JPH01181168A - Picture signal processing method - Google Patents

Abstract

PURPOSE:To eliminate the picture noises with no deterioration of contrast nor the blur of picture elements by replacing the density of a noted picture element with the mean density when the difference between the mean density of a matrix formed around the noted picture element and the density of this noted picture element itself is kept outside a fixed range. CONSTITUTION:A CPU sets a matrix containing 9 picture elements A-I around a noted picture element E and obtains the mean density e' from the arithmetic average of 8 picture elements with the element E deleted. A subtraction means 48 obtains the absolute value of the difference between the density e of the noticed picture element and the density e' via an absolute value means 50. This absolute value is compared with the set value DELTAX by a comparator 52. When the absolute value is larger than the value DELTAX, a switch means 56 selects and delivers the density e'. While the density e is delivered as it is for correction when the absolute value is smaller than the value DELTAX.

Description

Detailed Description of the Invention (Field of Industrial Application) The present invention removes noise by correcting the density of the pixel of interest using the density of each pixel in a fixed area centered on the pixel of interest. The present invention relates to an image signal processing method for improving image quality.

(Technical Background of the Invention) As a method for improving the image quality of an image, it is known to correct the density of a pixel of interest using the density of each pixel in a certain area centered on the pixel of interest. For example, as a method of smoothing processing to remove noise (3I sounds) included in an image, it is known to replace the density of a pixel of interest with the average value of the density of a fixed small area around the pixel. However, this method has a problem in that the contrast of the image decreases, resulting in blurring.

(Object of the Invention) The present invention was made in view of the above circumstances, and an object of the present invention is to provide an image signal processing method capable of removing noise without reducing the contrast of an image. be.

(Structure of the Invention) According to the present invention, it is an object of the present invention to provide an image signal processing method for correcting the density of a pixel of interest using the density of each pixel of a certain region centering on a pixel of interest. If the difference between this average density and the density of the pixel of interest is outside a certain range, replace the density of the pixel of interest with this average density, and if the difference is within a certain range, replace the density of the pixel of interest. This is achieved by an image signal processing method characterized by using the density as it is.

(Example) Fig. 1 is a block diagram showing the operation of an embodiment of the present invention, Fig. 2 is an output waveform diagram in each processing process, and Fig. 3 is an example of a matrix showing pixel arrangement in a certain area. 4 are diagrams showing an example in which the present invention is applied to a microfilm reader.

In FIG. 4, reference numeral 10 denotes a light source, and the light from this light source 10 is guided to a microfilm 18 as an original via a condenser lens 12, a heat shielding filter 14, and a mirror 16. The light transmitted through this microfilm 18 is guided to a transmission screen 28 by a projection lens 20 and mirrors 22, 24, 26, and an image of the microfilm 18 is formed on this screen 28.

A CCD line sensor 30 is disposed on the back side of the screen 28 so as to be movable one row below or to the left or right. That is, this line sensor 30 is arranged long in the direction that crosses the screen 28 horizontally or vertically, and the line sensor 30 performs main scanning by reading out the image signal α in the length direction and sequentially outputting the image signal α as a time-series signal. Further, sub-scanning is performed by moving this line sensor 30 vertically or horizontally. The image signal α from the line sensor 30 is input to the CPU 34 via the input interface 32, where various processes are performed as will be described later, and the image signal is binarized. This binarized image signal δ is outputted to the printer 38 via the output interface 36, where the image is printed out and a hard copy is obtained. Also, the binary signal δ outputted by the CPU 34
can be recorded on an external storage device such as the optical disk device 40.

Therefore, a necessary original image may be read out from the optical disk device 40 and printed out by the printer 38, if necessary. Note that 42 is a memory of the CPU 34.

The CPU 34 sequentially performs the operations shown in FIG. That is, C
The PU 34 first reads the image on the screen 28 while moving the line sensor 30, and stores the image signal α for one page of the screen in the memory 42 (memory means 44 in FIG. 1). In the case of a negative image, the image signal α at this time has a low density level in the background and a high density level in the image area, as shown in Fig. 2A, and especially the background contains a lot of noise. In FIG. 2, the horizontal axis represents time r (or pixel order), and the vertical axis represents density level h.

