JPH05344348A - Image encoding control method and its device - Google Patents

Abstract

PURPOSE:To obtain an image encoding control method and its device by which an image encoding processing for obtaining a high decoded image quality at a high encoding rate are attained at a high speed. CONSTITUTION:The input image signal at every color component is blocked into (n)X(n) pixels by a blocking part 1, a discrete cosine transformation is performed for the (n)X(n) pixel data by a discrete cosine converting part 2, and (n)X(n) pieces of conversion coefficients are calculated. Then, in what order of color components the input signal is placed is identified by a color component identifying part 3, when the input signal is the first color component, the feature of the image is discriminated by an image discriminating part 6, quantization table optimal to the image is selected, and the information is stored in an information storage part 7. A quantizing part 4 switches the quantization table from the information, and quantizes the conversion coefficients. An encoding part 5 encodes the quantization coefficients and the used table information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention divides an input image into blocks composed of a plurality of pixels, performs discrete cosine transform for each block, quantizes the transformed transform coefficients using a quantization table, The present invention relates to an image encoding control method and apparatus for encoding encoded coefficients.

[0002]

2. Description of the Related Art FIG. 7 shows a block diagram of encoding processing in a conventional method. First, the input image data is divided into a plurality of (n × n) pixels by the blocking unit 1. Then n ×
The discrete cosine transform unit 2 performs discrete cosine transform on the pixel data of n pixels to calculate n × n discrete cosine transform coefficients (DCT coefficients). The DCT coefficient has a low frequency region in the upper left direction and a high frequency region in the lower right direction of the n × n block. In the case of a general image, the value of the low frequency component is large and the value of the high frequency component is small.

The DCT coefficient for each block is quantized in the quantizer 4 using a quantization table with a different step size for each coefficient position. The DCT coefficients of the quantized two-dimensional array block are one-dimensionally arrayed from the low frequency side to the high frequency side, and variable length coding is performed by the coding unit 5. Further, the image determination unit 6 determines the features (characters, halftone dots, photographic images, etc.) of each block before the quantization unit 4 performs quantization, and switches the quantization table used by the quantization unit 4. On the other hand, on the decoding side, the decoding unit 8, the inverse quantization unit 9, the inverse discrete cosine transform unit 10, and the image restoration unit 11 perform the reverse processing of encoding to obtain image data.

[0004]

However, in the encoding process in the above-mentioned conventional example, when the input image is composed of a plurality of color components, the image determination unit 6 makes an image determination for each color component. Was complicated and time consuming.

The present invention has been made to solve the above problems, and provides an image coding control method and apparatus capable of performing an image coding process at a high coding rate to obtain a high decoded image quality at high speed. The purpose is to provide.

[0006]

In order to achieve the above-mentioned object, the structure of the present invention is such that an input image is divided into blocks composed of a plurality of pixels, a discrete cosine transform is carried out for each block, and transformed transform coefficients are obtained. Is quantized using a quantization table, and in the image coding control method of coding the quantized coefficient, the color component of the input image subjected to the discrete cosine transform is identified for each block, and the identified first A feature of an image is discriminated only for a color component, the discrimination result is stored as control information in the information storage means, and the quantization table is adaptively switched and controlled based on the control information stored in the information storage means. Is characterized by.

[0007]

In this structure, the color components of the input image are identified, the features of the image are discriminated only for the first color component, the discrimination result is stored as control information, and the control stored at the time of quantization is controlled. By controlling the quantization table to be adaptively switched based on the information, it is possible to perform the image coding process at a high coding rate at a high speed to obtain a high decoded image quality.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described in detail below with reference to the drawings.

<First Embodiment> FIG. 1 is a schematic block diagram showing the arrangement of an image coding apparatus according to the first embodiment. First, the input image signal for each color component is blocked by the blocking unit 1 into n × n pixels. Next, the discrete cosine transform unit 2 performs discrete cosine transform on the data of n × n pixels to calculate n × n transform coefficients. FIG. 2 is a diagram showing this discrete cosine transform formula.

Next, the color component identifying section 3 identifies which color component the input signal is. And
At the time of the first color component, the image discrimination unit 6 discriminates for each block from the characteristics of the conversion coefficient whether it is a character image block, a gravure (halftone dot) image block, or another block. .. Here, the method as described below is used for image discrimination. (1) Method of discriminating gravure image block In the coefficient (DCT coefficient) that has been discrete cosine transformed,
In the case of 8 × 8 coefficients, (b), (c),
DCT at the position of the diagonal line in either (d) or (e)
The gravure image has a feature that the coefficient has a peak and the absolute value of the DCT coefficient at the position of the diagonal line in FIG. 5A is relatively small. Using this feature, the sum of the absolute values (coefficient sums) of the DCT coefficients at the positions of the diagonal lines in (a), (b), (c), (d), and (e) of FIG. The maximum value of the sum of the coefficients of (b), (c), (d), and (e) is larger than that of (a), and (b), (c), (d),
When the maximum value of the sum of coefficients in (e) is larger than a predetermined threshold value (threshold value A), the block is determined to be a gravure image block.

