JPH0273475A - Method and device for processing blurred picture - Google Patents

Abstract

PURPOSE:To obtain the pictures free from any blur without deteriorating the picture quality even in the case an image having movements is used at an optional part of a screen by providing a picture quality improvement processing part and plural frame memories, etc. CONSTITUTION:The analog signals received from a TV 1, etc., are turned into digital signals via a RGB decoder 7 and an A/D converter 8 and inputted to a frame memory 9-1 or 9-2 of a frame memory 9 that is not used for the subsequent processes. The input picture signals stored in the memory 9 are used at a picture quality improvement processing part 13 for correction of the blurred pictures, etc. In other words, the picture elements equivalent to the slope part of a picture signal are eliminated and at the same time the picture quality is improved in detection of the signal level difference larger than the threshold value as well as in the detection and the correction of the blurred parts, etc. The processed picture signals are stored in a RAM 15 via a CPU 14 and stored successively in a picture memory 18 after undergoing an interpolation process as necessary. Thus the pictures free from any blur can be obtained without deteriorating the picture quality.

Description

[Detailed description of the invention] Industry and academia - 1-9 - Field of use - The present invention provides a hard copy of arbitrarily selected screens from continuous images of television broadcasting or recorded on video tape, etc. We will present a new method for correcting blur (moving parts) that occurs in still images when producing high-resolution images necessary for creating printing films and printing plates. It is something.

In recent years, due to the practical use of satellite broadcasting, advances in television technology centered on high-definition, and changes in publishing trends associated with the development of video technology and the diversification of audio and printed matter, television broadcasting and If a predetermined screen obtained from continuous video information on a video tape or the like can be used for photographs, manuscripts for printing, etc., various advantages can be expected.

For example, in the conventional method of shooting film using a still camera, photographers had to go to the location and actually take pictures, and after developing the film,
Since the photos are sent by various means of electronic transmission, they lack the ability to provide prompt reporting.

In particular, if it were possible to immediately obtain a manuscript for printing from a TV broadcast screen such as a news program or live sports broadcast while at the printing site, it would be easier and faster to obtain news photos for newspapers, etc., and it would greatly improve the production of newspapers. You can expect an increase in speed.

In addition, when shooting with a still camera, there are photo opportunities and other constraints, but when recording TV broadcasts on video tape, etc., you can freely select the screen you want, which makes it even better. It is also possible to obtain photos.

In addition to the press, we also provide hard copies and printing manuscripts such as direct photographs from television broadcast images, television images recorded on video tapes, etc., or images obtained from movie films via telecine equipment. If we can create one, we can expect the development of various needs.

However, when extracting still images from continuous images displayed on a television, there are the following problems.

That is, the television broadcasting currently commonly performed in Japan is called the NTSC system, and one screen is formed by 525 scanning lines, regardless of the size of the television image. One screen of television images is composed of two-field images, and the 525 television signals are divided into two and interlaced. Therefore, in order to obtain a hard copy etc. from one screen of image information, it is necessary to sequentially combine the scanning line information of the first field (odd field) and the scanning line information of the second field (even field). becomes. However, the first
There is a time lag between the image information of the field and the image information of the second field, so if there is a moving part on the screen, blur (movement) will occur in the still image. .

A simulation of this blurring situation is shown in Figure 1.

゛Figure 1 shows a partial image of one screen of a television, consisting of a moving part (Y) and a still part (X). Then, the screen status when the image is imported as odd field scanning line information is represented by a1. , even field scanning line information is indicated by d.

Since there is a slight time lag (t) when taking in information between the odd and even fields, simply combining the scanning line information of the odd and even fields results in the result shown by C. Originally, an image such as that shown by r is required, but if you try to obtain a still image from a moving screen, the screen will be blurred.

If the entire screen moves uniformly, blurring can be prevented by, for example, moving only the scanning line information of the odd field by the time lag and combining it with the scanning line information of the even field. However, in normal TV images, the entire screen rarely moves uniformly, and there are irregular movements in various parts of the screen, so the above-mentioned time shift compensation + E alone cannot correct blurred images. It is not possible.

