JPH0556263A - Printer - Google Patents

Abstract

PURPOSE:To prevent the omission of high frequency information and to realize the stability of a natural picture easily and at a low price by performing the switching processing of plural multi-value error diffusion processings in accordance with the frequency characteristic of the print data in a multi-value processing member. CONSTITUTION:Data are inputted through an interface processing part 101, and a multi-value processing part 102 switches the multi-value error diffusion processing in accordance with the frequency of the print data for a printer engine part for the inputted half tone data, and selects plural dot gradations. A correcting processing part 103 corrects the multi-value processed output in accordance with the number of gradations and the environmental change, and a gradation processing part 104 performs the converting processing so as to execute the gradation expression matched to the engine and outputs it to an engine part 105.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a personal computer, a workstation or the like, or an input device such as a scanner,
Further, the present invention relates to a printing device that receives and prints print image information and character information from a communication line or the like.

[0002]

2. Description of the Related Art Conventionally, in the case of expressing a halftone of an image or the like, a halftone data such as a systematic dither process or an error diffusion method is used in a device in which the dot expressive power of a printing device has only binary expressive power. The method of expressing gradation by converting the gradation data into binary data by the binarization processing of 2 is adopted. In such a binary output printing apparatus, in order to improve the quality of an image, the output resolution is increased to improve the gradation expression ability. On the other hand, when the print engine has a dot gradation expression, a high quality image can be obtained even if the resolution is low. Generally, the relationship between the resolution of natural images and the number of gradations is almost satisfactory.
The relationship is as shown in FIG. 11, and the higher the resolution, the lower the required number of gradations. Considering the stability of dot reproduction in terms of the number of gradations and the resolution, a lower resolution generally has a higher dot stability, and it is often easier to express halftones. Therefore, when expressing an image, it is effective considering the cost for reproduction to reproduce an image containing many halftones such as natural images by lowering the resolution and increasing the gradation. And is commonly practiced.

[0003]

When an image is expressed by decreasing the resolution and increasing the number of gradations, if the image to be expressed contains high resolution information, the high frequency information Will be missing. For an image containing high-frequency information, it is necessary to easily configure the stability of a natural image and the prevention of high-frequency information loss at a low price.

[0004]

SUMMARY OF THE INVENTION Therefore, the present invention provides an interface unit for a multi-valued printing device,
A multi-value processing unit, a correction processing unit, a gradation processing unit, and an engine unit are provided, and a process of switching a plurality of multi-value error diffusion processes in accordance with the frequency characteristics of print data is performed in the multi-value processing unit. Constitute.

[0005]

According to the present invention, it is possible to realize, at a low price, a stable image expression with little paper dependency for a natural image including both high and low frequency components and an image including characters and natural images. ..

[0006]

EXAMPLES The present invention will be described in detail below based on examples.

FIG. 1 shows an embodiment of the present invention, in which 101 is an interface processing unit, 102 is a processing unit for multivalued printing data, 103 is a processing unit for correcting multivalued data, and 104. Gradation processing unit for gradation printing, 10
An engine unit 5 includes a mechanical mechanism for printing.

FIG. 2 shows a more specific embodiment configured inside the correction processing unit 102 of FIG. 1, in which 201 is a multistage error diffusion processing unit, 202 is another multivalued error diffusion processing unit, and 203.
Is a frequency detection processing unit, and 204 is a selection unit.

A detailed description will be given according to the flow of operation.
In the interface processing unit 101, when data is delivered from another processing unit in the printer, the data for printing is transferred as a parallel data bus interface according to a predetermined rule. Also,
When sent from other equipment, such as a personal computer, it is a part that transfers data according to the standard interface format such as Centronics standard interface, SCSI interface, GPIB interface, etc. through this part. Data is entered. Next, the multi-value quantization processing unit 102
Is an image processing unit for converting the input halftone data into data for the printer engine. For example,
In the case of 8-bit input data, in order to accurately reproduce high image quality, the multilevel halftoning processing unit 102 is unnecessary if the engine 105 has the capability of expressing 8 bits. When the expression ability is limited, the processing for increasing the reproducibility is performed by the data processing according to the ability of the engine. Various methods have been proposed for increasing the quality of a reproduced image when the engine performance is lower than the number of gradations of the data to be reproduced. In the present invention, the configuration of multi-value quantization processing is the basic part. That is, the multi-valued error diffusion processing 201 shown in FIG. 2 is a processing means adapted to multi-valued output, which is a commonly used binarization means for converting multi-valued data into binary data. .. In addition, the multi-value error diffusion processing 202
Is an increase in the number of output expressions in the process of 201. In the multi-valued error diffusion processing of 201, the number of output dot gradation expressions is relatively small, and corresponds to processing for obtaining 2-level and 4-level outputs, for example. Further, the multi-valued error diffusion process 202 is a process for reproducing multi-gradation output intermediate value dots of 8 levels and 16 levels, for example. The frequency detection processing unit 203 determines the frequency characteristics of the input image data by performing both processes.
Then, the selection unit 204 switches between the two processes. Next, the correction processing unit 103 in FIG. 1 is a unit that further increases the number of gradations of the output that has been subjected to the multi-valued processing according to the engine and corrects it in response to environmental changes. The gradation processing unit 104 is a processing unit that converts the data that has undergone the correction processing so as to perform gradation expression that matches the engine. A pulse is generated according to the energization period, or a variable signal by voltage is generated. The engine unit 105 shows an engine centered on a mechanism. This may be a thermal transfer printer or another type of engine.

