JPH05347708A - Image converter - Google Patents

Abstract

PURPOSE:To provide an image converter which can perform a highly efficient density distribution inspection and the conversion of density with reduction of waste caused by the decoding job required especially for the density distribution inspection. CONSTITUTION:A density distribution inspecting part 102 applies an inspection to a DC component 107, i.e., the low frequency included in the frequency data. Then a density conversion value deciding part 103 decides the density conversion value based on the result 108 of the density distribution inspection. A density converting part 105 applies the conversion of density to the component 107 based on the deciding result of the part 103. A decoding part 105 outputs the decoding data 112 based on the AC data 111 and the density conversion DC component data 110.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention decodes gradation image data converted into frequency data by a technique such as discrete cosine transform (hereinafter referred to as DCT or DCT transform) and is suitable for an output target device. The present invention relates to an image conversion device that performs density distribution conversion.

[0002]

2. Description of the Related Art When gradation image data is output to an output device, density conversion for improving image quality is widely performed.

For example, in the chapter of "11 Image Processing Techniques (1) Preprocessing" of "Image Processing Handbook" (published by Shokoido Co., Ltd.), the density histogram of the original image is examined to obtain the image density data z.

[0004]

[Equation 1]

An example of density conversion into z'is described by. By this conversion, for example, when the original image data is in the density range from a to b, it is converted so as to be data in the density range from z1 to zk. a, b, z
The values of 1 and zk are determined to be appropriate values in consideration of the density histogram of the original image and the characteristics of the output device. As another example, as shown in FIG. 3, there is shown an example in which the density conversion is performed so that the density histogram is flattened to enhance the contrast.

Further, in the example of Japanese Patent Laid-Open No. 60-199286, when color image data consisting of three color components of red (hereinafter R), green (hereinafter G) and blue (hereinafter B) is used, all colors, A method is used in which the densities of all the pixels are not investigated, but the densities of only a part of the pixels of the green component are investigated, and this is used as a substitute for the entire concentration distribution investigation. Specifically, first, from among the input color image data of red (hereinafter R), green (hereinafter G), and blue (hereinafter B) that take values from 0 to 255, for G data, 10 pixels are used instead of all pixels. The brightness distribution obtained by sampling and surveying at a rate of 1 pixel is used instead of the brightness distribution of all pixels, and the point where the value of the cumulative brightness distribution is 1% of the whole is the dark point (the maximum value of the reproduced density) DP. ,
The 99% point is a highlight point (minimum density of reproduced) HP, and when red data is converted, for example, red data is R and cyan data is C,

[0007]

[Equation 2]

Describes an example in which R to C conversion, gamma correction, and normalization are performed simultaneously.

In any of the above examples, in order to carry out an appropriate density conversion, it is necessary to carry out a prior density distribution investigation.

On the other hand, since the gradation image data has an enormous data capacity, the opportunity for handling the compressed image data having a reduced data capacity by the compression means is increasing. Especially recently, DC
A technique of converting spatial data into frequency data by using T or the like and then compressing the data has attracted attention as a technique that achieves both high image quality and high compression rate. For example, JPEG (J
oint Photographic Expert
In the example of the Group standard, 8 × 8 pixels are set as one block, and 1) two-dimensional DCT 2) quantization of DCT coefficient to appropriate precision 3) Huffman coding of quantized DCT coefficient This enables highly efficient compression with little deterioration in image quality.

[0011]

In the prior art shown above, in order to perform the appropriate density conversion, before performing the density conversion, the entire image or a part of the image data, such as the density or the brightness, is changed. It was necessary to conduct a preliminary survey of the distribution. The example of Japanese Patent Laid-Open No. 60-199286 describes a method for directly investigating highlight points and dark points without making the survey results in the form of a histogram or cumulative frequency distribution. It is a kind of "preliminary survey of concentration distribution".

However, when the image data compressed by a complicated method such as JPEG is used as the input data, in order to perform the above-mentioned preliminary investigation of the density distribution, the compressed image is also decoded for the preliminary investigation. The problem arises that work must be done. Decoded image data usually has a capacity of tens of times, and it is often the case that all the decoded image data is stored at the time of the preliminary survey and used again at the time of density conversion. It is difficult in terms of memory capacity. For this reason, even though the decoding work is performed once at the time of the density distribution investigation, the complicated and time-consuming decoding work has to be repeated again even at the time of converting the density, and the work time and the man-hour increase. was there.

