JP2002209110A - Image encoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve encoding free from image deterioration and high in secondary a compression efficiency. SOLUTION: Inputted image data is cut off to be n×n pixels by a block dividing part 11 and is given filed length compression by the unit of blocks by a BTC conversion part at the next stage to output three kinds of compression codes of an average level LA within the block, a gradation width LD within the block and the quantized level ϕ of each pixel within the block. A feature value extraction part 13 makes decision on some feature by using these three kinds of compressed codes, and a data conversion part 14 converts the compressed codes of the LD, LA and ϕ based on the result of decision. The converted compressed cord is replaced with the secondary expressed code of a variable length by a Huffman encoding part 15.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding apparatus, and more particularly, to an image coding apparatus that performs fixed-length compression of image data by primary compression and then performs variable-length compression by secondary compression.

[0002]

2. Description of the Related Art In recent years, for example, A4
A rotation function for changing the size of the original image in the vertical and horizontal directions and outputting the image is provided. In order to perform such image rotation, it is necessary to once store image data for one sheet in a storage unit such as a memory. For this reason, as a compression method aimed at reducing the amount of memory used, a method of compressing data to a fixed length in a certain unit, for example, a unit of n × n pixels is used.

Further, by utilizing this fixed length compression method, all pages of a plurality of originals are fixed length compressed and stored in a storage means such as a hard disk drive (HDD), thereby providing a mechanical mechanism. An electronic sort function that changes the rotation direction for each copy and outputs the copy without using the sort function is possible. At this time, the data that has been primarily compressed to a fixed length is secondarily compressed to a variable length again, so that a larger number of document data can be stored in the HDD.

As this kind of prior art, there is, for example, an "image compression recording apparatus" described in JP-A-10-257432.

[0005]

However, in such an encoding apparatus, when an image that has been primary-compressed to a fixed length is further subjected to a secondary compression to a variable length, the conventional primary compression method is irreversible. Although compression is used, the secondary compression method is usually performed in a reversible method in order to prevent further deterioration in image quality. At this time, the secondary compression efficiency varies depending on the image. Depending on the image, the secondary compression efficiency may be high, but may be increased by expanding from the original data amount. Therefore, as for the number of sheets stored in the HDD, considering that the data amount expands, the secondary compression effect cannot be expected after all, and only the number of sheets corresponding to the compression ratio of the primary compression can be stored in the HDD.

That is, in the case of a conventional encoding apparatus, the higher the secondary compression effect is, the more data can be stored in a storage medium such as an HDD. However, there is a problem that it is not always possible to store a large amount of data.

SUMMARY OF THE INVENTION The present invention has been made in view of the above, and has been made in consideration of the above circumstances. It is an object of the present invention to provide an image encoding device capable of storing a large amount of image data by improving the compression effect while maintaining the same image quality as in the reversible time by adopting the next compression method.

[0008]

According to a first aspect of the present invention, there is provided a first compression means for compressing image data to a fixed length for each unit, and a first compression means for compressing the image data to a fixed length. Secondary compression means for compressing the next-compressed data to a variable length again, feature-value determination means for determining a feature value of an image for each unit, and primary compression based on the determination result by the feature-value determination means. Data conversion means for converting the converted data, wherein the primary compressed data converted by the data conversion means is compressed again by the secondary compression means.

According to the first aspect of the present invention, there is provided a characteristic amount judging means for judging an image characteristic amount for each unit, and the primary compressed data is converted by the data converting means based on the judgment result. As a result, high secondary compression efficiency can be obtained.

According to a second aspect of the present invention, in the image encoding apparatus according to the first aspect, the data conversion means reduces gradation.

According to the second aspect of the present invention, since the data conversion means is configured to reduce the gradation in a portion where the gradation difference is not so large, high secondary compression efficiency without image quality deterioration can be obtained. Can be

According to a third aspect of the present invention, in the image encoding apparatus according to the first aspect, the data conversion means reduces the resolution.