Conventionally, as a method for correcting blur, one screen of a television is used with only scanning line information of one field (only odd or even numbers), and this information is inserted between the scanning lines to form one screen. things were being done. However, in this case, the screen becomes distorted or blurred, making it impossible to obtain the desired clear image.

Another method is to use either the odd or even field scan line information, rather than simply inserting it, and use the information corresponding to the other field to create the information in the other field based on the information in one field. This was done by creating a new screen using all of the information through interpolation. but,
In this case, there is less distortion and blur compared to the above-mentioned method, but the image cannot be clear.

The present invention is a method for obtaining a blur-free image without deteriorating the image quality even when using a video with movement in any part of one screen. This is what we are trying to provide.

In order to achieve the purpose of this article, the blurred image processing method according to the present invention digitizes the image signal of an image configured in an interlaced format, and a first step of importing the image signal level of each pixel into a frame memory using the captured image as a source image; A second step in which the image signal level of each pixel in the other field is compared with the image signal level of each pixel at the corresponding position in the other field to be overlapped, and the pixels corresponding to the slope portion are removed. , Find the difference in the image signal level of the pixel at the position corresponding to the pixel in the other field corresponding to each pixel in the field, and calculate the difference in the image signal level of the pixel at the position corresponding to the pixel in the other field, and calculate the difference in the image signal level of the pixel where the magnitude of the difference in level exceeds a predetermined threshold. The third step is to detect the position of the original image captured in the frame memory. and the image signal level of the surrounding pixels of each pixel of the blurred portion in the original image,
A fifth step of determining a new image signal level and replacing it with the image signal level of each pixel in the -F blurred portion.

Here, the image signal processing method according to the present invention will be explained in more detail.

First, as the image signal of the original image (original image) to which the image signal processing method according to the present invention is applied, a TV signal from a television, a VTR signal from a VTR, a VDR signal from a laser disc, etc. can be used. . Furthermore, a telecine signal obtained from a film or the like via a telecine device can also be used. (Hereinafter, these signals will be collectively referred to as TV.
It is written as a signal. ) Since these TV signals are analog signals, they are quantized by an A/D converter and stored in the frame memory as signals with addresses. On the other hand, a pre-digitized signal, such as a digital signal from an optical disk or computer graphics (CG), can also be used as an input signal and is stored in the frame memory without going through an A/D converter. In addition, if the TV signal is a color image signal, an RGB decoder is used to convert R, G,
After being separated into B signals, each is converted into a digital signal by an A/D converter, and each of the R, G, and B signals is stored in each frame memory.

As shown below, in the case of an input image signal such as a TV signal, each address (e.g., 512 x 512) (corresponding to a pixel) of this frame memory has, for example, 8 bits (
Therefore, it is stored as a density value of 256 levels).

Next, the blurred image processing method according to the present invention will be explained with reference to the simulated image of a moving object shown in FIG.
This will be explained in more detail.

Figure 2a shows the state of the portion of the image signal stored in the frame memory from the jth line of the scanning line (however, the jth line is an odd-numbered line) to the (j++o)th line. There is. In this figure, each square represents one pixel, and each pixel in the shaded area belongs to a moving object and has a non-zero (for example, positive) image signal level.
Each pixel in the white portion belongs to the background and has an image signal level of O.

Here, a method for extracting a moving part of the object will be explained. First, the image signals of each pixel in each row (hereinafter simply referred to as the odd-numbered row) corresponding to the scanning line of the odd-numbered field (first field) stored in the frame memory shown in FIG. It is assumed that the signal is an image (hereinafter referred to as the basic image) of a motionless (still) part of the image. Then, for the portion represented by each pixel in the odd-numbered row, each row (hereinafter simply referred to as the even-numbered row) corresponding to the scanning line of the even-numbered field (second field) shown in FIG. Check whether the image signal of the pixel is moving.