The CPU 34 sets a matrix consisting of pixels in a fixed area centered on the pixel of interest, performs a predetermined calculation, and replaces the density e of the pixel of interest E with the calculated density d under certain conditions. That is, as shown in FIG.
Set up a matrix consisting of 9 pixels. CPU3
4, the average density d is determined by the arithmetic mean of eight pixels of the matrix excluding the pixel of interest E in accordance with the operating program (FIG. 1, calculation means I46). That is, when the density of each pixel A-I of the matrix is ai, the following calculation is performed: e=(a+b+c+d+f+g+h+i)/8.

The CPU 34 reads the density e of the pixel of interest E from the memory 42, calculates the difference Δe=e−c3 from this average density C in the subtraction means 48, and then calculates the absolute value 1Δe1 of this difference Δe in the absolute value means 50. . This absolute value IΔe1 is compared with a set value Δχ in a comparator 52. This set value Δχ is set by a setter 54.

The average density O and the density e of the pixel of interest E read from the memory 42 are input to the switch means 56. This switch means 56 selects and outputs the average density d when the comparator 52 determines that 1Δe1≧Δχ,
When it is determined that l<Δχ, the density e of the pixel of interest E is selected and output as is. In other words, if the density e of the pixel E of interest is within the range of ±Δχ between the average density d of pixels in a certain area around it, it is assumed that there is no large noise in the pixel E of interest, and the density e is output as is. adopt. However, if the density e is outside the range of density d±Δχ, it is assumed that the pixel E of interest contains noise, and the average density C is used without using the density e. As a result, the output β of the switch means 56 is as shown in FIG. 2B.
Only fine noise is removed, leaving the signal portion unmodified in the output waveform.

The CPU 34 stores this series of calculations in the memory 42.
This is performed sequentially for each pixel of the page, and the output β after this calculation is
Image enhancement processing is performed using (emphasis means 58). This image enhancement is performed by a known method, such as using a second-order differential (Laplacian) of the output β or using a filter that emphasizes high spatial frequencies. As a result, the emphasized output γ becomes as shown in FIG. 2C. This output γ is further binarized at a predetermined binarization level (binarization means 6
0). This binarized signal δ is stored in the memory 42 and output to the printer 38, or stored in an external memory such as the optical disk device 40.

In the above embodiment, the average density d of the pixels other than the pixel of interest, which is the center of the matrix, is used, but depending on the image, the average density including the pixel of interest may be used. The present invention also includes such things.

Note that it is also possible to further develop the present invention and use something other than the average concentration.

For example, instead of this average concentration d, the median of the matrix (
It is also possible to use the average density obtained from only a certain number of pixel densities near the center of the density distribution from the pixel density distribution of the matrix, or to use the root mean square value or standard deviation of the density of each pixel. It is.

(Effects of the Invention) As described above, the present invention calculates the average density of a matrix centered on the pixel of interest, compares it with the density of the pixel of interest, and when the difference between the two is outside a certain range, the density of the pixel of interest is determined. is replaced with this average density, so the rotation is to remove noise from the image without blurring the image or reducing the contrast.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the operation of an embodiment of the present invention, FIG. 2 is an output waveform diagram in each processing process, FIG. 3 is a diagram showing an example of a matrix showing pixel arrangement in a certain area, and FIG. The figure shows an example in which the present invention is applied to a microfilm reader. 30... Line sensor, α... Image signal, E...
Pixel of interest, e...density, e...average density Patent applicant Fuji Photo Film Co., Ltd. Agent Patent attorney Mt.
1) Yu Fumi (1 other person) Figure 1 Figure 2A Figure 2B "su" Figure 4

Claims (2)

[Claims] (1) In an image signal processing method in which the density of each pixel in a certain area centered around the pixel of interest is used to correct the density of the pixel of interest, the average density of each pixel in the certain area is determined, and this average density and the An image characterized in that if the difference from the density of the pixel of interest is outside a certain range, the density of the pixel of interest is replaced by this average density, and if the difference is within a certain range, the density of the pixel of interest is used as is. Signal processing method. (2) The image signal processing method according to claim 1, wherein the average density is an arithmetic mean of pixel densities in a certain area excluding the pixel of interest.