That is, when the condition shown in FIG. 3 is satisfied, it is discriminated as a gravure image block. (2) Method of discriminating character image block In the character image, two lines /
It has a high value at a low frequency of about mm. DCT
The coefficient is also characterized in that the absolute value of the low frequency DCT coefficient is large. Using this feature, among the DCT coefficients, the sum of the absolute values of the coefficients at the hatched positions shown in FIG. 6 is calculated, and the sum of the coefficients is larger than a predetermined threshold value (threshold value B) and is a gravure image block. When it is not determined, it is determined to be a character image block.

That is, when the condition shown in FIG. 4 is satisfied, the character image block is determined.

The characteristic of the block is discriminated by the discrimination method as described above, and the optimum one is selected from the quantization tables prepared for the character image block, the gravure (halftone dot) image block and the other blocks. Is selected and the information is stored in the information storage unit 7.

Next, the quantizer 4 switches the quantization table from the information in the information storage 7 to quantize the transform coefficient. Then, the encoding unit 5 encodes the quantized coefficient and the table information used for the quantization.

The image discrimination processing is not performed for the second and subsequent color components, and the quantization unit 4 performs quantization by switching the quantization table from the information already stored in the information storage unit 7. Then, the quantized coefficient and the table information are encoded by the encoding unit 5 in the same manner as the first color component.

On the other hand, in the decoding process, contrary to the encoding process, first, the decoding unit 8 decodes the encoded image signal and the encoded table information. Then, the table information is stored in the information storage unit 7.

Next, in the inverse quantization section 9, the quantization table is switched from the information in the information storage section 7 to perform inverse quantization on the decoded coefficient. Then, the inversely quantized transform coefficient is subjected to inverse discrete cosine transform in the inverse discrete cosine transform unit 10. Next, the image restoration unit 11 restores an image signal from the blocked data.

<Modification of First Embodiment> The present invention is not limited to the above-described embodiment, and the color component identifying section 3 and the image determining section 6 shown in FIG. Alternatively, the image determination may be performed directly from the provided pixel data.
In this case, for example, a determination method based on an average value can be considered.
First, the average value H of a block of 8 × 8 pixels is calculated, and when the absolute value of the difference between the pixel value P and the average value H is larger than the threshold value A, the count is incremented, and the pixels larger than the average value H in the block And count C. Then, it is determined whether or not the count value C is smaller than the first threshold value A1, and when it is smaller, it is determined as a flat portion, and when the count value is between the first and second threshold values A1 and A2, it is determined as a character image, If it is larger than the second threshold A2, it is determined to be a gravure image.

According to the first embodiment described above, the discrete cosine transform section, the quantization section, the coding section, the color component identification section, the image determination section, and the information storage section are provided, and a plurality of units are provided. The input image signal made up of color components is divided into blocks made up of a plurality of pixels for each color component in the blocking unit, the discrete cosine transform is performed for each block in the discrete cosine transform unit, and the image is determined in the image determination unit from the transformed transform coefficients. In the image coding method, in which the quantized part is quantized and the quantized coefficient is coded by using the variable length code, the color component is The number of the color component input in the identification unit is identified, the image feature is identified in each block in the image identification unit only for the first color component, and the identification result is stored in the information storage unit. In order to obtain a high decoding quality with a high coding rate, the quantization table used when quantizing the DCT coefficient of each block of each color component is adaptively controlled using the information storage unit information. The image encoding process of can be performed at high speed.

<Second Embodiment> Next, a second embodiment of the present invention will be described in detail with reference to the drawings.

Usually, in the image coding apparatus, after the digital original image signal (for example, RGB, YMCK, etc.) is subjected to the discrete cosine transform, the image discrimination is performed based on the characteristic of the image, and the optimum quantization table is obtained based on the discrimination result. Is used for quantization and then for encoding.

However, since the digital original image signal includes both the luminance signal and the color difference signal, it is difficult to accurately determine the image.

The second embodiment is made to solve the above-mentioned problems, and more accurate image discrimination can be performed, an optimum quantization table can be set, a high compression rate and high image quality can be obtained. An object is to provide an image encoding device that can obtain a decoded image.