According to the third aspect of the present invention, since the resolution of the portion where the gradation difference is not large is reduced as the data conversion means, high secondary compression efficiency without image quality deterioration can be obtained.

According to a fourth aspect of the present invention, in the image encoding apparatus according to the first aspect, the data conversion means approximates an inclination shape between pixels.

According to the fourth aspect of the present invention, since the data conversion means approximates the inclination shape between pixels in a certain unit, high secondary compression efficiency without image quality deterioration can be obtained.

According to a fifth aspect of the present invention, in the image encoding apparatus of the first aspect, the image data is color image data, and the data conversion means performs different data conversion depending on colors. There is a feature.

According to the fifth aspect of the invention, a color image is used as input data, and the data conversion means performs different data conversion depending on the color utilizing the characteristics of the color. And high secondary compression efficiency can be obtained.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of an image coding apparatus according to the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram illustrating a schematic configuration of an image coding apparatus according to the present embodiment. The image encoding apparatus shown in FIG. 1 includes a scanner 1, an image processing unit 2, a primary compression unit 3 as a primary compression unit, a memory 4, a secondary compression unit 5 as a secondary compression unit, an HDD 6, and a secondary decompression. It comprises a unit 7, a primary extension unit 8, a plotter 9, and the like.

In FIG. 1, 8-bit / pixel data read from a scanner 1 is corrected by an image processing unit 2 and then compressed by a primary compression unit 3 into a fixed length of n × n blocks. One page of the fixed-length compressed code is stored in the memory 4 two-dimensionally like the original image. As described above, when the compressed code for one page is stored in the memory 4, irreversible variable length compression is performed in the secondary compression unit 5, and the compressed code is stored in a storage medium such as the HDD 6.

The code stored in the HDD 6 is expanded into a primary compression code by the secondary expansion unit 7 in response to a request from the plotter 9, and one page is expanded in the memory 4 two-dimensionally again.

Here, in response to a request to rotate or not rotate the output direction of the paper, a process of changing the order of the primary compression codes input to the primary decompressor 8 in accordance with the paper direction is performed. The primary decompression unit 8 converts the primary compression code to 8 bit / pixe
The data is decompressed and sent to the plotter 9.

FIG. 2 is a detailed explanatory diagram of components from the primary compression section 3 to the secondary compression section 5 in FIG. 1, and includes a block division section 11, a BTC conversion section 12, and a feature quantity as feature quantity determination means. Extractor 13, data converter 1 as data converter
4 and a Huffman encoding unit 15. It is not necessary to store one page in the memory 4 at the time of compression, so that the description is omitted here.

As shown in FIG. 2, the input image data is cut into n.times.n pixels by a block dividing unit 11, and is fixed-length compressed in block units by a BTC conversion unit 12 in the next stage. As the fixed-length compression method, there is a BTC method in addition to DCT used in JPEG or the like. In the present embodiment, a case where fixed length compression is performed by the BTC method will be described as an example.

FIG. 3 is a diagram for explaining fixed length compression by the BTC method. For example, in the BTC system with a compression ratio of 1/2 shown in FIG. 3, the LA (8 bi
t) and the average level in the block called LD (8bi
t) and φ (3bit × 16
The above three types of codes at the quantized level for each pixel in the block referred to as “number” are output.

In the feature quantity extraction unit 13 shown in FIG.
The determination of a certain feature is performed using three types of compressed codes of D, LA, and φ. The data conversion unit 14 determines the LD, LA,
The conversion of the compression code of φ is performed. The compressed code thus converted is supplied to the Huffman encoding unit 1 shown in FIG.
5 is replaced with a variable length secondary compressed code.

The data converter 14 can perform various data conversions described below.

(Conversion for Reducing Gradation) For example, the data conversion unit 14 performs a conversion when converting each of the LD, LA, and φ compression codes based on the determination result of the feature amount extraction unit 13. A conversion that reduces tonality can be performed.

First, in the feature quantity extraction unit 13 of FIG.
A determination of a feature amount for gradation is performed. For example, B
In the case of the TC method, there is data indicating a gradation width in a block called LD indicating gradation. If the LD has a certain condition, for example, LD <TH, it can be determined that the gradation width in the block is small.