Further, as in the example of Japanese Patent Laid-Open No. 60-199286, a method of sampling and examining some percentage of data instead of performing density distribution survey on all image data is a compressed image such as JPEG. In the case of adapting to data, another problem arises. For example, if it is desired to sample only one pixel from the data compressed with 8 × 8 pixels as one block, a decoding operation is required at the time of sampling, but even if only one pixel is decoded, Since the decoding operation processing must be performed with reference to all 8 × 8 = 64 pieces of DCT data, the efficiency cannot be improved even if the number of samples is reduced.

Further, in the conventional method, since the density conversion processing is performed for each pixel, there is a problem that the processing amount becomes extremely large when handling image data having a large number of pixels.

Therefore, an object of the present invention is to solve the above-mentioned problems, and when receiving image data in the form of frequency data, it is possible to reduce the waste of performing the decoding work twice and to achieve high efficiency. An object of the present invention is to provide an image conversion apparatus which can perform an accurate density conversion by performing an appropriate density distribution investigation and can significantly reduce the density conversion processing amount.

[0016]

An image conversion apparatus of the present invention is an apparatus which uses, as input image data, frequency data obtained by converting grayscale image data into spatial frequency components, wherein low frequency components in the frequency data are A density distribution investigation unit that investigates the density distribution as a target, a density conversion amount determination unit that determines the density conversion amount based on the investigation result of the density distribution investigation unit, and a low frequency component of the density conversion amount determination unit. A density conversion unit for converting density according to the determination result to obtain a density converted low frequency component;
It is characterized by comprising a decoding unit for decoding the low-frequency component in the frequency data by the density-converted low-frequency component to obtain the density-corrected image data.

[0017]

First, the principle of image data encoding by orthogonal transformation will be briefly described. The transform coding by orthogonal transform is to realize efficient coding by removing the correlation between sample values by converting sample values into mutually orthogonal coordinate systems. Also, by converting to frequency components, allocating a large number of bits to low-frequency components where power is concentrated and also has a great influence on visual characteristics, high-frequency components are quantized with a small number of bits, Increase data compression. The orthogonal transform for transforming an image into frequency components includes Hadamard transform, DCT, and the like.

Next, regarding DCT and its inverse transformation, 8 ×
An example of encoding 8 pixels as one block is shown in FIG.
It will be explained based on. The original image data is divided into blocks of 8 × 8 pixels and converted in block units. According to the relative position of each pixel in one block, the upper left pixel is the zeroth data of the origin, and the data of the i-th pixel in the right direction and the j-th pixel in the downward direction is f (i, j) (0 ≦ i <8, 0 ≦ j <8, i and j are integers). The origin data in the upper left corner is f (0,
0) and the data in the lower right corner is f (7,7). DC at this time
The T-transformed data F (u, v) (0 ≦ u <8, 0 ≦ v <8, u, v are integers) is

[0019]

[Equation 3]

It becomes like.

F (0,0) represents a value corresponding to the average data of the image block, and is called a DC coefficient in the sense of a DC component. In the case of the formula 3, it is actually eight times the average value of the data in the block. The other 63 elements represent AC components and are also called AC coefficients. The larger the values of u and v in Expression 3, the higher the frequency component. In the case of performing data compression, the fact that the higher the frequency component in a natural image is, the smaller the value is and the more difficult it is to be recognized by human eyes, the lower bits of the high frequency component are discarded and quantization is performed with a small number of bits.

The inverse discrete cosine transform (IDCT) for obtaining f from F is

[0023]

[Equation 4]

It becomes However, c (u), c (v) in Expression 4
The definition of is the same as the case of Expression 3. Now, an embodiment of the image conversion apparatus of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram showing an embodiment of the present invention in the case of using DCT as an orthogonal transform and handling image data encoded with 8 × 8 pixels as one block.

