to a black character text by applying compression coding only to a lightness signal when a color difference signal is included in a prescribed range. CONSTITUTION:A lookup table 109 quantizes a means value of color difference data a<x>, b<x> into a code Cab to decide whether or not the block is achromatic based on the mean value of the color difference data a<x>, b<x> and outputs a 1-bit discrimination flag S. That is, when the color difference signal is included in a prescribed range, it is a known single color or achromatic and then it is not required to code the color difference signal, only the lightness signal is subjected to compression coding to assign lots of bits. Thus, the color picture information is coded efficiently.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an encoding method for color image information, particularly m×
The present invention relates to a color image information encoding method in which three primary color signals of n pixels are separated into a brightness signal and a color difference signal and compressed and encoded.

[Conventional technology] FIG. 4 is a block diagram of a conventional image processing device.

In the figure, 201 is a reader, such as a color page scanner or a TV camera. The reader 20
1 outputs digital image data R1G and B regarding the color separation signals R, G, and B of the image in parallel. 2o2 is L
"a"b1 conversion circuit, digital image data R9
For example, G and B are brightness signals L1 in CIE uniform color space.
and color difference signals a'' and b''. 2o3 is a line buffer for 4 lines, and image data Lll, am
, b*, image data L8°a'',b'' are stored so that block processing can be performed for each 4×4 pixel. 20
Reference numeral 4 denotes an orthogonal transformation circuit, which orthogonally transforms the brightness signal L'' for each 4×4 pixel. The orthogonal transformation is, for example, Hadamard transformation, cosine transformation, etc. This orthogonally transformed information is appropriately scalar quantized. .205 is a vector quantizer.
Then, the orthogonally transformed and scalar quantized information is vector quantized (information compressed) to generate an encoded code C with a predetermined number of bits.
Output L. On the other hand, 206°207 is an averaging circuit (Σ
), and the 0 color difference signal a'', which averages each color difference signal a'', b' in the block and outputs the average value D'', τ'', respectively.
Averaging b is based on the fact that, as a human visual characteristic, the spatial resolution of color difference signals is inferior to that of brightness signals. A lookup table 208 encodes the average value of the color difference signals a", b" and outputs a code Cab of a predetermined number of bits. 209 is a latch, which is a combination of codes CL and Cab (Ci, ”
Cab). + means logical combination. An image memory 210 stores code data (CL+cab) for one page. In this way, a huge amount of image data R, G, and B is compactly degenerately encoded and stored in the image memory 210. A decoding circuit 211 decodes the code data (C+, +Cab) into image data L II, ",b" for each pixel, and further converts and outputs the four color data Y, M, C, K. 212
is a printer with colors Y, M, C, K, and Y, M
, C, to form a color image based on the signals.

By the way, considering the bit allocation of the code CL and Cab, the code CL requires a sufficient number of bits so that the fine brightness structure within the block is not lost.
In a flat area where colors change smoothly, a sufficient number of bits is required to prevent color jumps, and the overall bit allocation is determined to balance these two.

However, in general, when the input image is a document or the like, most of the image is achromatic characters, and the number of bits is taken up by the code Cab even though there is almost no change in hue.

For this reason, the character quality (sharpness) of the reproduced image deteriorates to the extent that the number of bits of the code CL is limited.

Furthermore, since bits of the hue signal were assigned to achromatic colors, redundancy increased unnecessarily.

[Problems to be Solved by the Invention] The present invention eliminates the above-mentioned drawbacks of the prior art, and its purpose is to provide an encoding method for color image information that can efficiently encode color image information. The objective is to propose a method.

[Means for Solving the Problems] In order to achieve the above object, the color image information encoding method of the present invention separates the three primary color signals of m×n pixels into a brightness signal and a color difference signal and converts them into compression codes. The outline of the encoding system for color image information is to compress and encode only the brightness signal when the color difference signal is included within a predetermined range.

[Operation] In such a configuration, if the color difference signal is included within a predetermined range, it is a known single color or an achromatic color, so there is no need to encode the color difference signal, and only the lightness signal is compressed and encoded. allocate many bits.

[Description of Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[First Embodiment] FIG. 1 is a block diagram of an image processing apparatus according to a first embodiment. Components in the configuration shown in FIG. 4 that can be used in this embodiment are given the same reference numerals, and their explanation will be omitted. In Figure 1, 1
05 is a vector quantizer which vector quantizes (compresses information) the orthogonally transformed vector information and outputs an encoded code CLI having a first predetermined number of bits. Further, 106 is a vector quantizer, which vector quantizes orthogonally transformed vector information and outputs a second predetermined number of bits of encoded code CL2. On the other hand, 109 is a look-up table, and the average value of the color difference data a", b" is stored as a code C.
At the same time, it is determined whether the block is an achromatic color based on the average value of the color difference data a'' and b'', and 1
A bit determination flag S is output. As a method of determining whether the color is achromatic or not, each average value of the color difference data a", b" is "0".
” represents an achromatic color, so we set the threshold T in advance, and when 11” 12+1τ” 12 T, S=
1 (achromatic color), otherwise 5 = O (chromatic color)
Create a lookup table 1°9 so that 110 is a latch, and the code (CL++Cab+S
) to latch. 111 is also a latch, and the code (CL
2+S); 112 is a selector that switches the outputs of the latches 110 and 111 and outputs them. That is,
When S=1, it is determined that the color is achromatic, and the latch 111 (C
L2+S) is output to the image memory 210 as encoded data. Further, when S=O, it is determined that the color is a chromatic color, and (CL++Cab+S) of the latch 110 is outputted to the image memory 210 as encoded data.