Therefore, using the result of this determination, the data conversion unit 14 in FIG. 2 performs conversion of the compressed code according to the BTC method. Specifically, in a block having a small gradation width, the quantization level of φ can be reduced while preventing the image quality from deteriorating. In the BTC system with a compression ratio of 1/2, φ is originally quantized with 3 bits per pixel, that is, with an accuracy of 8 levels. The precision can be 2 bits, 1 bit, or even 0 bits. For example, when LD <TH1, φ is 2 bits (4 levels). When LD <TH2, φ is 1 bit (2 levels). When LD <TH3, φ is 0 bit (that is, φ is not output). It becomes 3 bits (8 levels). However,
It is assumed that TH1>TH2> TH3. in this way,
By reducing the output level of φ, the Huffman coding efficiency can be improved, and the secondary compression ratio can be finally improved.

(Conversion for Reducing Resolution) The data conversion unit 14 reduces the resolution when converting each of the compressed codes of LD, LA, and φ based on the determination result in the feature amount extraction unit 13 described above. Conversion can be performed.

In the same manner as described above, the gradation width in the block is determined by the LD in the feature amount extraction unit 13 in FIG. In a block having a very small gradation width, the property that the image quality does not significantly deteriorate even if the resolution is lowered is used. For example, as shown in FIG. 4, in a block where LD <TH, 1
Φ, which is one pixel (a total of 16 pixels / block), is used
Averaging 2 × 2 4 pixels as in 1
/ block). Furthermore, by combining with the above-described conversion example for reducing the gradation, the accuracy (the number of bits of φ) can be changed by the threshold value of the LD. By thus reducing the number of outputs of φ, it becomes possible to improve Huffman coding efficiency, and ultimately to improve the secondary compression ratio.

(Conversion to Approximate the Shape of Inclination Between Pixels) Further, the data converter 14 converts each of the compressed codes of LD, LA, and φ based on the determination result of the feature amount extractor 13 described above. , A conversion that approximates the inclination shape between pixels can be performed.

FIG. 5 is a diagram for explaining the processing in the data converter 14 of FIG. That is, φ output for each block is divided into four in 2 × 2 units, and simple frequency conversion is performed by the S conversion shown in FIG. At this time, L (LL) is a low frequency component (DC component) in the block.
And H (LH, HL, HH) indicates a high frequency component (AC component).

Then, the high frequency component H is approximated by a three-dimensional vector registered in advance as shown in FIG. 6, and replaced with a new code. By doing so, it becomes possible to replace the four φs with one low frequency component L and one high frequency component H.

Further, the feature quantity extraction unit 13 shown in FIG.
The accuracy of the three-dimensional code of the high frequency component H can be changed according to the threshold value of the LD in the block. For example, L
If D is large to some extent, it can be said that the gradation difference within the block, that is, the slope adjustment is large. In this case, even if the three-dimensional vector is coarse, that is, even if the number of vectors is set to be small, the quantization error is hardly noticeable.

Conversely, when the LD is small, it can be said that the gradation difference in the block, that is, the inclination adjustment is small.
In this case, the three-dimensional vector is made fine, that is, the number of vectors is prepared to increase the accuracy. As described above, by approximating the inclination shape between the pixels in the block, the secondary compression ratio can be improved without deterioration of the image quality.

FIG. 7 is a detailed explanatory diagram of the components from the primary compression section 3 to the secondary compression section 5 in FIG. 1 when a color image is used as input data.
C conversion unit 32, feature amount extraction unit 3 as feature amount determination means
3. It is composed of a data conversion unit 34 as a data conversion unit, a Huffman encoding unit 35, and the like.

In the example of FIG. 7, when a color image is used as an input image, the characteristics of the color are used for each color plate. For example, when an RGB image is converted to CMYK, the K plane tends to be generated with a relatively large gradation width compared to other colors. In a place where the gradation width is large, even if the gradation is reduced somewhat and coarsely quantized, the deterioration is hardly noticeable. Therefore, utilizing this property, the above-described (conversion for reducing the gradation) or (conversion that approximates the inclination shape between pixels) is performed only on the K plane to reduce the gradation.

