JPH029389B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a noise removal device for removing minute isolated points included in a binary image.

BACKGROUND TECHNOLOGY In a conventional noise removal device, when removing a one-pixel isolated point of a binary image as shown in FIG. 2a as noise, the conditions of the target pixel and surrounding pixels for noise removal shown in FIG. 6 are satisfied. A local image R consisting of each pixel is extracted, and noise is removed by logical operations on each pixel (for example, Japanese Patent Laid-Open No. 34385/1983). FIG. 7 shows a block diagram of a conventional device. 1 is a binary signal when scanning an image.
4 and 5 are shift registers, which have a storage capacity for one raster and send signals from left to right. From signals 2 and 3, past signals are obtained for a time corresponding to 1 raster and 2 rasters, respectively, from the signal 1. 6 is a 3×3 array of shift registers that send signals from left to right.
pixel information can be read simultaneously. A logical operation is performed on each pixel of the local image R obtained in this way using an OR circuit 7 and an AND circuit 8 to obtain an image signal 9 after noise removal.

Problems to be Solved by the Invention However, in the noise removal device having such a configuration, in order to obtain a 3×3 local image R, the target image must be scanned pixel by pixel. This is also true for images that are already stored as binary data in a memory or the like, since the data corresponding to each pixel must be extracted one by one, which takes time.

The present invention has been made in view of these points, and an object of the present invention is to provide a noise removal device that can process images already stored in a memory or the like as binary data at high speed.

Means for Solving the Problems In order to solve the above problems, the present invention stores the target image line by line, shifts the data for each line to the right and stores it, and further shifts the data for each line to the left. The data is stored and logically operated on the data for each line to remove minute isolated points.

Effect: By the means described above in the present invention, even images that are already stored as binary data in a memory, etc.
Instead of extracting data corresponding to each pixel one by one, the data stored for each line is shifted to the right or left to create data for surrounding pixels for noise removal of the pixel of interest for each line, resulting in high-speed noise removal. This makes removal possible.

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of a noise removal device according to the present invention. Reference numeral 10 denotes an image input unit which scans the target image and inputs the image as a binary signal. 11 is a line storage section, and an image input section 10
The input binary signal is stored line by line. 1
2, 13, 14 are registers R _o of the line storage section
1, R _o 2, R _o 3, where R _o 2 is the data of the line of interest, R _o 1 is the data of the line one line above the line of interest, and R _o 3 is the data of the line one line below the line of interest. Remember. 15 is an image right shift section;
Data for each line in the line storage section 11 is shifted to the right by one pixel and stored. 16, 17, and 18 are registers R _r 1, R _r 2, and R _r 3 of the image right shift section,
The data obtained by shifting the data of R _o 1, R _o 2, and R _o 3 by one pixel to the right is stored. Reference numeral 19 denotes an image left shift unit, which shifts the data for each line in the line storage unit 11 to the left by one pixel and stores the data. 20,2
1 and 22 are registers R _l 1 and R _l of the image left shift section
2 and R _l 3, and store data obtained by shifting the data of R _o 1, R _o 2, and R _o 3 by one pixel to the left, respectively. 23 is a noise judgment section, and register R _o
1・R _o 2・R _o 3・R _r 1・R _r 2・R _r 3・R _l 1・R _l
Perform the logical operation 2.R _l 3 to remove minute isolated points on the line of interest. 24 is an image storage unit that stores image data from which noise has been removed by the noise determination unit 23.

The operation of the noise removal apparatus configured as described above will be explained using the image shown in FIG. 2a as an example.
P _j i in the figure represents each pixel, i represents a pixel number in the horizontal direction, and j represents a line number. Moreover, P ₂₇ is a minute isolated point.

The target image shown in FIG.
Valued. FIG. 2b shows the data D _j after binarization. Each pixel corresponds to each bit, with a white pixel being "0" and a black pixel being "1". The binarized data is sequentially stored in the line storage unit 1 starting from j=1.
1 and stored in registers R _o 1, R _o 2, and R _o 3. Figure 3a shows the contents of each R _o register when data up to the third line is input.
The data stored in the R _o 1, R _o 2, and R _o 3 registers are shifted to the right by one pixel by an image right shift unit 15 and stored in the registers R _r 1, R _r 2, and R _r 3 in order. Furthermore, the data in the R _o 1, R _o 2, and R _o 3 registers are shifted to the left by one pixel in the image left shift section 19 and stored in the registers R _l 1, R _l 2, and R _l 3 in order. Each R _o up to the third line data is input to the 3rd register.
The contents of the R _r register and R _l register are shown in Figure 3b.
Shown in c. The noise determination section 23 performs the logical operation shown in the first equation on the line-by-line data obtained in this manner and stored in each register to remove minute isolated points. If D′ is the data after removing minute isolated points, D′=R _o 2∩(R _r 1∩R _r 2∩R _r 3∩R _o 1∩R _o 3∩R _l 1∩R
_l 2∩R _l 3)……(1). Figure 3d shows the contents of D' that are sequentially determined. The data obtained in this manner after the removal of minute isolated points is stored in the image storage unit 24 as a noise-removed image. FIG. 4 shows the image data obtained after noise removal. Comparing this data with the data in FIG. 2b, it can be seen that the minute isolated point _P27 has been removed.

