JPH11262004A - Device and method for processing image and storage medium thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable recognizing the outline of an image from a part of encoding data at an early stage, to eliminate redundant properties and to improve compression efficiency by distributing data converted into variable length codes to bit planes, based on the frequency distribution of predicted coefficients and hierarchically outputting it. SOLUTION: A variable length encoding part 102 refers to the code list of a code list memory 105 and successively converts input image data into the variable length codes. When the range of pixel data is identified and the variable length codes are assigned to the highest-order bit of respective kinds of pixel data in the code list so as to output encoding data at every bit plane, a density area is restricted efficiently, highness and lowness is recognized with respect to the density with the highest frequency degree in each pixel and, then the outline of the image is recognized. A bit plane forward scanning part 103 stores encoding data for the portion of two screens in a buffer 104, successively executes reading from the higher-order bit plane, in order after that a predicted blank area is skipped and encoding data by bit plane unit is outputted from a code output part 106.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and method, and a storage medium storing the method.

[0002]

2. Description of the Related Art An image, especially a multi-valued image, contains a great deal of information, and there is a problem that the amount of data becomes enormous when storing and transmitting the image. For this reason, when storing and transmitting images, remove the redundancy of images,
Alternatively, high-efficiency coding is used in which the amount of data is reduced by changing the content of the image to such an extent that the deterioration of the image quality is difficult to visually recognize.

[0003] However, even if the data amount can be reduced to some extent by high-efficiency encoding, it may take time to transmit or read the encoded data. In such a case, the side that receives the transmitted encoded data can recognize the outline of the image at the initial stage of data reception, and further receives the subsequent encoded data to gradually improve the image quality. It is preferable to use hierarchical coding that can be recognized as

Conventionally, as a general hierarchical coding, image data in which each pixel is represented by multi-values is converted into a plurality of bit planes, and these bit planes are converted from an upper bit plane to a lower bit plane. A method of transmitting data in order is performed.

For example, in JPEG recommended by the ISO and ITU-T as an international standard encoding method for still images, several encoding methods are used according to the contents of an image to be encoded and the intended use of encoded data. Are defined, and SS (Spectrum Selection) and SA (Successive Approval) for realizing hierarchical coding in the extended DCT process.
ximation) is defined.

For details on JPEG, see Recommendation ITU
-T Recommendation T.81 | ISO / IEC 109
Since it is described in 18-1 etc., it is omitted here,
In Successive Approximation, a discrete cosine transform (DCT) is performed for each block of an image, all of the obtained frequency components are quantized into n-bit coefficients, and the obtained quantized coefficients are divided into n layers (n to 1). ), And the data is transmitted in order from the upper (layer n) bit plane to the lower (layer 1) bit plane.

[0007]

However, conventionally, in a bit plane encoding method in which multi-valued image data is converted into bit planes of a predetermined number of layers and is output hierarchically for each bit plane, the bit planes are still not used. There is a problem in that the redundancy includes redundancy.

Further, in the conventional hierarchical coding method, when the receiving side receives only the upper bit plane, there is a problem that the outline of the coded multi-valued image may be difficult to understand at an early stage.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and enables an outline of an image to be recognized efficiently from a part of encoded data at an early stage. It aims to provide technology.

[0010]

According to the image processing apparatus of the present invention, a plurality of coefficients representing an image (corresponding to a pixel value or a quantized value in the present embodiment).
(Also corresponding to the image input unit 101 and the coefficient quantization units 604, 1304, and 1604) and a plurality of coefficients generated by the generation unit are calculated based on the frequency distribution of the coefficients predicted in advance. Variable-length coding means (variable-length coding unit 102, Go
lomb encoding units 605, 1305, and 1605);
Each bit of the variable length coded data corresponding to each coefficient obtained by the variable length coding of the variable length coding means (also,
For example, a code of 3 bits to several bits in FIG. 3) is distributed to a plurality of bit planes by corresponding to the order of each bit (similarly, for example, to the distribution of FIG. 4), and the plurality of bit planes are hierarchically arranged. It is characterized by having a hierarchical output means for sequentially outputting (similarly, for example, equivalent to the hierarchical output in FIG. 5).

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) Hereinafter, a representative embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows an image processing apparatus for carrying out the first embodiment of the present invention.

Referring to FIG. 1, reference numeral 101 denotes an image input unit;
Is a variable length coding unit, 103 is a bit plane forward scanning unit,
104 is a buffer, 105 is a code table memory, and 106 is a code output unit.

In the present embodiment, description will be made on the assumption that monochrome image data representing each pixel by 4 bits is encoded. However, the present invention is not limited to this, and is applicable to the case of encoding a monochrome image represented by 8 bits for each pixel or a color multi-valued image representing each color component (RGB / Lab / YCrCb) in each pixel by 8 bits. It is also possible. Also, the present invention can be applied to a case where multi-valued information representing the state of each pixel constituting an image is encoded, for example, a case where a multi-valued index value representing the color of each pixel is encoded. When applied to these, each type of multi-value information may be encoded as monochrome image data described later.

Hereinafter, the operation of each section in the present embodiment will be described in detail.

First, image data (pixel data) representing an image to be encoded is continuously input from the image input unit 101 in raster-scan order. This image input unit 101
For example, an imaging device such as a scanner or a digital camera, an imaging device such as a CCD, an interface of a network line, or the like is used. The image input unit 101 includes a RAM, a ROM, a hard disk, a CD-R
A recording medium such as OM may be used.

FIG. 2 shows a frequency distribution of pixel data generated from the image input unit 101.