That is, in the K version, the primary compression code is converted through the feature amount extraction unit 33 and the data conversion unit 34 shown in FIG. 7, and in the CMY version, these processes are passed. Without saving.

As described above, when a color image is used as input data, when a primary compression code is converted, a different conversion method is used by utilizing color characteristics, so that the image quality is deteriorated in all colors in total. Therefore, the secondary compression ratio can be improved.

As described above, according to the image encoding apparatus of the present embodiment, it is possible to perform encoding at a high secondary compression rate without image quality deterioration, and thus it is possible to store a large amount of image data. .

[0043]

As described above, according to the first aspect of the present invention, there is provided a feature amount judging means for judging the feature amount of an image for each unit, and the data converting means performs the judgment based on the judgment result. Since the primary compressed data is converted, encoding with high secondary compression efficiency can be performed.

According to the second aspect of the present invention, since the data conversion means is adapted to reduce the gradation of a portion where the gradation difference is not large, a code having high secondary compression efficiency without image quality deterioration is provided. Is possible.

According to the third aspect of the present invention, since the resolution of the portion where the gradation difference is not large is reduced as the data conversion means, encoding with high secondary compression efficiency without image quality deterioration can be performed. It becomes possible.

According to the fourth aspect of the present invention, as the data conversion means, the inclination shape between pixels in a certain unit is approximated, so that encoding with high secondary compression efficiency without image quality deterioration can be performed. It becomes possible.

According to the fifth aspect of the present invention, a color image is used as input data, and the data conversion means uses color characteristics to perform different data conversion depending on the color. Encoding with high secondary compression efficiency without deterioration is enabled.

[Brief description of the drawings]

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

FIG. 2 is a detailed explanatory diagram of components from a primary compression unit to a secondary compression unit in FIG. 1;

FIG. 3 is a diagram illustrating fixed-length compression by the BTC method.

FIG. 4 is a diagram illustrating a case in which a data conversion unit performs conversion for reducing resolution.

FIG. 5 is a diagram illustrating a case in which a data conversion unit performs conversion that approximates a slope shape between pixels.

FIG. 6 is a diagram illustrating a three-dimensional vector of a high-frequency component.

FIG. 7 is a diagram of FIG. 1 when a color image is used as input data.
FIG. 3 is a detailed explanatory diagram of components from a secondary compression section to a secondary compression section.

[Explanation of symbols]

 Reference Signs List 1 scanner 2 image processing unit 3 primary compression unit 4 memory 5 secondary compression unit 6 HDD 7 secondary expansion unit 8 primary expansion unit 9 plotter 11 block division unit 12 BTC conversion unit 13 feature amount extraction unit 14 data conversion unit 15 Huffman coding unit 21, 22 block 31 Block division unit 32 BTC conversion unit 33 Feature extraction unit 34 Data conversion unit 35 Huffman coding unit

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C078 BA23 BA53 BA57 BA64 5J064 AA01 AA02 BA09 BA15 BC01 BC02 BC14 BD01

Claims (5)

[Claims] 1. A primary compression unit for compressing image data to a fixed length for each unit, a secondary compression unit for recompressing data primary-compressed by the primary compression unit to a variable length, and a unit. A feature amount determining unit for determining a feature amount of an image for each time; and a data converting unit for performing conversion of primary-compressed data based on a determination result by the feature amount determining unit. An image encoding apparatus, wherein the compressed primary compressed data is compressed again by the secondary compression means. 2. The image encoding apparatus according to claim 1, wherein said data conversion means reduces gradation. 3. The image coding apparatus according to claim 1, wherein said data conversion means reduces a resolution. 4. The image coding apparatus according to claim 1, wherein said data conversion means approximates a slope shape between pixels. 5. The image encoding apparatus according to claim 1, wherein said image data is color image data, and said data conversion means performs different data conversion depending on colors.