(It should be noted that the original image of the even field shown in FIG. 2c includes pixels in a portion that does not move relative to the basic image and pixels in a portion that moves.) The level of the image signal of each pixel in the th row, and the level of the image signal of the pixels in the even-numbered rows adjacent thereto (for example, the (++1)th row and the (i + 1)th row),
Compare and find out the difference. Then, each pixel in the even-numbered row where the level difference is larger than a predetermined threshold is checked to see if it is a moving part (i.e., a blurred part of the image) with respect to the basic image of the object, and then the image signal is sequentially determined. It is determined that this is a portion where the level changes (slope portion) or may be an edge portion. A pixel in which the level difference is smaller than a predetermined threshold is regarded as not a moving part but a part in which the image signal of the pixel of the basic image is continuously changing. For example, when representing an image signal level with 8-bit resolution (ie, 256 steps), the size of the threshold value is set to 32 in advance. As a specific operation, first, for the first pixel in the (++2)th row shown in FIG. 2, the corresponding first pixel in the (++1)th row shown in FIG. 2C, Compare image signal levels. in this case,
It is assumed that the magnitude of the difference between the image signal levels of one pixel is smaller than the predetermined threshold magnitude 32. Furthermore, Part (j) 2)
The image signal levels of the (i+1)th and (++2)th pixels in the row are compared with the corresponding (j+1)th and (++2)th pixels in the (++1)th row. In these cases, it is also assumed that the magnitude of the difference in the image signal level of the corresponding one pixel is smaller than the predetermined threshold value 32.Furthermore, as shown in FIG.
Regarding the (++3)th pixel in the ++2)th row, the corresponding (i+)th pixel in the (++1)th row shown in FIG.
3) Compare the image signal level of the th pixel. Furthermore, in a similar manner, each pixel after the (i+-4)th row in the (++2)th row shown in FIG. Compare signal levels. In this way, the image signal level of each pixel in each odd-numbered row shown in FIG. 2C is similarly compared with the corresponding pixel in the even-numbered row shown in FIG. 2C. As a result of such image signal level comparison, each pixel in the even field where the difference in image signal level from the corresponding pixel in the odd field is larger than a predetermined threshold is indicated by a △ mark or an X mark. is shown in Figure 2d. Here, the △ mark indicates an even field whose image signal level is lower than the image signal level of the corresponding odd field pixel, and the X mark indicates an even field whose image signal level is lower than the image signal level of the corresponding odd field pixel. Each pixel in an even field whose image signal level is higher is shown. Furthermore, the pixels in the portion where the dotted pattern is applied are pixels in an even field belonging to the object, and the difference in image signal level from the corresponding pixel in an odd field is smaller than a predetermined threshold. Shows pixels. Therefore, Fig. 2d
Pixels indicated by a Δ mark or an X mark are blurred in the image; they indicate the 1st part (moving part), the slope part, or the even part.

Here, first, a method for distinguishing between a blur portion and a [1-brake portion] will be explained.

I-Each 1. f In comparing the image signal levels of corresponding pixels, if the magnitude (absolute value) of the difference in image signal level is larger than a predetermined threshold value, the part of the level difference caused by movement and the signal level are This includes a portion (slope; 11) that changes sequentially.

When comparing the image signal levels of consecutive pixels between the blurred portion and the sloped portion, the following relationship exists.

That is, the image signal level D(j) of the i-th pixel in the odd field (for example, the first row) and the image signal level D of the first pixel in the even field (the (j + I)th row). (j1) and the next odd field (j+2)
)-th row, i-th pixel image signal level D (j 4
-2), the image signal level becomes uneven in the blurred portion and stepped in the sloped portion. Therefore, the level difference between D(j) and D(j+2) with respect to the image signal level of D(j+I), that is, d i r r (j+l, D = D (j+1
) −D (D.

d i r r (j+I, j+2) = D (
j+1) -D (j'2), in the blurred part d i f r (j+1. D .