FIG. 8 is a schematic block diagram showing the structure of the image coding apparatus according to the second embodiment. In the figure, the input digital original image signal is converted into a luminance signal and a color difference signal by the color signal conversion unit 21, and the converted data is divided into blocks by the 8 × 8 blocking unit 22. Then, the discrete cosine transform unit (DCT) 23 performs discrete cosine transform, and the image determination unit 24, which will be described later, determines the type of image from the characteristics of the luminance signal. Next, the quantization table setting unit 25 sets an optimum quantization table based on the result determined by the image determination unit 24, and the DCT-transformed data is used by the quantization table set by the quantization unit 26. And quantize. Then, the direct current (DC) of the encoding unit
The component encoding unit 27 and the alternating current (AC) component encoding unit 28 encode DC and AC components, and further select signal encoding unit 2
Information indicating which quantization table is selected in 9 is encoded, multiplexed in the multiplexing unit 30 and output as encoded data.

Next, a method for identifying the type of the input image in the image determining section 24 will be described in detail.

Since the character image has a feature that the absolute value of the value of the position of the diagonal line shown in FIG. 6 is large, when the sum of the values of this portion is larger than a certain threshold value A, this block is a character image block. And the quantization table setting unit 2
In step 5, the quantization table R1 having a small quantization table value is selected so that the resolution does not deteriorate. When the sum of the absolute values of the shaded positions shown in FIG. 6 is smaller than a certain threshold value A, it is determined that this block is a graphic image, and the quantization table setting unit 25 increases the compression rate by Quantization table R having a large value in the quantization table
Select 2. Further, at this time, information indicating which quantization table is used in the quantization processing is also encoded.

<Modification of Second Embodiment> FIG. 9 is a schematic block diagram showing the structure of an image coding apparatus according to a modification. In this example, as shown in the figure, the input digital original image signal is converted into a luminance signal and a color difference signal by the color signal conversion unit 31, and the converted data is divided into blocks by the 8 × 8 blocking unit 32. .. Then, the discrete cosine transform unit (DCT) 33 performs discrete cosine transform, the edge determination unit 34 determines an edge, and the type of the input image is determined. Next, the quantization table setting unit 25 sets an optimum quantization table based on the result determined by the image determination unit 24, and the DCT-transformed data is used by the quantization table set by the quantization unit 26. And quantize. And
Direct current (DC) component encoding unit 27 of the encoding unit and alternating current (A
C) The component coding unit 28 codes the DC and AC components, the selection signal coding unit 29 codes information indicating which quantization table is selected, and the multiplexing unit 30 multiplexes the information. It is output as encoded data.

As described above, according to the second embodiment,
More accurate image determination can be performed, an optimum quantization table can be set, and a high-quality decoded image can be obtained with a high compression rate.

<Third Embodiment> Next, a third embodiment of the present invention will be described in detail with reference to the drawings.

In the second embodiment described above, the character in the color digital original image signal is regarded as an edge portion, and the character image area is determined.

However, in the second embodiment, the character image area and the area including the edge of the other image are both determined to be the character image area. Therefore, only the character image area can be selectively determined. It was difficult.

The third embodiment is made to solve the above-mentioned problems, and selectively determines a character image area in mixed color digital image data including a character image area and other image areas. An object of the present invention is to provide an image encoding device that can be performed.

FIG. 10 is a diagram showing a partial structure of an image coding apparatus according to the third embodiment. In the figure, 41
FIG. 1 is a system configuration diagram showing a character area determination device for selectively determining a character area. By this system, each image area is separated from mixed color digital image data including a character image area and other image areas. You can Hereinafter, the configuration of the character area determination device 41 will be described in detail.

First, reference numeral 42 is a color signal converter for converting the input color digital image data into Y, Cb, Cr signals. Here, when the color signals are Y, Cb, and Cr, the image data can be treated with luminance and color difference. Four
Reference numeral 3 denotes a block division unit for dividing the color digital image data converted into Y, Cb, Cr signals by the color signal conversion unit 42 into n × n blocks.
When used in G, the unit of block division is 8 ×
It becomes 8. Reference numeral 44 denotes a discrete cosine transforming unit for performing discrete cosine transform on the color digital image data divided into n × n blocks by the block dividing unit 43.

Reference numeral 45 is a character area determination unit for selectively determining the character area of the color digital image data discrete cosine transformed by the discrete cosine transformation unit 44 from the characteristics of the color signal transformed by the color signal transformation unit 42. , A character image area (white, black, etc.) can be determined by the difference in luminance.
Reference numeral 46 is a character image area storage unit for temporarily storing the data of the area determined to be a character image by the character area determination unit 45, and the data of the character image area can be output from here if necessary. it can. Reference numeral 47 denotes another area storage unit that temporarily stores the data of the area that is not determined as the character image, and can output the data of the area other than the character image, if necessary.