In this embodiment, as shown in FIG. 2, a plurality of pieces of pixel data to be encoded have a high frequency of occurrence of small-valued pixel data and a high frequency of occurrence of large-valued pixel data. Explanation is made as low.

Such a bias in the frequency distribution is detected by the image input unit 10.
1 and the characteristics of the image to be encoded. In particular, when the image input unit 101 is a CCD, the frequency distribution tends to be biased unless gamma correction is performed. Although not shown in the present embodiment,
The case where a bias in the frequency of occurrence as shown in FIG. 2 is generated by intentionally performing preprocessing or the like while being input from the image input unit to the variable length coding unit is also included in the scope of the present invention.

The variable-length coding unit 102 includes an image input unit 101
Is variable-length coded with reference to the code table stored in the code table memory 105.

FIG. 3 shows an example of a code table stored in the code table memory 105. It is assumed that the code table is stored in the code table memory 102 before variable-length coding. The code table stored in the code table memory 102 considers the occurrence frequency distribution of pixel data representing a certain sample image as shown in FIG. 2 as a general distribution, and is generated based on this distribution. It is a thing. The length of the code shown in FIG. 2 is basically such that a short code is assigned to pixel data (pixel value) having a high frequency of occurrence. The present invention is not limited to the case where one code table is used, but also includes the case where a plurality of code tables are selectively used. In this case, the content of the image to be encoded (the frequency of occurrence of each pixel data) is actually identified, and an optimal one is selected from a plurality of code tables according to the identification result.

In the code table of FIG. 3 in the present embodiment,
Considering that transmission is performed for each bit plane in the subsequent stage, for example, the MSB (most significant bit) of each pixel data is 0
The variable length code is assigned so that it is possible to identify whether the pixel data is in the range of 0 to 2 or in the range of 3 or more depending on whether the pixel data is 1. That is, it is determined that candidate values of decoded pixel values corresponding to the next lower bit are continuous when the variable length code is recognized in order from the upper bit. This is realized by constructing a code tree by limiting two adjacent two in the process of repeatedly integrating two infrequently occurring algorithms in an algorithm for forming a Huffman code to create a code tree. be able to.

By performing the above-described variable length code allocation, when encoded data is output for each bit plane, which will be described later, an efficient density range based on the density range of each pixel is hierarchically determined. Limitations can be made.
That is, even if the receiving side of the encoded data receives only the first one bit (MSB) of each pixel as a bit plane to be described later, whether the value is higher than the density at which the occurrence frequency is highest in each pixel. Alternatively, it can be recognized at an early stage whether the value is low. This gives a very good overview of the image. Similarly, if a subsequent high-order bit plane is also received, the density can be more efficiently limited for each pixel. On the other hand, when multi-valued pixel values are hierarchically output for each bit plane as in the prior art, when only the first bit (MSB) is received, the density of each pixel is higher than the intermediate value. I can only tell if it is low or low. Therefore, when encoding an image in which the density of the entire image is concentrated in a low-density region or a high-density region, it becomes difficult for the receiving side to know the outline of the image.

In the variable length coding unit 102, if the input pixel data is "0", the output code is "000", if the pixel value is "1", "001", and if the pixel value is "2", the output code is "001". 0
The input pixel data is sequentially coded as in “1”.

The bit-plane sequential scanning unit 103 temporarily stores the variable-length encoded data output from the variable-length encoding unit 102 in the buffer 104. Then, the most significant bit (MSB) of the coded data of variable length is stored as binary data in the first bit plane, and the next higher bit is stored as binary data in the second bit plane. The position of each bit plane stored as binary data is controlled so as to correspond to the position of each pixel of the original image to be encoded. Similarly, each bit forming the variable-length coded data includes a third bit plane, a fourth bit
Are stored in the buffer 104 as binary data in the order of the bit planes.

As will be described later, since the coded data is variable-length coded, up to which bit plane the bits constituting the coded data are stored differs for each pixel.

For example, when the variable-length encoded data output from the variable-length encoding unit 102 is “101”, “1” is used as the first bit plane, and “0” is used as the second bit plane. In this case, "1" is stored in the third bit plane, and no data is stored in the fourth and subsequent bit planes. On the other hand, when the variable-length encoded data output from the variable-length encoding unit 102 is “11010”, data is stored up to the fifth bit plane.

FIG. 4 shows the pixel data sequence "0, 1, 3,...
.., 1, 2, 3,..., 2, 3,... Indicate the state in which the data encoded by the variable length encoding unit 102 is stored as a bit plane.

The hatched portion in the figure does not require bit information because the variable length code is terminated at the upper plane.
It indicates that it is not stored.

The bit plane forward scanning unit 103 receives one screen of encoded data from the variable length encoding unit 102 and stores it in the buffer 104. Subsequently, the bit plane forward scanning unit 103 sequentially outputs the first bit plane (MSB), the second bit plane,... From the buffer 104 to the lower bit plane in the order from the upper bit plane to the lower bit plane. The bit information "1/0" of the bit plane is read out in the raster scan order.

FIG. 5 shows the order of the encoded data (bit information) returned from the buffer 104. The hatched portion shown in FIG. 4 is skipped and read.
That is, in the second line of the third bit plane in FIG. 4, "1" is read from the left and then "0" is read by skipping the blank in the shaded area, and in the third line, one pixel from the left is read out. "0" is read first by skipping the shaded area of the minute. When the receiving side decoding the data shown in FIG. 5 obtained by this reading receives the data, it is known that the receiving side has read and output these data in order from the upper bit plane. It is possible to predict at which position the blank shown in FIG.

According to the data form of the present embodiment described above, the code amount can be greatly reduced as compared with the case where each pixel is simply output for each bit plane expressed as fixed bits.