If d i f f (j+l, j+2) is either positive or negative, in the slope part, if one is positive, the other is negative. Therefore, the level difference between the two levels diff(
j41. j) and d i f f (j+1.3+2)
It is possible to find the relationship between the two and distinguish the part where it is a positive value as a blurred part.

For example, the pixel indicated by O in FIG. 2d can be regarded as a pixel corresponding to a slope portion rather than a blur portion. Therefore, it can be regarded as a pixel corresponding to a slope portion.

Therefore, the signal level of the even field corresponding to the slope portion is excluded from the processing of the blurred image.

Next, since an edge portion is generally a portion of an image where the level of the image signal (density level) changes more rapidly than the surrounding area, it does not appear continuously over a plurality of scanning lines (a plurality of rows).

On the other hand, a blurred portion (a moving portion) is a line that appears as a continuous block across multiple scanning lines (multiple rows). Therefore, a portion where the difference in image signal level appearing across three consecutive rows of scanning lines has a large rise above a predetermined threshold can be regarded as a blurred portion (portion with movement). . For example, the part indicated by the bold line in Figure 2d is the blurred part (
It can be regarded as a moving part).

As shown in E below, either the image signal information of the even field or the image signal information of the odd field is fixed,
By comparison, it is possible to extract a portion where the signal level difference is greater than a certain threshold value, and to detect the portion where this appears over multiple rows of the scanning line as a blurred portion (portion with movement). .

Next, the image signal of the original image (basic image) of the odd field (first field) and the image signal of the original image of the even field (second field) are converted into the moving part of the even field (second field). The pixels of the image from which the image signals of the corresponding pixels have been removed are combined to form a second original image of the object (FIG. 2C). In the second original image of the object obtained in this way, the image signal level of the pixel at the position indicated by an By performing interpolation processing on the image signal level of a pixel based on the image signal values of surrounding pixels, the image signal level of a moving portion can be corrected, and thereby a target image can be obtained.

The part marked with a circle in FIG. 2f corresponds to the pixel for which the image signal level is newly interpolated.

In the above description, the scanning line information for odd fields is fixed, but it is also possible to fix the scanning line information for even fields.

In addition, as a threshold value for calculating the image signal level difference l: between corresponding pixels of odd and even fields and extracting edge parts or blurred parts (portions with movement), for example, in the case of It is necessary that the image signal level is 32 with an 8-bit (256 step) resolution and that the number of image signal levels changes depending on the number of image signal levels of the entire image. Normally, it is desirable to set the threshold value to a value in the range of about I/I 0-115 of the total number of signal levels (stages). When the level is set to 1/1O or higher, even the density change part naturally becomes a blurred part, and conversely, when the level is set to 115 or higher, it becomes difficult to extract moving parts.

In addition, regarding the number of scanning lines to distinguish between an edge part and a moving part, it is preferable to define a blurred part (a moving part) when the number of scanning lines is 3 or more. .

In the case where the image spans two lines, there is a possibility that even the edge portion may be treated as a blurred portion (a moving portion).

Note that as an interpolation method for forming a new image signal for a pixel with no signal, the conventionally known nearest key method (N+method) is used.
), Bi-1inear method
d), Cubic convolution method (Cubic
convolution method) etc. can be used. In addition, the patent application filed in 1986-
It is also possible to use the linear extrapolation averaging method disclosed in Japanese Patent Application No. 286351, Japanese Patent Application No. 61-286352, and Japanese Patent Application No. 62-2+7419. In particular, in the case of the present invention, since pixels in one field arranged every other row are interpolated, it is desirable to use a linear extrapolation averaging method that can enhance the signal.

It is preferable that the crossing occurs in a blurred part (a moving part).

In the case of two edges, there is a possibility that even one edge may be treated as a blurred portion (a moving portion).

Note that as an interpolation method for forming a new image signal for a pixel with no signal, the conventionally known nearest neighbor method is used.
, Bi linear method
), cubic composition method (Cubic
convolution method) etc. can be used. In addition, the patent application No. 1986-2 related to the proposal of the applicant of the present application
It is also possible to use the linear extrapolation averaging method disclosed in Japanese Patent Application No. 86351, Japanese Patent Application No. 61-286352, and Japanese Patent Application No. 62-217419. In particular, in the case of the present invention, since the pixels of one field arranged every other row are interpolated, it is desirable to use a linear extrapolation averaging method that enables signal enhancement.