Next, a method of selectively determining a character area in the above-mentioned character area determining section 45 will be described below.

In the discrete cosine transform, the block dividing unit 4
When there is a character image area in the block divided by 3, the absolute value of the discrete cosine transform coefficient (DCT coefficient) in the low frequency part is large. With this feature,
For example, when the unit of the block divided by the block division unit 43 is 8 × 8, in the determination of the character image area, among the 8 × 8 DCT coefficients shown in FIG.
When the sum of absolute values of T coefficients is larger than a certain threshold value A, this block is determined as a character image area. When this threshold value becomes large, it is possible to determine that the difference in luminance is large, for example, a black character on a white background or a white character on a black character is a character area, and the others are other areas.

<Modification of Third Embodiment> FIG. 11 is a diagram showing the structure of a character area determination device 51 in a modification. In the figure, the parts indicated by 52 to 57 are the same as the parts indicated by 42 to 47 in FIG. 10 described above, and in this example, the character recognition code attachment part 58 is replaced by the character image area storage part 56.
To prepare before or after. This configuration is intended to recognize which part is a character image area by newly attaching a code. This makes it possible to determine which part is the character image area.

Here, the character recognition code attaching section 58 is for distinguishing and recognizing the character image area and the other area, and any one can be used as long as the difference between them can be recognized. By providing the character recognition code attachment unit 58, when compressing an image using JPEG or the like,
By reducing the value of the quantization table only for the character part,
It is possible to prevent the image quality of the character image from being deteriorated and to color the character portion.

In the third embodiment described above, the data input to the color signal converter 42 is converted to R, G, B or Y, M.
When the color signal is converted into C, K, etc., the image data can be treated with chromaticity. In this case, the character area determination unit 4
5 depends on chromaticity (red and green, yellow and cyan, etc.)
The character area can be determined.

According to the third embodiment, by determining the character image area from the characteristics of the discrete cosine transform coefficient, the mixed color digital image data including the character image area and the other image areas can be used to detect the respective image areas. Can be separated.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device.

Needless to say, the present invention can also be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0044]

As described above, according to the present invention,
It is possible to perform image coding processing at high speed with a high coding rate to obtain high decoded image quality.

[Brief description of drawings]

FIG. 1 is a schematic block diagram showing the configuration of an image encoding device according to a first embodiment.

FIG. 2 is a diagram showing a discrete cosine transform formula.

FIG. 3 is a diagram showing a discriminant of a gravure image.

FIG. 4 is a diagram showing a discriminant of a character image.

FIG. 5 is a diagram showing DCT coefficient positions for gravure image block discrimination.

FIG. 6 is a diagram showing DCT coefficient positions for determining a character image block.

FIG. 7 is a schematic block diagram showing a configuration of an image encoding device in a conventional example.

FIG. 8 is a schematic block diagram showing the configuration of an image encoding device according to a second embodiment.

FIG. 9 is a schematic block diagram showing the configuration of an image encoding device according to a modification of the second embodiment.

FIG. 10 is a diagram showing a partial configuration of an image coding apparatus according to a third embodiment.

FIG. 11 is a diagram showing a partial configuration of an image encoding device according to a modification of the third embodiment.

[Explanation of symbols]

 1 Blocking Unit 2 Discrete Cosine Transform Unit 3 Color Component Discrimination Unit 4 Quantization Unit 5 Encoding Unit 6 Image Discrimination Unit 7 Information Storage Unit 8 Decoding Unit 9 Inverse Quantization Unit 10 Inverse Discrete Cosine Transform Unit 11 Image Restoration Unit

Claims (3)

[Claims] 1. An input image is divided into blocks composed of a plurality of pixels, discrete cosine transform is performed for each block, and the transformed transform coefficients are quantized using a quantization table,
In the image coding control method for coding the quantized coefficient, the color component of the input image subjected to the discrete cosine transform is identified for each block, and the feature of the image is identified only for the identified first color component. An image coding control method, wherein the discrimination result is stored as control information in information storage means, and the quantization table is adaptively switched and controlled based on the control information stored in the information storage means. 2. A discrete cosine transform means for performing a discrete cosine transform, a quantizing means for quantizing using a quantization table, and an encoding means for coding the data quantized by the quantizing means. In the image encoding control device, a color signal conversion unit that converts color image data into a predetermined color signal, and an image determination unit that determines the type of image based on the color signal converted by the color signal conversion unit, An image coding control device, comprising: a quantization table setting means for setting an optimum quantization table according to a discrimination result by the image discrimination means. 3. A color signal conversion means for converting color image data into a predetermined color signal, and a character area determination means for selectively determining a character area based on the color signal converted by the color conversion means, An image coding control apparatus, comprising: a storage unit that stores the determination result of the character area determination unit.