To the data input unit 101, DCT image data 106 converted into frequency data using DCT as orthogonal transformation is input. The DCT image data 106 may be obtained directly from a file in the magnetic disk storage device or a communication line, or may be obtained in the decoding process of data compressed according to the JPEG standard or the like. The data input unit 101 uses the DCT in the first concentration distribution investigation process.
From the image data 106, only the DC component data 107 is selected as the low frequency component and output to the density distribution investigation unit 102. In Equation 3, the DC component F (0,0) has a value that is eight times the average value of the data in the block, but in this embodiment, the DC component has the same 8-bit and 256-gradation as the original image data. Shall be output after being converted.

Next, the concentration distribution investigation unit 102 conducts a concentration distribution investigation on the DC component data 107. Details of the concentration distribution investigation unit 102 will be described with reference to FIG. In the present embodiment, the distribution of the DC component value is divided into 64 sections in order to simplify the concentration distribution inspection process. Therefore, the density distribution investigation unit 102 includes 64 counters 202 from the counter 0 to the counter 63 and the counter control unit 2.
Has 01. The counter control unit 201 sets the initial values of all counters to zero before starting the investigation. Next, the counter control unit 201 reads the value of the DC component data 107, and when the value is n × 4 or more and n × 4 + 3 or less (n is an integer of 0 or more and 63 or less), increments the value of the counter n by 1,
The counting operation is performed. This count operation is all DC
By performing this on the component data 107, the value of the n-th counter is such that the DC component level is n × 4 or more n × 4 +
It shows the number of data in the range of 3 or less. As a result, in the counter 201, the DC component data 10 when the lower 2 bits are ignored and the density is divided into 64 areas
7 distribution data were obtained, and as the concentration survey result 108,
It is output to the density conversion amount determination unit 103.

The density conversion amount determination unit 103 determines the optimum density conversion amount of the DC component based on the density distribution inspection result obtained from the concentration distribution inspection unit 102.

There are many methods for determining an appropriate density conversion amount based on the result of density distribution investigation, but in this embodiment, as in the example of Japanese Patent Laid-Open No. 60-199286, the high density side is used. The value of the cumulative luminance distribution counted from the above is obtained by defining 1% of the points as the dark point: DP and 99% of the points as the highlight point: HP. It will be applied to. That is, a constant conversion equation 5 is determined by assigning constants such as b = DP, a = HP, zk = 255, and z1 = 0 in the equation 1.

[0029]

[Equation 5]

However, when z '<0, z' = 0, z '> 255
In this case, z '= 255 is set so that the conversion result does not exceed the 8-bit range.

Next, the decoding unit 105 converts the density-converted DC data 110 as the DC component of the DCT image data into AC.
The AC component data 111 is used as a component, and the IDCT processing as described above outputs the decoded data 112 obtained by converting the frequency data into the spatial image data.

In the above embodiment, the central processing unit of the computer may perform various arithmetic processing in the density conversion amount determining unit 103, the density converting unit 104, etc. by software calculation, or by a multiplier or addition by an electronic circuit. You may use the arithmetic unit by hardware comprised by means, such as combining a device.

In the above embodiment, the density conversion based on the density conversion equation such as the expression 5 is performed. However, when the output device has a specific input / output characteristic, the density conversion processing for correcting it is performed at the same time. You can go. For example, JP-A-60-199
Like Equation 2 in the example of No. 286, log may be introduced into the equation to perform gamma correction depending on the output device at the same time.

Further, as in the example of Japanese Patent Laid-Open No. 60-199286, the density distribution inspecting unit and the density conversion amount determining unit may be integrated so that the distribution inspection and the conversion amount determining are performed simultaneously.

In the above embodiment, the DC data is 8
Although it was handled with bit precision and the concentration distribution investigation was performed with 6-bit precision, these may be performed with arbitrary precision. The accuracy and the number of used bits may be changed before and after the density conversion.

The density conversion amount does not necessarily have to be represented by a specific formula. For example, by using the density distribution histogram flattening process as shown in FIG. 3, the density conversion table for flattening the DC component distribution histogram obtained by the density distribution research unit is obtained, and the conversion table is used to D
You may perform the density conversion of C component data.

Further, the concentration distribution investigation unit does not necessarily have to be all DC.
It is not necessary to investigate the data, and it is possible to use the data disclosed in JP-A-60-19928
You may investigate only the data sampled by the appropriate method like the example of No. 6.