The encoded data in the unit of bit plane shown in FIG. 5 is stored in the memory in the code output unit 106 or transmitted to an external device. For the code output unit 106, for example, a recording medium such as a hard disk, a RAM, a ROM, or a DVD may be used, or an interface for transmitting data to a line such as a public line, a wireless line, or a LAN. May be used.

By the above-described encoding processing, even when data is transmitted hierarchically from the upper bit plane, the outline of the image can be efficiently grasped on the receiving side. Further, the total code amount can be reduced as compared with the normal bit plane coding.

The coded data generated in the above embodiment includes the size of the image and the code table memory 102.
Table specification information on the code table stored in the table (index indicating which code table among a plurality of code tables was used, or specific data indicating each correspondence between the pixel data indicated in the code table and the variable length code) ) Etc. are added as attached data as appropriate. For example, when an image is performed in units of lines, blocks, or bands, information indicating the size of the image is required. Further, when a plurality of code tables are stored in the code table memory 105 and are selectively used according to the content of the image to be coded, the table specification information is necessary.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to the drawings.

In the present embodiment, description will be made on the assumption that 8-bit monochrome image data is encoded. However, the present invention is not limited to this, and a monochrome image represented by 4 bits per pixel or each color component (RGB / Lab) in each pixel
/ YCrCb) can be applied to the case of encoding a color multi-valued image that is expressed by 8 bits. Also,
The present invention can be applied to a case where multi-valued information representing the state of each pixel constituting an image is encoded, for example, a case where a multi-valued index value representing the color of each pixel is encoded. When applied to these, each type of multi-value information may be encoded as monochrome image data described later.

FIG. 6 shows an image processing apparatus for carrying out the second embodiment of the present invention. In the figure, 601 is an image input unit, 602 is a discrete wavelet transform unit, 603 is a buffer, 604 is a coefficient quantization unit, 605
Is a Golomb encoding unit, 606 is a bit plane forward scanning unit,
607 is a buffer, and 608 is a code output unit.

First, pixel data constituting an image to be encoded is input from the image input unit 601 in raster-scan order. As the image input unit 601, for example, an imaging device such as a scanner or a digital camera, an imaging device such as a CCD, an interface of a network line, or the like is used. In addition, the image input unit 601
Recording media such as M, ROM, hard disk, and CD-ROM may be used.

The discrete wavelet transform unit 602 converts each pixel data of one screen input from the image input unit 601 into
The data is temporarily stored in the buffer 603. Next, the buffer 603
A known discrete wavelet transform is applied to each pixel data of one screen stored in the image data, and the image data is decomposed into a plurality of frequency bands. In the present embodiment, the discrete wavelet transform for the image data sequence x (n) is performed by the following equation.

R (n) = floor ｛(x (2n) + x
(2n + 1)) / 2｝ d (n) = x (2n + 2) −x (2n + 3) + flow
r ｛(− r (n) + r (n + 2) +2) / 4｝ r (n) and d (n) are conversion coefficients, r (n) is a low frequency component, and d (n) is a high frequency component. . In the above equation, floor {X} represents the maximum integer value not exceeding X. This conversion formula is for one-dimensional data,
By applying this conversion in the horizontal and vertical directions in order, it is possible to perform a two-dimensional conversion, as shown in FIG. 7A.
It can be divided into four frequency bands (sub-blocks) of L, HL, LH, and HH.

By subjecting the generated LL component to discrete wavelet transform in the same procedure, the LL component is decomposed into seven frequency bands (sub-blocks) as shown in FIG. In the present embodiment, the discrete wavelet transform is repeated once more to perform LL, HL3, LH3, HH3, HL2, LH2, HH2, and LL as shown in FIG.
It is divided into ten frequency bands (sub-blocks) HL1, LH1, and HH1.

The conversion coefficients are LL, HL3, LH3, HH3,
The sub-blocks HL2, LH2, HH2, HL1, LH1, and HH1 are output to the coefficient quantization unit 604 in the order of raster scan in the order of sub-blocks.

The coefficient quantization unit 604 quantizes each of the wavelet transform coefficients output from the discrete wavelet transform unit 602 at a quantization step determined for each frequency component, and converts the quantized value into a Golomb code. Output to the conversion unit 605. When the coefficient value is X and the value of the quantization step for the frequency component to which the coefficient belongs is q, the quantized coefficient value Q (X) is obtained by the following equation.

Q (X) = floor ｛(X / q) +0.
In the above expression, floor {X} represents the maximum integer value not exceeding X. FIG. 8 shows the correspondence between each frequency component and the quantization step in the present embodiment. As shown in the figure, higher frequency components (HL1, LH) than lower frequency components (LL, etc.)
1, HH1) and the like have a larger quantization step.

The Golomb encoding unit 605 includes a coefficient quantization unit 60
4 to encode the quantized value and output a code. The encoded data for this quantized value is positive / negative (+ /
Golo for the sign bit representing-) and the absolute value of the quantized value
It is composed of mb code.

It should be noted that the Golomb code has k occurrence frequencies different in the degree of decrease in occurrence frequency from the occurrence probability of the value having the highest occurrence frequency (0 in the case of Golomb coding) to the value having a lower occurrence frequency. A variable length code corresponding to the distribution can be easily generated by setting the encoding parameter k. Specifically, if the parameter k used for Golomb encoding is set to be small, a pixel data group in which the occurrence frequency (occurrence probability) of pixel data to be encoded decreases from the highest value to the lowest value in occurrence frequency (occurrence probability) is large. Can be efficiently encoded, and if the parameter k is increased, the pixel data group in which the frequency of occurrence (occurrence probability) of the pixel data to be encoded decreases from the highest value to the lowest value is small. Can be efficiently encoded. For example, when encoding an image having the highest occurrence frequency of 0, if k = 0, the occurrence probability of 0 is 1/2 and the occurrence probability of 1 is 1 /
For example, a pixel data group having an occurrence frequency distribution in which the occurrence probability is greatly reduced can be efficiently encoded.