FIGS. 2A and 2B are diagrams showing the bit allocation of the encoded code in the first embodiment.

FIG. 2(A) is a diagram showing the bit allocation of chromatic colors with S=O. If the total number of bits is a fixed length of, for example, 32 bits, then for example, nl = 21 bits, m = 10 pits, S = The bit allocation will be 1 bit.

FIG. 2(B) is a diagram showing the bit allocation for achromatic color with S=1, for example, n2=31 bits and S=1 bit. For achromatic blocks, the number of quantization bits of the brightness signal increases by as much as 10 bits, and the quality deterioration (sharpness) of black characters and the like is significantly improved.

Returning to FIG. 1, 114 is a decoding circuit, and the code (CL
l+Cab) as printer color data y, , Mt, C
Decode to I+Kl. 115 is also a decoding circuit,
An appropriate temporary hue code Cab' is given to code CL2, and these are converted into printer color data Y2, M2, C.
2, decode into 2.

116 is a selector which, according to the determination flag bit S from the image memory 210, selects color data Y+, M+, C+, and + to the printer 21 when 5=0 (chromatic color).
2, and when S#1 (achromatic color), color data Y
2. M2. C2, 2 is sent to the printer 212.

[Second Embodiment] The second embodiment is the same as the first embodiment up to the point where the flag bit S for determining whether the hue of the processing block is achromatic is output. Then, for the block determined to be S=1 (achromatic color), only the brightness signal is changed to 5=0 (
It is configured to perform the same encoding as for chromatic colors).

That is, vector quantizers 105 and 106 in FIG. 1 can be considered to perform exactly the same quantization.

FIGS. 3(A) and 3(B) are diagrams showing the bit allocation of the encoded code of the second embodiment.

Figure 3 (A) shows the case when 5 = 0 (chromatic color), and the brightness signal is quantized into an nl bit code CLI, the color difference signal is quantized into m bit codes C and lh, and each is quantized into a 1 bit flag. Add S. .

・Figure 3 (B) shows the case where S=t (achromatic color), and only the brightness signal is the same as in Figure 3 (A) -n! Bit code C
It is encoded into LI and a 1-bit flag S is added.

When the above method is adopted, the resulting code string becomes a variable length code, so it is originally necessary to construct a word-initial code, but as shown in FIG. 3(A), nu = m.

Flag bit S for closing, brightness signal, and color difference signal
If you add , the code string will be standardized, and as shown in Figure 3 (A)
It becomes possible to distinguish between (B) and (B). In this way, the number of bits in the achromatic part becomes 1/2 and the image memory 2
10 is saved, and when communicating, it becomes possible to reduce the communication time.

In the above-mentioned embodiment, since the target is black characters, it is determined whether the color is achromatic or not.
Of course, it can also be applied to red characters, blue characters, etc.

Further, as a third embodiment, if the decoding circuit 115 in FIG. 1 is configured so that S=1 in the achromatic part as in the first embodiment, color data Y2 = M2 = C2 =
It is also possible to set it to O and output only K2.

Furthermore, as a fourth embodiment, if the temporary hue codes C, b' given to the decoding circuit 115 in FIG. 1 are given as codes of chromatic colors such as red and blue, black characters can be converted to characters of other colors. It is also possible to output the

[Effects of the Invention] As described above, according to the present invention, color image data can be efficiently compressed and encoded, and the deterioration of black characters, which require high resolution, can be minimized, or black characters, which generally have a high probability of appearing, can be compressed and encoded. It is possible to reduce the amount of information in a document.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram of the image processing device of the first embodiment, FIGS. 2(A) and (B) are diagrams showing bit allocation of the encoding code of the first embodiment, FIG. 3(A), (B) is a diagram showing the bit allocation of the encoded code of the second embodiment, and FIG. 4 is a block configuration diagram of a conventional image processing device. In the figure, 206, 207 are averaging circuits (Σ).

Claims (1)

[Claims] In a color image information encoding method in which three primary color signals of m×n pixels are separated into a brightness signal and a color difference signal and compressed and encoded, when the color difference signal is included within a predetermined range, is a color image information encoding method characterized in that only the brightness signal is compressed and encoded.