In the above embodiments, the term "density" is used, but this may be applied to other parameters indicating brightness or darkness of an image such as brightness and brightness.

The above embodiments are examples of density conversion in a single color image, but of course the present invention may be applied to a color image composed of a plurality of color component data such as red, green and blue. . In that case, the density conversion of the present invention may be carried out independently for each color component.
As in the example of No. 199286, the density distribution investigation may be performed on only one of the color components, and the density conversion amount determined as a result may be applied to all the color components.
In addition, the DC component data of red, blue and green,
It is also possible to convert the luminance data into monochrome luminance data by an arithmetic process and perform a density distribution investigation on the luminance data.

Further, in the above embodiment, one pixel is 8 × 8 pixels.
An example of orthogonally transformed image data has been described as a block, but the pixel size of one block is 4 × 4, 16 ×
Any number such as 16 may be used. Also, a number other than a power of 2 such as 10 × 10 may be used.

As described above, in the present invention, instead of directly investigating the density distribution of the original image, the DC in the DCT image is
We are conducting concentration distribution surveys for low-frequency components such as components, and as a result, the amount of data to be surveyed is significantly reduced. As in the embodiment of FIG. 1, when 8 × 8 pixels are set as one block, the number of DC component data becomes 1/64 of the number of original image data. As described above, the DC component is a value corresponding to the average value of the data in the block. When a natural image is treated as image data, the AC component amount is generally not so large, and therefore the problem caused by targeting only the DC component does not occur much.

Similarly, in the present invention, the density conversion processing is not performed on all image data, but only on low frequency components such as DC components, thereby significantly reducing the processing amount. Increasing or decreasing the DC component value changes all data in the block by an equal amount. For example, when decoding the DCT data using the equation 4, changing the DC component F (0,0) by a certain value X
This is equivalent to uniformly changing all the decoded data by X / 8. This shows that the density correction according to each pixel data after decoding is not performed, but this also does not cause a big problem when a natural image with a small AC component is targeted, The benefits of reduced throughput are much greater.

[0043]

Therefore, an object of the present invention is to solve the above-mentioned problems, and when receiving image data in the form of frequency data, it is possible to reduce the waste of performing the decoding work twice and to improve the performance. An object of the present invention is to provide an image conversion device that can perform an appropriate density distribution investigation efficiently and can realize accurate density conversion, and can significantly reduce the amount of density conversion processing.

As described above, in the image conversion apparatus of the present invention, the density distribution inspecting unit is configured to perform the density distribution inspection for the low frequency components in the frequency data, and therefore, for the preliminary investigation of the density distribution. In addition, it is not necessary to bother to decode the image, and the number of survey target data can be significantly reduced. Therefore, when density correction is performed on a compressed image such as JPEG, the number of steps for determining the density conversion amount can be significantly reduced.

Further, since the density conversion section is also configured to perform density conversion only for low frequency components, the density conversion processing amount can be greatly reduced.

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment of the present invention.

FIG. 2 is a diagram showing an embodiment of a concentration distribution investigation unit of the present invention.

FIG. 3 is a diagram showing an example of a density conversion method.

FIG. 4 is a diagram showing DCT conversion.

[Explanation of symbols]

 102 density distribution investigation unit 103 density conversion amount determination unit 104 density conversion unit 105 decoding unit

Claims (3)

[Claims] 1. A device which uses, as input image data, frequency data obtained by converting grayscale image data into spatial frequency components, the density distribution for investigating the density distribution of low frequency components in the frequency data. And a density conversion amount determining unit that determines the density conversion amount based on the result of the density distribution investigating unit, and performs density conversion of the low frequency component according to the determination result of the density conversion amount determining unit. A density conversion unit for obtaining a frequency component, and a decoding unit for decoding the low frequency component in the frequency data with the density conversion low frequency component to obtain density corrected image data by decoding the density conversion frequency data. Image conversion device characterized by. 2. An image conversion apparatus, wherein the frequency data according to claim 1 is data obtained by orthogonally converting gradation image data. 3. An image conversion apparatus, wherein the frequency data according to claim 1 is data obtained by performing discrete cosine conversion of gradation image data.