In particular, in the present embodiment, encoding efficiency is high when encoding a natural image. That is, the probability distribution of the transform coefficients obtained by performing the wavelet transform on the pixel data representing the natural image is centered on 0 (the highest occurrence frequency) in each of the sub-blocks of the transform coefficients HL3... HH1 other than the LL component. Value) as positive (+1 to ...) and negative (-1 to
...), The occurrence frequency distribution tends to gradually decrease in both directions. In Golomb coding, the transform coefficients (that is, quantized values) are 0, 1,-
Variable length coding is performed by arranging the codes in the order of 1, 2, -2,... And assigning the variable length codes with the shortest code length in this order.

Therefore, when performing the wavelet transform, the obtained transform coefficients are not simply variable-length coded,
Executing Golomb coding makes it possible to particularly improve the compression efficiency.

In a general natural image encoded in the present embodiment, the frequency of occurrence of the conversion coefficient of the high frequency component (eg, HH1 rather than HH3) from the highest frequency to the lower value is higher for the lower frequency component. (Probability of occurrence) tends to increase. Further, in the present embodiment, since the transform coefficient of the high-frequency component is quantized more coarsely than the transform coefficient of the low-frequency component, the frequency distribution of the quantized value corresponding to the high-frequency component is equal to that of the quantized value corresponding to the low-frequency component. By predicting that the degree of decrease in the occurrence frequency (occurrence probability) from the value with the highest occurrence frequency (0 in the present embodiment) to the lower value is greater than the frequency distribution, the coding parameter
k is set. Note that the correspondence between the frequency components and the encoding parameters k in the present embodiment is as shown in FIG.

The Golomb coding unit 605 performs Golomb
Since the basic method of encoding is known, only the basic operation of encoding and the characteristic parts of the present invention will be briefly described.

The Golomb encoding unit 605 first checks the positive / negative of the sequentially input quantized value, and outputs a sign (+/−) bit. Specifically, when the quantized value is 0 or positive, “1” is set as the sign bit, and when the quantized value is negative, “0” is set as the sign bit.

Next, Golomb encoding is performed on the absolute value of the quantized value. When the absolute value of the quantization value to be encoded is V and the encoding parameter for the frequency component to which the coefficient belongs is k, Golomb encoding is performed in the following procedure. First, V is shifted right by k bits to obtain an integer value m. Go for V
A lomb code is composed of a combination of m “0” s followed by “1” and V lower k bits. FIG. 10 shows an example of a Golomb code at k = 0, 1, and 2.

In the Golomb coding, coding and decoding can be performed without holding a code table (table indicating correspondence between input values and variable length codes as shown in FIG. 3). As described with reference to FIG. 3 of the first embodiment, when hierarchically recognizing sequentially from the upper bit of each variable length code, the range of the decoded value corresponding to the next lower bit can be sequentially limited. When the variable-length encoded data is hierarchically output for each bit plane, the outline of the decoded image can be quickly and efficiently recognized on the receiving side.

As described above, the coded data including the code (+/-) bits and the Golomb code for the input quantized value is generated, and output to the bit plane forward scanning unit 606.

The bit plane forward scanning unit 606 performs processing for each of the above-described frequency components (sub-blocks). First, the encoded data generated by the Golomb encoding unit 605 is stored in the buffer 607 for one frequency component (one of the subblocks LL to HH1). The sign (+/-) bits corresponding to each pixel generated by the Golomb coding unit 605 are stored in a sign plane indicating positive / negative, and the first bit (MSB) of the Golomb code corresponding to each pixel is stored in the first bit. And store the second bit in the second
Is stored in the bit plane. Similarly, the third and subsequent bits are sequentially stored in the third and subsequent bit planes. This method is the same as in the first embodiment. As described above, the encoded data corresponding to each pixel is stored in the buffer 607 as a plurality of bit planes.

For example, when the code output from the Golomb coding unit 605 is “0110”, “0” is a code plane indicating positive / negative, “1” is a first bit plane,
"1" is stored in the second bit plane and "0" is stored in the third bit plane. The data “0110”
If, no bit information is stored in the fourth bit plane.

FIG. 11 shows a data sequence "3, 4, -2, -5,-" of coefficient values (quantized values) quantized for the HL3 component.
.., "Are stored by the Golomb coding unit 605 as bit planes. In the figure, a hatched portion indicates a portion where bit information is not necessary since the encoded data is terminated in the upper plane, that is, a portion where bit information is not stored. The bit plane forward scanning unit 606 includes:
Golomb encoding section 605 outputs one frequency component (LL to H
H1), all encoded data representing any one sub-block of H1 is received and stored in the buffer 607 as described above, and a sign bit plane indicating positive / negative, a first bit plane, and a second bit plane are indicated. The information of each bit plane is read out in the order of raster scan in the order of-, that is, in order from the higher bit plane to the lower bit plane following the sign bit plane, and is output to the sign output unit 608. FIG. 12 shows a data format when the bit information stored in the buffer 607 is output in bit plane order.

The hierarchical output for each bit plane is composed of sub-blocks LL, HL3, LH3, HH3,
HL2, LH2, HH2, HL1, LH1, and HH1 are performed in this order.