Naturally, a mechanism for controlling the engine and a processing unit are present, though not shown. Any control mechanism may be used.

Next, the basic dot stability for carrying out the present invention will be described. For example, when observing the growth of dots with a thermal transfer printer or the like, when the dots are continuously energized on the line head, the growth becomes as shown in A of FIG. The key is limited. On the other hand, when there is no adjacent dot, the dot growth is stably increased as shown in FIG. 9B. In this case, in the case of B, since the area is larger than that of A, stability is naturally high and a large number of gradations are generated.
It will be stable. FIG. 10 shows the relationship between the energy input and the density when printed on paper for each of A and B in FIG. That is, when the dots are separated as in B, a large amount of energy is applied to express a large number of gradations. From the viewpoint of stability and gradation, how to increase the dot size, and from the viewpoint of resolution, how to increase the number of dots. The present invention is a method of simultaneously satisfying this, and will be described more specifically.

FIG. 3 shows an embodiment of the present invention, which is an embodiment of the concrete contents of the multi-value quantization processing of FIG. Reference numeral 301 is a frequency detection processing unit for sensing the frequency characteristic of data, 302 is a block data multiplication unit for processing data according to the frequency detection result, and 303 is a multi-value slice level for generating comparison data of error diffusion. , A comparator 304 for comparing the data with the slice level, an adder 305 for adding the input data and the data from the error memory,
A weight matrix unit 306 weights the data in the error memory, and an error memory 307 stores the error amount. It will be explained according to the flow of data. The data input to the multi-value quantization processing unit 301 is compared with the data in the block by the frequency detection processing unit 301, and is classified into high frequency blocks or low frequency blocks. That is, for example, as shown in FIG. 7, each pixel data is classified into a high frequency component region A or a low frequency component region B in units of 2 × 2 pixels. The classified result is determined for each block, for example, as shown in FIG. This determination method may be performed in the horizontal direction for each pixel or in blocks. In order to improve accuracy, it is desirable to make a determination for each pixel. As a specific method for the determination, it is possible to compare the data values of the respective pixels and determine that the data having a certain difference or more is in the high frequency region. Alternatively, there is also a method of comparing with the average value of each block. Classification is performed regardless of which method is used. The classification result is supplied to the block data multiplication unit 302 and the multi-value slice level generation unit 303. The block data multiplication unit 302 is provided with a multiplication table according to the respective frequency component areas A and B as shown in FIG. 4, for example, and the constant of the corresponding portion of each pixel is multiplied by the data. FIG. 4A is an example of constants corresponding to the low frequency block B, and FIG. 4B is an example of constants corresponding to the high frequency block A. Here, the B block in FIG. 4A is multiplied by a constant four times, because the B block does not increase its concentration unless a large amount of energy is supplied, as shown in FIG. On the other hand, it becomes multiplication for correction. Further, once the high energy of B is applied, the dot next to the dot produces an effect of being automatically corrected to dot data of a small size by the feedback from the error memory. In the multi-value slice level generation unit 303, a threshold table as shown in FIG. 6 is prepared.
A threshold value corresponding to each block is generated. Each cell corresponds to each pixel, and each number shows an example of a threshold value when an 8-bit 256 level is assumed. Figure 6 (a)
Is an example of a high frequency region, and the threshold level is 3 levels (64,
128, 192), and dot output is 0, 64, 12
It is an example classified into 5 levels of 8, 192, and 256.
FIG. 6B shows an example of 16 levels in the low frequency region. The two tables are switched by a frequency detection signal to select a plurality of dot gradations. The data compared by the comparator 304 is output to the next stage as multi-value output information. Although it is possible to take out as binary output data simply by using a comparator alone, multi-valued output becomes possible by taking out outputs that match a plurality of types of multi-valued slice level tables. Although various methods of making the output level multi-valued are conceivable, it is fundamental to store the error between the output level and the level input as data in the error memory 307. The error memory 307 is the same as the general error diffusion, and stores the error data of each dot of the line, and the weight matrix 306 adds the called error value to the input data and the addition unit 305 to perform feedback control. Will be costly. FIG. 5 shows an example of the weight matrix. The shaded area means the dot data to be compared, and the error values around it are added together with such weighting. This matrix can be freely selected depending on the ease of calculation.

[0013]

According to the present invention, high resolution and high gradation can be utilized more easily, and the image quality of the printing apparatus can be improved.

[Brief description of drawings]

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

FIG. 2 is a diagram showing details of a correction processing unit.

FIG. 3 is a diagram showing details of a multi-value quantization processing unit.

FIG. 4 is a diagram showing a multiplication table according to each frequency component region.

FIG. 5 is a diagram showing a weight matrix.

FIG. 6 is a diagram showing a threshold table.

FIG. 7 is a diagram showing classification of each pixel data.

FIG. 8 is a diagram showing the classification result of FIG. 7.

FIG. 9 is a diagram showing dot growth in a thermal transfer printer.

FIG. 10 is a graph showing the relationship between input energy and concentration.

FIG. 11 is a graph showing the relationship between resolution and the number of gradations.

[Explanation of symbols]

 102 multi-value processing unit 201, 202 multi-value error diffusion processing unit 301 frequency detection processing unit 307 error memory

Claims (1)

[Claims] 1. A multi-valued printing device comprising an interface section, a multi-value processing section, a correction processing section, a gradation processing section, and an engine section, including dot modulation in the multi-value processing section. A printing apparatus, wherein a plurality of multi-valued error diffusion processes are switched according to frequency characteristics of print data.