JPH0436502B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an image information processing method for changing the gradation of an image.

In recent years, office automation has progressed, and various requests for image information processing have emerged. Furthermore, the work of accumulating image information and providing various information services is about to enter the practical period. However, the reality is that there are no very effective means for speeding up image processing (for example, gradation processing) when handling a large amount of image information. Therefore, in order to increase processing speed, improvements are being made mainly to electronic components and hardware configurations.

On the other hand, as for information recording media, advances in electronics technology have led to the development of large capacity memories, and the capacity of RAM, hard disks, optical disks, magnetic disks, etc. has increased significantly. or,
With the practical use of optical disks and research and development of magneto-optical disks, it has become possible to write large amounts of image information into these recording media. For example, it is possible to store 10,000 A4-sized documents on one optical disk. For this reason, storage of image information has come to rely exclusively on these large-capacity memories. Although the memory capacity has been increased in this way, in order to further increase the amount of information stored in the memory, image information is encoded and compressed before being written.

By the way, conventional image processing that involves encoding, such as gradation conversion, has been performed before encoding and compression or after decoding. For this reason, the amount of data was large and it was not possible to perform gradation conversion at high speed.

The present invention was made in view of this problem, and
It is an object of the present invention to provide an image information processing method that can perform various gradation conversions at high speed.

The method of the present invention that achieves this objective involves preparing a plurality of conversion functions for converting the density representative value of each block of image information encoded by the block coding method into a new density representative value, and The image is transformed by selecting one of the conversion functions of It is characterized by changing the gradation of the image.

Hereinafter, the present invention will be explained in detail using the drawings.

The method of the present invention is based on the block coding method. First, the block encoding method will be explained.

In the block encoding method, the image is encoded vertically and horizontally, for example, by 4
Divide into approximately 4 x sub-blocks (see Figure 1),
The density of the i and j-th pixels in this sub-block is
Xij, number of pixels is N, average density Xt (= 〓 ^ij 〓Xij/N)
When the number of pixels on the higher density side is N ₀ and the number of pixels on the lower density side is N ₁ (N ₀ + N ₁ = N), the pixel density in the block is the average value of the density on the lower density side and the higher density side. (referred to as density representative value) _, a ₀ = 〓 〓 ^Xij > ^Xt Xij／N ₀ a ₁ = 〓 _〓 ^Xij < ^Xt , a ₁ , which side has the density value is indicated by resolution information φ (takes 0 or 1 for each pixel),
According to this encoding method, the information in each block is
Determined by a ₀ , a ₁ , and φ. When the block image signal is the simplest image signal as shown in FIG. 2, it can be expressed as follows.

a ₀ = 0, a ₁ = 1, φ = {01111000} The compression rate by the above block encoding method is, for example, the number of original bits in one pixel; P ₁ = 8 the number of pixels in one block; N = 4× 4 = 16 Number of bits of a ₀ and a ₁ ; P ₂ = 8 Number of bits of φ; 1, then the number of bits of one pixel after encoding P ₃ is P ₃ = (N + 2P ₂ )/N = ( 16+16)/16=2, and the compression ratio is _P3 / _P1 =1/4.

During this encoding, it is also possible to achieve higher compression efficiency by utilizing the correlation between each block, or to perform processing such as adaptively changing the block size depending on the image content.

In the method of the present invention, gradation processing is performed by converting the density representative values a ₀ and a ₁ of each block of the image information encoded in this manner into other values.
The gradation conversion includes cases in which image information on the high density side is converted (shifted) to the low density side, and conversely, cases in which image information on the low density side is converted to the high density side. Images, especially halftone images, can cover a very wide density area, but if many pixel densities are concentrated on the high density side, it is very difficult to see, and in order to reproduce minute gradations, many gradations must be on the low density side. It is necessary to assign a key. Therefore, it is often necessary to convert information on the high-density side to low-density information. In the method of the present invention, as mentioned above, the representative density value of the block is
Convert a ₀ and a ₁ (hereinafter, both will be simply expressed as density representative value a) to other values. During this conversion, for example, the density histogram of the density representative value a is often set to a constant value. That is, histogram equalization is performed on the density representative value a. As this method, any known method can be used.