The code output unit 608 sequentially transmits the plurality of bit plane data obtained by the above output in a hierarchical manner. The code output unit 608 includes a public line, a wireless line,
An interface such as a LAN can be used. Further, the code output unit 608 may be a recording medium such as a hard disk, a RAM, a ROM, or a DVD in which the hierarchical data is stored.

By the above-described encoding, images are transmitted hierarchically in order from low-frequency components to high-frequency components, and the receiving side can hierarchically grasp the outline of the images. Further, since hierarchical transmission is performed for each bit plane for each frequency component, it is possible for the receiving side to further hierarchically grasp the outline of an image for each frequency component. Also, as in the first embodiment, each pixel (transformation coefficient) is represented by a variable length, so that the overall code amount can be reduced as compared with normal encoding for each bit plane.

The coded data generated in the above-described embodiment includes an image size, the number of bits per pixel, a quantization step for each frequency component, an encoding parameter k, etc. necessary for the decoding side. Information is appropriately added. For example, when an image is performed in units of lines, blocks, or bands, information indicating the size of the image is required.

(Third Embodiment) In the second embodiment, the bit information of each bit plane is output as it is. In this case, for each quantized value quantized by the wavelet transform, one sign (+/-) bit indicating positive / negative is used, and at least one bit is used to represent the absolute value of the quantized value by Golomb code. Bits are required, for a total of 2 bits. This means that the method described in the second embodiment cannot realize compression of 2 bits or less per transform coefficient.

In the present embodiment, the bit information finally output in the second embodiment is further efficiently coded, instead of directly outputting the bit information to the code output unit, thereby obtaining the entire data. This is to reduce the code amount. Hereinafter, a specific example will be described.

FIG. 13 shows a block diagram of the third embodiment. In the figure, 1301 is an image input unit, 1302 is a discrete wavelet transform unit, 1303 is a buffer, 1304 is a coefficient quantization unit, and 1305 is Golomb.
An encoding unit, 1306, a bit plane forward scanning unit, and 130
7 is a buffer, 1308 is a run length encoding unit, 13
09 is a code output unit.

In the present embodiment, description will be made on the assumption that 8-bit monochrome image data is encoded. However, the present invention is not limited to this, and a monochrome image represented by 4 bits per pixel or each color component (RGB / Lab) in each pixel
/ YCrCb) can be applied to the case of encoding a color multi-valued image that is expressed by 8 bits. Also,
The present invention can be applied to a case where multi-valued information representing the state of each pixel constituting an image is encoded, for example, a case where a multi-valued index value representing the color of each pixel is encoded. When applied to these, each type of multi-value information may be encoded as monochrome image data described later.

Image input unit 1301, discrete wavelet transform unit 1302, buffer 1303, coefficient quantization unit 13
04, the operation of the Golomb encoding unit 1305, the bit plane forward scanning unit 1306, and the buffer 1307 are the same as those of the second embodiment. Therefore, the description of these parts is omitted.

The bit-plane forward scanning unit 1306 includes a second
The bit information of each bit plane is sequentially output to the subsequent run-length encoding unit 1308 in the same data form as the bit plane forward scanning unit 606 of the embodiment.

The run-length encoding unit 1308 includes a code plane indicating positive / negative (+/−) and a first bit (MSB) in the bit information corresponding to each bit plane received from the bit plane forward scanning unit 1306. ) For the plane, a number is generated in which the bit information "1" continues, and this continuous number is subjected to variable length coding according to the correspondence table of FIG. FIG.
Reference numeral 4 denotes a state in which a bit number “1” generates a continuous number. In FIG. 14, three bit information “1” are consecutive at first, so the first consecutive number is “3”. Since the next bit information is found to be "0", this one is skipped. Since the following bit information is followed by two “1” s, the second consecutive number is “2”.
Similarly to the above, the next bit information becomes “0” and is skipped, but the next bit information is also “0”, which means that the bit information “1” is not continuous. Therefore, the third consecutive number is “0”. Since two consecutive “0” s can be skipped, attention is paid to three consecutive bit information “1”, and “3” is output as the fourth consecutive number. You. Fig. 1
By performing encoding based on the correspondence table of No. 5, run-length encoding is performed as a result.

The run-length encoding is successively performed in the order of the second bit plane and the third bit plane. Also, due to the nature of hierarchical coding for each bit plane, it is necessary to reset the run length count for each bit plane.

The code output unit 1309 receives the run-length encoded data output from the run-length encoding unit 1308, receives the additional information separately output from the bit-plane forward scanning unit 1306, and combines the data. This is the final encoded data.

The code output unit 1309 sequentially transmits a plurality of bit plane data (data further run-length coded) obtained by the above output in a hierarchical manner. The code output unit 1309 includes a public line, a wireless line, and a LAN.
Interface can be used. The code output unit 1309 may be a recording medium such as a hard disk, a RAM, a ROM, or a DVD that stores the hierarchical data.

By the above-described encoding, images are transmitted hierarchically in order from low frequency components to high frequency components, and the receiving side can hierarchically grasp the outline of the image. Further, since hierarchical transmission is performed for each bit plane for each frequency component, it is possible for the receiving side to further hierarchically grasp the outline of an image for each frequency component. Further, the total code amount can be further reduced by further performing run-length encoding for each bit plane.
Also, as in the first embodiment, each pixel (transformation coefficient) is represented by a variable length, so that the overall code amount can be reduced as compared with normal encoding for each bit plane.

Note that the coded data generated in the above-described embodiment includes an image size, the number of bits per pixel, a quantization step for each frequency component, and additional parameters necessary for the decoding side such as a coding parameter k. Information is appropriately added. For example, when an image is performed in units of lines, blocks, or bands, information indicating the size of the image is required.

