JPS58114282A - Binary coding device - Google Patents

Abstract

PURPOSE:To correctly perform binary coding of two-dimensional images, by dividing a reading area into blocks, and implementing slice levels individually at every block correspondingly to the maximum printing depth at each block. CONSTITUTION:A memory 1 samples two-dimentional image signals and stores analog signal levels corresponding to the black and white of the image signals as two-dimensional image data. The two-dimensional image data are given to a white/black judging circuit 3 and a slicing circuit 6 after they are divided into plural blocks by a block dividing circuit 2. By the white/black judging circuit 3, a maximum black level detecting circuit 4, and a slice implementing circuit 5, slice levels are individually implemented at every block in accordance with the maximum printing depth of each block, and two-dimensional image signals which are binary-coded are outputted from the slicing circuit 6.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical Field of the Invention The present invention is an improvement of a signal binarization device, particularly a device for binarizing a two-dimensional image such as a seal stamp or printing by a type printer.

(2) Foreground of technology When automatically reading or failing a seal stamp or printing by a type printer, first use a sensor such as a t-TV camera to convert the two-dimensional image into an eirin signal, sample it, and convert it into a binary code. However, the general fixed slice level method is not suitable for such a binarization device for reading print. That is.

If the pressing force is uneven when printing a seal stamp or type, there may be deviations even within the print density A1 character, and if this is binarized at a fixed slice level, it will not be possible to accurately represent the original seal impression or type pattern. This may result in cases where two-dimensional data cannot be created.

(3) Purpose of the Invention In view of the above points, an object of the present invention is to enable accurate bipolarization of two-dimensional images such as printed patterns, and in particular to create appropriate slice levels for rounding. The purpose is to provide a new device for performing

(4) Structure of the Invention The binarization device according to the present invention is a device that creates a binary signal corresponding to black and white from a two-dimensional image, and divides a three-dimensional image into a plurality of blocks. At each first slice level, it is detected whether an island portion exists within the block.

In a block where it is determined that the island portion remains, 1g is applied according to the signal level closest to the black level within the block.
The slice level of 2 was set and binarization was performed.

(5) Embodiments of the Invention Figure 1 Axis) IIi This is a diagram showing an embodiment of the present invention, 1
Fi is a memory that samples the two-dimensional image signal t and stores it as analog signal level 1i-two-dimensional image data according to its black and white. Reference numeral 2 denotes a block division circuit which divides this two-dimensional image data into a plurality of blocks and reads out each block. This block division is performed, for example, by equally dividing the two-dimensional image to be read into 16 4×4 areas H-p as shown in FIG. 1(b). Image data read in blocks.

First, the black and white judgment circuit 3 determines whether or not an island portion exists in the block based on the slice level x't-0 This slice level x#'i is a fixed slice level for normal binarization rounding. Equivalent to or higher than (
For example, the level is set to be the most distant from the white level by about 15 inches of the black and white maximum level width.

In this example, it is assumed that the paper on which the print is printed is of a standard size, and the white level is always constant, but the main consideration is to deal with variations in print and stamp density. Therefore, the black and white judgment slice level X is set to a constant level as described above. However, this may be variable depending on the white level. FIG. 1() shows a slice Im of the image data signal at this black-and-white determination level X, and here, an island portion remains within the block f t 1ieJ.

Block LgeJtk where the island part remains? During reading, the maximum black level is detected for each block. The ninth circuit 4#i can be constructed by known means such as a suitable peak hold circuit. The slice creation circuit 5 creates a slice level corresponding to the maximum black level for a block including an island portion.

For example, this slice level is determined by the maximum black level detection circuit 4.
It is sufficient that the slice generation circuit generates a level close to the level determination slice level X of 50- the maximum black level generated from the maximum black level, or the moth level.

Using the slice level thus generated for each block, the slice circuit 6 binarizes the two-dimensional image data within that block.

When binarizing through the above process, the two-dimensional image data is divided into blocks, and a slice level is used that corresponds to the black level within each block (simply, proportional to the maximum black level). By appropriately setting the 9-division block size, it is possible to improve printing/impression density accuracy. Specifically, when compared with the case of a fixed slice level, in the settings of the above embodiment, it is assumed that an island exists in a block where the black level signal strength is 15- or more with respect to the maximum level width, and the maximum Since the level of 50 degrees of the black level signal becomes the final slice level, the lowest slice level that can be taken, the value is 7 for the maximum repetition width.
．． It is 5%. Therefore, the black level of 7.5 to 15 inches, which was lost in the conventional fixed slice level method, can be recovered and detected. Of course, there is a possibility that the black level signal in this range is noise, but this signal is only detected as black when the island part in the block has a low concentration, so such a If the black level is relatively low, the signal part is originally black and &! There are an overwhelming number of cases in which this should be recognized. On the other hand, if there is no low-density black level part, signals with levels within the above range are treated as noise, so less noise is picked up. Therefore, even if there is bias or variation in print density, more accurate fkZ value conversion can be achieved.

(6) Effects of the Invention According to the present invention, even if the print density of a seal stamp, typeface, etc. is partially uneven, the reading area is divided into blocks and each block is adjusted according to its maximum print density. Since slice levels are created individually, it is possible to more accurately binarize a two-dimensional image such as a seal impression or print pattern.

[Brief explanation of the drawing]

f[Figures 1-) to (c) are circuit diagrams for explaining embodiments of the present invention, blocks of image data are shown, and signal waveform diagrams are shown, respectively.

Claims (1)

[Claims] This is a device that creates a binary code corresponding to black and white from a two-dimensional image, and divides the two-dimensional image into a plurality of blocks and creates an island part within the block at the 1th slice level for each block. In a block where it is determined that an island exists, the slice level of Yoshiji 2 is set according to the signal level closest to the black level in the block, and binarization is performed. A binarization device 0 characterized by