Here, we will discuss the case where conversion is performed using a simple table lookup. FIG. 3 shows a conversion table, in which a new density representative value a' is calculated from the density representative value a using the conversion function a'=f(a). If this conversion function is changed, the new density representative value a' will naturally change, and the degree of gradation conversion will also change. Therefore, it is necessary to select a conversion function that matches the intended gradation conversion. For example, in the case of a downwardly convex monotonically increasing function as shown in FIG. 3, conversion to the lower density side can be achieved. Conversely, in the case of an upwardly convex function, conversion to a higher density side can be achieved. Further, by using a function that forms a curve such as an S-shape, more complicated conversion can be performed.
Of course, a straight line may be used, and in this case, linear transformation with uniform magnification is performed. In addition, the conversion is performed using a conversion table written in ROM etc.
Or, perform calculations each time.

FIG. 4 is a block diagram schematically showing an example of an apparatus for performing the above conversion, and FIG. 5 is a flowchart of the processing. In FIG. 4, a CCD image sensor (charge coupled device) is also shown. This device stores the image output of a CCD image sensor 10 in a buffer memory 11 and divides it into subblocks, converts it into a digital signal in an A/D converter 12, and converts it into a digital signal by a microprocessor (hereinafter referred to as
(written as CPU) 13.
Note that 14 is a memory (RAM) for recording data from the CPU 13 and supplying the recorded data to the CPU 13, and 15 is a memory (ROM) for storing calculations and other programs for the CPU 13. CPU1
3 performs calculations based on the block coding method with the help of these memories 14 and 15, calculates a ₀ , a ₁ , and φ for all blocks, and stores them in the image memory 1.
6 (see flowchart in Figure 5 A),
After that, gradation conversion processing is performed. When performing gradation conversion processing (see the flowchart in Figure 5 B), the CPU
13 reads out a signal indicating the density representative value a from the image memory 16, performs a conversion operation such that a'=f(a) for all blocks, and stores it in the image memory 16 again. Then, the image after the conversion operation is displayed on the CRT 17, and the image is checked. If an unsatisfactory image is obtained during image checking, a′=f
Use another function as (a) and repeat the conversion operation.
When a satisfactory gradation is obtained during image checking, gradation conversion is completed. At this point, the image information before conversion may be deleted. Alternatively, if the image check is not satisfactory, the system may be configured to return to the flowchart in FIG. 5A and start over again from block division. If you change the block division like this,
The effect of gradation changes may be more pronounced.

Although the above explanation mainly assumes monochrome images, it goes without saying that the method of the present invention can also be applied to color images. Furthermore, the apparatus for implementing the method of the present invention is not limited to the one shown in FIG.
For example, if the gradation conversion is fixed,
There is no particular problem even if there is no CRT17 etc.

As explained above, in the present invention, since gradation conversion is performed by increasing or decreasing the density representative value a,
Compared to the conventional method that processes all pixels, high-speed processing is possible and the processing time is significantly shortened, and various tone conversions can be performed by selecting a conversion function that converts to a new density representative value. I can do it.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of a block encoding method, FIG. 2 is an explanatory diagram of an image signal, FIG. 3 is an explanatory diagram of a conversion table, and FIG. 4 is a block diagram showing an example of a device that implements the method of the present invention. FIG. 5 is a flowchart for explaining the operation of the apparatus shown in FIG. 10...CCD image sensor, 11...Buffer memory, 12...A/D converter, 13...
CPU, 14...Memory (RAM), 15...Memory (ROM), 16...Image memory, 17...
CRT.

Claims (1)

[Claims] 1 Prepare a plurality of conversion functions for converting the density representative value of each block of image information encoded by the block encoding method into a new density representative value,
By selecting one of the plurality of conversion functions and converting the density representative values of all blocks of image information encoded by the block encoding method into new density representative values based on the selected conversion function. , an image information processing method characterized by changing the gradation of an image.