(Fourth Embodiment) In the third embodiment described above, run-length encoding is used as a technique for efficiently encoding bit information corresponding to each bit plane.
Instead of run-length encoding, it is also possible to further reduce the entire code amount by using another high-efficiency encoding method. Hereinafter, the modified examples will be described.

FIG. 16 is a block diagram showing a fourth embodiment according to the present invention. In the figure, 160
1 is an image input unit, 1602 is a discrete wavelet transform unit, 1603 is a buffer, 1604 is a coefficient quantization unit, 1
Reference numeral 605 denotes a Golomb coding unit, 1606 denotes a bit plane forward scanning unit, 1607 denotes a buffer, 1608 denotes an arithmetic coding unit, and 1609 denotes a code output unit.

In the present embodiment, description will be made on the assumption that 8-bit monochrome image data is encoded. However, the present invention is not limited to this, and a monochrome image represented by 4 bits per pixel or each color component (RGB / Lab) in each pixel
/ YCrCb) can be applied to the case of encoding a color multi-valued image that is expressed by 8 bits. Also,
The present invention can be applied to a case where multi-valued information representing the state of each pixel constituting an image is encoded, for example, a case where a multi-valued index value representing the color of each pixel is encoded. When applied to these, each type of multi-value information may be encoded as monochrome image data described later.

The image input unit 1601, discrete wavelet transform unit 1602, buffer 1603, coefficient quantization unit 16
04, the operation of the Golomb encoding unit 1605, the bit plane forward scanning unit 1606, and the operation of the buffer 1607 operate in the same manner as in the second embodiment. Therefore, the description of these parts is omitted.

The bit-plane forward scanning unit 1606 includes a second
The bit information of each bit plane is output to the subsequent arithmetic coding unit 1608 in the same data form as the bit plane forward scanning unit 606 of the embodiment.

The arithmetic coding unit 1608 separates the sequence of bit information output from the bit plane forward scanning unit 1606 into 64 states separated by 6 bits immediately before the bit of interest, and encodes them by the QM-coder. I do. The operation of QM-coder is described in Recommendation ITU-T Recommendation T.
81 | ISO / IEC 10918-1 and the like, and a description thereof will be omitted.

The arithmetic coding is successively performed in the order of the first bit plane and the second bit plane. Also, due to the nature of hierarchical coding for each bit plane,
It is necessary to reset every bit plane.

The code output unit 1609 is an arithmetic coding unit 160
The plurality of bit plane data (data further subjected to arithmetic coding) generated in step 8 are sequentially and hierarchically transmitted. For the code output unit 1309, an interface such as a public line, a wireless line, or a LAN can be used. The code output unit 1309 may be a recording medium such as a hard disk, a RAM, a ROM, or a DVD that stores the hierarchical data.

By the above-described encoding, images are transmitted hierarchically in order from low-frequency components to high-frequency components, and the receiving side can hierarchically grasp the outline of the images. Further, since hierarchical transmission is performed for each bit plane for each frequency component, it is possible for the receiving side to further hierarchically grasp the outline of an image for each frequency component. Further, by performing arithmetic coding for each bit plane, the total code amount can be further reduced. Also the first
Since each pixel (transformation coefficient) is represented by a variable length as in the case of the first embodiment, the entire code amount can be reduced as compared with ordinary coding for each bit plane.

Note that the coded data generated in the above-described embodiment includes an image size, the number of bits per pixel, a quantization step for each frequency component, and additional parameters necessary for the decoding side such as a coding parameter k. Information is appropriately added. For example, when an image is performed in units of lines, blocks, or bands, information indicating the size of the image is required.

(Other Embodiments) The present invention is not limited to the above embodiments.

For example, in the second to fourth embodiments, an example of encoding using the discrete wavelet transform has been described, but the discrete wavelet transform is also limited to the one used in the present embodiment. Instead, the type of filter and the frequency band division method may be changed. Furthermore, in addition to discrete wavelet transform, DCT transform (discrete cosine transform)
For example, the present invention may be applied to an encoding method based on another conversion method.

The method of quantizing frequency components and the method of variable-length coding are not limited to the above embodiment. For example, one frequency component (sub-block) is classified into several classes by discriminating the local properties of each divided block, and the quantization step and coding parameters are set more finely for each of these classes. You may.

Further, the configuration of Golomb coding is not limited to the above embodiment. For example, in the above embodiment, the Golomb code for the non-negative integer value V when the coding parameter is k is “1” (variable) following m (m is obtained by shifting V right by k bits) “0”. It is configured by combining the lower k bits (referred to as a fixed length part) of V and the lower k bits of V (referred to as a fixed length part). However, the usage of “0” and “1” is reversed, ie, “0” and “1” Golomb codes may be generated as “1” and “0”. Also the final Go
As the lomb code, a fixed length portion may be synthesized after the variable length portion, or a variable length portion may be synthesized after the fixed length portion.