In the following, an image signal processing device that is a preferred embodiment of the present invention will be described in more detail.

Figure 3 is a diagram illustrating an image signal processing method according to the present invention.

As an input image signal, TV signal i from a television;
(+), VTR signal (2) from the VTR.

VDR signals (3) from laser discs etc. can be used. Furthermore, a telecine signal (4) obtained from a film or the like via a telecine device can also be used.

Hereinafter, these signals will be collectively referred to as TV signals. These TV signals are analog signals L-me, A
/D conversion '5 (8) and is quantized and stored in the frame memory (9) as a signal with an address. On the other hand, a signal that has been digitized in advance, such as an optical disc (
5) Or Combi Coater Graphics CCG) (6)
It can be used with 12 input signals such as digital signals from A
The data is stored in the frame memory (9) without passing through the /D converter (8). The digitized human image signal is stored in one of the frame memories 9-1 and 9-2 that is not used for subsequent processing, for example, in frame memories 9-1 in FIG. 3. Incidentally, this is performed by the Fure Gobu (10). A D/A converter (11) and a monitor display (I2) are devices for projecting signals from the frame memory (9) onto the display 1- as necessary.

Next, the human image signal stored in the frame memory (9-2) is supplied to the image quality processing section (13). At this time, the other frame memory (9-1) is connected to the control unit (10).
The next manual image signal is stored according to the instructions from . In this way, frame memories (9-1) and (9-2)
It is used after being switched to σ using the method described above. In addition,
When it is necessary to store many image signals in parallel,
You can prepare as many frame memories as you need.

In addition, if the TV signal is color image signal No. 1, after being separated into R, G, and H signals by the RGB decoder (7),
Each is converted into a digital signal by an A/D converter (8), and each of the R, G, and B signals is stored in each frame memory.

As mentioned above, in the case of input image signals such as TV signals,
Each address (corresponding to a pixel) (for example, 512×512) of this frame memory is stored as a density value of, for example, 8 bits (therefore, 256 levels).

The human image signal stored in the frame memory (9) is then subjected to necessary image quality improvement processing in the image quality improvement processing section (13). In addition, this image quality improvement processing section (
In step 13), the blurred image (correction) process is performed.

In this image quality processing ff1 (13), in addition to performing the above-mentioned processing on the image signal read out from the frame memory (9), processing such as aspect ratio correction and gradation correction is performed as necessary. The subsequent image signal is sent to the CPU (
+4) and stored in the RAM (15). At this time, the image signal stored in the RAM (15) can be stored in an image file (6) on a floppy disk, hard disk, etc., or on a CRT display (17) as necessary. It can also be displayed and confirmed.

The image signal read from the rtAM (15) or image file (16) and stored in the RAM (15) in this way can be subjected to interpolation processing to obtain a high-resolution image if necessary. The image signals of new pixels obtained by the inter-hli calculation are sequentially stored in the image & memory (18).

The image data stored in the image memory (I8) is output by the output device (20) via the interface (19).
For example, it will be output on photographic film 14, photographic paper, or the like. Although not shown in FIG. 4, the output device means a laser beam prober or the like, and it is also possible to use the output section of a scanner. In addition, if the output device is equipped with a buffer memory, it is possible to temporarily store image data in it and then output it.Also, if a halftone image is required for printing, etc., the image data can be halftone It is also possible to output the data after applying it.

4(A), (B), and (C) are flowcharts illustrating the flow of procedural operations when implementing the image signal processing method according to the present invention using the apparatus shown in FIG. 3. .