Next, the fact that the first equation corresponds to the logical operation OR circuit 7 and AND circuit 8 of the prior art pixel of interest and surrounding pixels for noise removal on a line-by-line basis will be explained using the minute isolated point _P27 as an example. Figure 5 shows each _Ro・
The contents of the R _r and R _l registers are shown in bits, and the line including the pixel of interest P ₂₇ is stored in R _o 2 as the line of interest. _P27 is the 9th bit of R _o 2. Each R _o register stores data for each line in order, so the surrounding pixels in the upper direction
P ₁₇ is the 9th bit of R _o 1, the lower peripheral pixel P ₃₇
is stored in the 9th bit of R _o 3. Also,
Each R _r register is obtained by shifting each R _o register by 1 bit to the right, and the surrounding pixel P ₁₆ in the upper left direction is R _r
The 9th bit of 1, the surrounding pixel P ₂₆ to the left is the 9th bit of R _r 2, and the surrounding pixel P ₃₆ to the lower left is 9 of R _r 3.
Each bit is stored separately. Similarly, each
The R _l register is obtained by shifting each R _o register by 1 bit to the left, and the surrounding pixel _P18 in the upper right direction is R _l 1.
The 9th bit of , rightward surrounding pixel P ₂₈ is R _l 2 no 9
The peripheral pixel _P38 in the lower right direction is stored in the 9th bit of _Pl3 . In this way, the pixel of interest P ₂₇ and the surrounding pixels for noise removal are each R _o
It can be expressed by the 9th bit of the R _r and R _l registers.
This is the same for each pixel on the line of interest.
In each of the R _o , R _r , and R _l registers, data of surrounding pixels for noise removal is stored at the same bit position as the bit position of the R _o 2 register where the data of each pixel of the line of interest is stored. In other words, the local image R shown in Figure 6 is represented by data at the same bit position in each R _o , R _r , and R _l register shown in Figure 1, and noise can be removed by line-by-line data processing. can.

Effects of the Invention The present invention stores target image data line by line, shifts the data for each line to the right and stores it, further shifts the data for each line to the left and stores it, and performs logical operations on the data for each line. By removing small isolated points, it is possible to perform noise removal processing at high speed, even for images that are already stored in memory as binary data, without having to extract data corresponding to each pixel one by one. can.

[Brief explanation of drawings]

Fig. 1 is a block diagram of a noise removal device according to an embodiment of the present invention, Fig. 2 a and b are diagrams showing an image to be noise removed and its binarized data, and Fig. 3 a to d are diagrams showing each register and noise An explanatory diagram showing changes in data content after noise removal. Figure 4 is a diagram showing data after noise removal from an image to be noise removed. Figure 5 is a diagram showing the bit contents of each register and a local image of the image to be noise removed. Relationship diagram explaining relationships, No. 6
The figure is an explanatory diagram of a local image for noise removal according to a conventional technique, and FIG. 7 is a block diagram of the conventional example. 11...Line storage unit, 12...Register R _o
1, 13...Register R _o 2, 14...Register
R _o 3, 15... Image right shift section, 16... Register R _r 1, 17... Register R _r 2, 18... Register R _r 3, 19... Image left shift section, 20...
...Register R _l 1, 21...Register R _l 2, 22...
...Register R _l 3, 23...Noise judgment section.

Claims (1)

[Claims] 1. When removing minute isolated points in a binary image, a line storage unit stores the target image line by line, and an image right shift unit stores the line-by-line data of the line storage unit by shifting it to the right line by line. and,
an image left shift section that shifts and stores the data for each line in the line storage section to the left line by line; and a logical operation of the data for each line in the line storage section, the image right shift section, and the image left shift section. 1. A noise removal device comprising: a noise determination unit that removes minute isolated points.