Further, in the above embodiment, the positive / negative (+/−)
Are output separately from the bit plane corresponding to the code indicating, and the bit plane corresponding to each transform coefficient (quantized value), but the present invention is not limited to this. For example,
The sign indicating the sign may not be output as a bit plane. For example, the sign plane indicating the sign may be output while the bit plane is output hierarchically. For example, the coefficient values (quantized values) "3", "4", "-2", "-5", "-4", and "-4" shown in FIG.
When data including “0” and “1” is hierarchically output for each bit plane, a positive / negative sign is required for a quantized value other than “0”. Output the first plane. At that time, the first bit (MS
Since the original quantization value (B) corresponding to “1” (corresponding to “3”, “−2”, “0”, and “1” in FIG. 11) may be “0”, the sign is negative. Is not inserted. Meanwhile, 1
The original quantized value (“4”, “−5”, “−4” in FIG. 11) whose bit (MSB) is indicated by “0” is not likely to be “0”, so Corresponding signs “1”, “0”, “0” are inserted after all of the first bit planes (before the second bit plane) and output. The above-described insertion of the sign is performed in the same manner. It should be noted that the sign inserted and outputted for each quantized value once is sufficient, and therefore, whether to insert a sign between the second bit plane and the third bit plane is determined in FIG. This is performed for “3”, “−2”, “0”, and “1”, but is not performed for “4”, “−5”, and “−4” in FIG. It should be noted that the above-mentioned insertion and output method does not insert between the first and second bit planes, but one bit indicating each value “4”, “−5” and “−4” in the first bit plane. After each of the eyes (MSB) “1”, “1”, “1”, the plus / minus signs “1”, “0”,
"0" is added, and "1,1", "1,0", "1,0"
May be output.

Further, it is also possible to temporarily convert each of the above-mentioned quantized values having a positive / negative sign into an intermediate value of an integer having no positive / negative sign, and then to subject the intermediate value to variable length encoding (Golomb encoding).
In this case, 0. -1.1. .. Are converted into intermediate values of 0, 1, 2, 3, 4,.

Further, in the above embodiment, the transform parameter (quantized value) subjected to the wavelet transform is predicted in advance that the value of 0 occurs most frequently, and the encoding parameter k is set in advance to perform the Golomb encoding. Although the present invention has been executed, the present invention is not limited to this, as long as the setting method enables efficient Golomb coding based on the frequency of occurrence of the transform coefficient. For example, more effective encoding can be performed by actually analyzing the frequency of occurrence of transform coefficients to be variable-length encoded and setting an optimal encoding parameter k each time.

In the above-described embodiment, all of the plurality of bit planes forming one sub-block are hierarchically output, and then all of the plurality of bit planes forming the next sub-block are hierarchically output. , The hierarchical encoding is performed. For example, after outputting the first bit plane configuring the first sub-block, outputting the first bit plane configuring the next sub-block, After outputting the first bit planes of all the sub-blocks, the second sub-block constituting the first sub-block is output.
Even if the bit planes are output in an order such that hierarchical encoding can be performed.

In the third and fourth embodiments, examples have been described in which run-length coding and arithmetic coding are applied to further efficiently code bit information corresponding to each bit plane. The present invention is not limited to this, and other high-efficiency coding methods can be used.

The present invention can be applied to a single device (for example, a host computer, an interface device, a reader, a printer, etc.) even if it is applied as a part of a system. For example, the present invention may be applied to a part of an apparatus including a copying machine, a facsimile machine, a digital camera, and the like.

Further, the present invention is not limited to only the apparatus and method for realizing the above-mentioned embodiment, and the above-described embodiment is applied to a computer (CPU or MPU) in the above-mentioned system or apparatus. A software program code for realizing the program is supplied.
The present invention also includes a case where the computer of the system or the apparatus operates the various devices according to the above-described modes to realize the above-described embodiment.

In this case, the program code of the software itself realizes the function of the above-described embodiment, and the program code itself and the program code are supplied to the computer. For this purpose, a storage medium storing the above-mentioned program code is included in the scope of the present invention.

Examples of storage media for storing such program codes include floppy disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, magnetic tapes, nonvolatile memory cards, A ROM or the like can be used.

The computer controls various devices in accordance with only the supplied program codes, so that the functions of the above-described embodiment are realized. Even when the above-described embodiment is realized in cooperation with an OS (operating system) running on a computer or other application software, such a program code is within the scope of the present invention. included.

Further, after the supplied program code is stored in a memory provided on a function expansion board of a computer or a function expansion unit connected to the computer, the program code of the program code is stored. The scope of the present invention includes a case where the CPU provided in the function expansion board or the function expansion unit performs part or all of the actual processing based on the instruction, and the above-described embodiment is realized by the processing. .

[0100]

As described above, according to the present invention, coded data that has been subjected to variable-length coding based on the frequency distribution of the coefficient predicted in advance is distributed to bit planes and output hierarchically. , An outline of an image can be efficiently recognized at an early stage from a part of the encoded data. Further, it is possible to provide a technique of hierarchical coding with high compression efficiency.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first embodiment;

FIG. 2 is a diagram showing a frequency distribution of an image to be encoded in the first embodiment;

FIG. 3 is a diagram showing an example of a code table stored in a code table memory 105.

FIG. 4 is a diagram illustrating a state of bit plane division;

FIG. 5 is a diagram showing an encoded data string output by a bit plane forward scanning unit 103;

FIG. 6 is a block diagram of a second embodiment.

FIG. 7 is a schematic view of a two-dimensional wavelet transform.

FIG. 8 is a diagram showing correspondence between frequency components and quantization steps;

FIG. 9 is a diagram showing the correspondence between frequency components and coding parameters k.

FIG. 10 is a diagram illustrating an example of a Golomb code.

FIG. 11 is a diagram showing a state of bit plane division.

FIG. 12 is a diagram showing an example of a code string output from a bit plane forward scanning unit 606.

FIG. 13 is a block diagram of a third embodiment;

FIG. 14 shows an example of conversion from a bit string to a run length in run-length encoding section 1308

FIG. 15 is a diagram showing a state of encoding by a run-length encoding unit 1308.

FIG. 16 is a block diagram of a fourth embodiment according to the present invention.