First, initialize the image size (number of pixels) and threshold data of the original digital image (original image) (Step S+)
. That is, data on the number of pixels in the horizontal and vertical directions of the original image are set to variables x and y, respectively, and data on the threshold value described below is set to the variable thrO. Then, the data of the image signal level of the original image stored in the frame memory is read into the array d (step S2). In the following, for the elements of the array d, for example, the data of the image signal level of the i-th pixel in the j-th row will be expressed as di,j. Furthermore, the value of the image signal level of the corresponding element of d is set as the value of each element of the array d' that stores the corrected image data (step S3).

Next, the values of control variables (counters) i and j are set to 1 (step S4).

Here, first, the image signals of the i-th pixel in the j-th row of the original image, the corresponding i-th pixel in the (j+1)-th row, and further the i-th pixel in the corresponding (j+2)-th row The following calculation is performed from the data (step S5).

dirr21 = di, j+1 di, j ... (ri1fr23 1ne21 di, ji-1 di, jto2 ... (2) 1rr21 l1nc23 = difr23sign
=diff21 * diff23... (5) Next, determine whether the value of this sign is positive or not, and 1i
Respective values of ne21 and l 1ne23 and threshold th
A comparison is made with the value of re (step 6). If the value of sign is positive and each value of 1ine21 and l1ne23 is larger than the threshold value thre, that is, if there is a possibility of a blurred part (a part with movement) or an edge part, The value of the variable r lagf (i) is set to the value of difr21 described in (step S7). In other cases, the value of the variable r 1aKl (i) is set to O (step S8). Next, step S9) increases the value of the control variable i by 1, and further increases the value of the control variable i.
It is checked whether the value of is greater than the number of pixels in the horizontal direction of the original image of variable X (step 510).

If the value of the control variable i is smaller than the number of pixels in the horizontal direction of the original image, the above steps S5 to SIO are further performed for the next pixel in the j-th row of the original image, and the number of pixels in the j-th row of the original image is Similar processing is repeated until all pixels of jM< are processed. If the value of the control variable i is greater than the number of pixels in the horizontal direction of the original image, proceed to step S11, increment the control variable J by 1, and perform the same steps as steps S5 to SIO for each pixel in the next row of the original image. Process. That is,
First, in step SI2, the control variable i is set to 1 to start processing for the first pixel, in step 13, the above equations (1) to (5) are calculated, and then in step I4, step 6 is performed. Perform similar comparison processing. here,
The value of sign is positive and one 1ine21 and l1
If each value of nc23 is larger than the threshold value thre, that is, if there is a +-4 probability of a blurred part or an edge part, then the value of the variable flag2(i) is set to Assign the value of dirr21 (step 5I5), otherwise set the variable flag
The value of 2(i) is set to 0 (step 516). Next, step 5I7) increases the value of the control variable i by 1, and the control a
It is checked whether the value of the variance i is greater than the number of horizontal pixels of the original image of the variable X (step 818), and if the value of the control variable i is less than the number of horizontal pixels of the original image, the original image is further Steps S13 to S16 described above are performed for the next pixel in the current j-th row of the original image, and the same process is repeated until all pixels in the current j-th row of the original image are processed. If the value of the control variable i is greater than the number of pixels in the horizontal direction of the original image, the process proceeds to the following. First, in step S19, the control variable i is set to 2, and then in step S20
The value of control variable j' is set to (j-1). Furthermore, in step s2I, 2r lagl (i+I)tr l
Whether the value of ag2 (i) is positive or not, 2f l a tr,
I (i −1) * r la<2 (+) Compare whether the value is positive or not. And flagl(i+1)*f
lag2(i)>0... (6) or rlagl(i-1)*flag2(i)>0...
In the case of (7), it is recognized as a blurred part that appears as a lump over three scanning lines, and in step 22, the image signal level of the first pixel in the j'th row of the above corrected image is calculated. Replace the values d' i, j' with the values obtained by the following formula. However, the above item (6)
If neither the condition in Equation nor the condition in Equation (7) above are satisfied, it is assumed that there is no blurring and step 2 is performed.
2 is not modified by equation (8) (N branch of step 21). Then, the process proceeds to step S23, where the value of the @J variable i is incremented by I, and whether the current value of the control variable i obtained thereby is greater than (the number of pixels in the horizontal direction of the original image - I), that is, (X1). Compare (step 524).

If it is larger, that is, if the necessary correction processing has been completed for all pixels in the j'th row (excluding both ends), the processing from step S25 to step 928 is further performed to set the value of each element of the variable array flagl. , is replaced with the value of the corresponding element of array flag2.

Furthermore, in step S29, it is checked whether the value of the control variable j is larger than the value of (the number of pixels in the vertical direction of the original image - I), which is exactly y - 1), and if it is not larger, the process returns to step S11 and the next The above step S12 for the (j+1)th row
Processing from step S29 is performed. and step 29
Then, the above step S is repeated until the value of the control variable j becomes larger than the value of (y-1), that is, for all rows of the original image.
The processes from step S12 to step S29 are repeated.

Note that in the above embodiment, when pixels in an even field whose image signal level difference with corresponding pixels in an odd field is equal to or greater than a threshold appear as a cluster over three consecutive scanning lines, it is considered as a blurred part. However, it is also possible to capture a blurred portion if it appears as a lump over a predetermined number of scanning lines depending on the magnification of the original image, etc.

As explained above, according to the method of the present invention, a screen randomly selected from a continuous video such as a television signal being broadcast or recorded on a videotape, etc. It is possible to obtain a blur-free image from an image that has a blurred portion (a moving portion) without degrading the image quality.

[Brief explanation of the drawing]

Figures 1 (a) to (f) are diagrams showing examples of images from the back of one television screen, consisting of a moving part and a still part of an object, and Figure 2 (a) to (r) are
FIG. 3 is a schematic block diagram showing an apparatus for carrying out the image processing method according to the present invention; FIG. Figures (A), (n
) and (C) are flowcharts showing the flow of procedural operations for carrying out the method for correcting blurred images according to the present invention. In addition, in the drawings, 7...RG B decoder, 8...Δ/D converter, 9
...Frame memory, 10...Control unit, 11...
D/A converter, 12...Monitor display, 13
...Image quality processing unit, 14...CPU115...
RAM, 16...Image file, I7...CR
T, 18... Image memory, 19... Interface, 20... Output device. N school Figure 4 (A) Figure 4 (8)

Claims (1)

[Claims] 1) A first step of digitizing the image signal of an image configured in an interlaced format and importing the image signal level of each pixel into a frame memory using the digitized image as an original image; The image signal of either the odd field or the even field captured in the frame memory is fixed, and the image signal level of each pixel of the other field is adjusted to the pixel level of each pixel of the corresponding position of the other field. Compare the image signal level of
The second step is to remove the pixels corresponding to the slope part, and calculate the difference in the image signal level of the pixel at the position corresponding to the pixel in the other field corresponding to each pixel in the one field, and calculate the difference in level. a third step of detecting the position of a pixel whose size is larger than a predetermined threshold; and a third step of detecting the position of a pixel whose size is larger than a predetermined threshold; A fourth step is to detect a portion that appears over the entire area as a blurred portion, and a new image signal level is obtained using the image signal level of the surrounding pixels of each pixel of the blurred portion in the original image, and the image signal level of the blurred portion is determined. A method for processing a blurred image, comprising: a fifth step of replacing the image signal level of each pixel. 2) In the method according to claim 1, when the image signals are differentiated into 0-255 stages during the digitization, the threshold value for the difference in image signal level is 30 or more, and the method is for detecting the blurred portion. The blurred image processing method described above, wherein the predetermined number of scanning lines is three or more. 3) A first storage unit that stores the image signal of the digital image after selecting the image signal of the desired digital image as the original image, reading it from the medium, and performing A/D conversion as necessary. An input unit reads the image signal stored in the first storage unit, performs image quality improvement processing on the image by the blurred image processing method according to claim 1, and produces an image signal subjected to the image quality improvement processing. and an output unit for outputting the image signal stored in the second storage unit. Device.