[Explanation of symbols]

 101 Image Input Unit 102 Variable Length Encoding Unit 103 Bit Plane Forward Scanning Unit 104 Buffer 105 Code Table Memory 106 Code Output Unit 602 Discrete Wavelet Transform Unit 604 Coefficient Quantization Unit 605 Golomb Encoding Unit 1308 Run Length Encoding Unit 1608 Arithmetic encoder

Claims (21)

[Claims] A generating means for generating a plurality of coefficients representing an image; and a plurality of coefficients generated by the generating means are subjected to variable-length coding for each coefficient based on a frequency distribution of the coefficient predicted in advance. A plurality of bit planes by associating each bit of the variable-length encoded data corresponding to each coefficient obtained by the variable-length encoding with the variable-length encoding means with the position of each bit; And a hierarchical output unit for sequentially outputting the plurality of bit planes in a hierarchical manner. 2. The image processing apparatus according to claim 1, wherein the variable-length coded data obtained by the variable-length coding includes a Golomb code. 3. The image processing apparatus according to claim 1, wherein the plurality of coefficients are conversion coefficients obtained by converting image data representing an image into frequency components. 4. The image processing apparatus according to claim 3, wherein the conversion into the frequency component uses a wavelet transform. 5. The image processing apparatus according to claim 3, wherein the conversion into the frequency component uses DCT. 6. The image processing apparatus according to claim 1, wherein said hierarchical output means further performs run-length encoding for each bit plane to be output hierarchically. 7. The image processing apparatus according to claim 1, wherein said hierarchical output means further performs arithmetic coding for each bit plane to be output hierarchically. 8. A generating step for generating a plurality of coefficients representing an image, and a plurality of coefficients generated in the generating step are subjected to variable-length coding for each coefficient based on a frequency distribution of the coefficient predicted in advance. A plurality of bit planes by associating each bit of the variable-length encoded data corresponding to each coefficient obtained by the variable-length encoding step with the variable-length encoding step with each bit position. And a hierarchical output step of sequentially outputting the plurality of bit planes in a hierarchical manner. 9. An input step of inputting a plurality of coefficients representing an image, and a plurality of coefficients generated in the generating step are subjected to variable-length coding for each coefficient based on a frequency distribution of the coefficient predicted in advance. A variable length encoding step, and distributing each bit of the variable length encoded data corresponding to each coefficient by the variable length encoding in the variable length encoding step to a plurality of bit planes by associating each bit with a bit position. And a computer-readable storage medium storing an image processing program having a hierarchical output step of sequentially outputting the plurality of bit planes hierarchically. 10. A generating means for generating P coefficients representing a predetermined image, and a variable length encoding unit for each of the P coefficients based on a frequency distribution of the coefficients predicted in advance. Length encoding means, and 1 to m [k] (k = k) constituting variable length encoded data of each coefficient obtained by the variable length encoding by the variable length encoding means.
The 1-P, m [k] ≦ n) -th bits are assigned as the 1-m [k] -th bit planes included in the 1-n-th bit planes by associating them with the order of each bit. An image processing apparatus comprising: a hierarchical output unit that hierarchically outputs the 1st to m [k] th bit planes for each of P coefficients. 11. The image processing apparatus according to claim 10, wherein the variable length coded data obtained by the variable length coding includes a Golomb code. 12. The image processing apparatus according to claim 10, wherein the P coefficients are a part of conversion coefficients obtained by converting image data representing an image into frequency components. 13. The image processing apparatus according to claim 12, wherein a wavelet transform is used for the conversion into the frequency component. 14. The image processing apparatus according to claim 1, wherein the input unit inputs Q coefficients different from a part of the conversion coefficients obtained by converting the image data representing the image into frequency components. 13. The image processing apparatus according to claim 12, wherein the same processing as the P coefficients is performed using the variable length coding unit and the hierarchical output unit. 15. The conversion to the frequency component is performed by DCT.
The image processing apparatus according to claim 12, wherein conversion is used. 16. The image processing apparatus according to claim 10, wherein said hierarchical output means further performs run-length encoding for each bit plane to be output hierarchically. 17. The image processing apparatus according to claim 10, wherein said hierarchical output means further performs arithmetic coding for each bit plane to be output hierarchically. 18. The image processing apparatus according to claim 10, wherein the frequency distribution of the coefficient is predicted based on the content of the predetermined image that is actually subjected to variable-length coding. 19. The image according to claim 10, wherein the frequency distribution of the coefficient is predicted in advance based on a sample image instead of the content of the predetermined image which is actually subjected to variable length coding. Processing equipment. 20. A generating step of generating P coefficients representing a predetermined image, and a variable length coding of the P coefficients for each coefficient based on a frequency distribution of the coefficient predicted in advance. A length encoding step, and 1 to m [k] constituting variable length encoded data of each coefficient obtained by the variable length encoding in the variable length encoding step.
The (k = 1 to P, m [k] ≦ n) -th bit planes correspond to the order of each bit to form the 1-m [k] -th bit plane included in the 1-n-th bit plane. An image processing method, comprising a hierarchical output step of hierarchically outputting the 1st to m [k] th bit planes for each of the P coefficients. 21. A generating step of generating P coefficients representing a predetermined image, and a variable length coding of the P coefficients for each coefficient based on a frequency distribution of the coefficient predicted in advance. A length encoding step, and 1 to m [k] constituting variable length encoded data of each coefficient obtained by the variable length encoding in the variable length encoding step.
The (k = 1 to P, m [k] ≦ n) -th bit planes correspond to the order of each bit to form the 1-m [k] -th bit plane included in the 1-n-th bit plane. A storage medium storing, in a computer-readable state, an image processing program having a hierarchical output step of hierarchically outputting the 1st to m [k] th bit planes for each of the P coefficients.