JPH08168004A - Device and method for processing image - Google Patents

Abstract

PURPOSE: To form a high-definition image by deciding a threshold value based on the density value of an extracted formed material picture element and further thresholding the density value based on the threshold value to form an edge suitable for a degraded image. CONSTITUTION: Image data stored in frame memories 103R, 103G and 103B are read into a RAM 107 for each line. Among the data read into the RAM 107, the data for which density difference between adjacent two picture elements is larger than α are detected as inclination presence. When the picture elements provided with this inclination are continued more than β in the same direction, those picture elements are extracted. A threshold value Th1 is calculated from the density value of these extracted picture elements, and the density value is replaced with the threshold value Th1. In this case, the average value of maximum and minimum density value stored in the RAM 107 is defined as the threshold value Th1. When the picture element provided with the density higher than this threshold value Th1 is converted to the maximum density value and the picture element provided with density lower than the threshold value Th1 is converted to the minimum density value, the image can be threshold processed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and a method thereof, and more particularly to an image processing apparatus and a method thereof for forming a high quality image.

[0002]

2. Description of the Related Art In a conventional image processing apparatus, an image signal is input as an analog signal by an image input apparatus such as a scanner or a video input device, converted into a digital signal, and then subjected to various image processing. The image data is stored in the memory according to the resolution and density of the image input device.

[0003]

However, in the conventional image processing apparatus, since random noise is superposed on the analog signal input from the image input apparatus, if the digitalization is performed as it is, the deteriorated image data is deteriorated. Will be stored in memory. Then, when a small image having a small number of pixels is enlarged to a large image such as A4 size, the enlarged image is entirely blurred. Furthermore, depending on the image, the subject may not be distinguishable.

In the conventional image processing apparatus, in order to remove such blurring of an image, edge enhancement processing using a spatial filter is performed. There is a drawback in that the effect of the edge enhancement processing is hardly obtained because the difference in density between them is small.

The present invention has been made in order to eliminate the above-mentioned drawbacks of the prior art. The purpose of the present invention is to form an appropriate edge for an image deteriorated by the influence of noise or the like, and An object of the present invention is to provide an image processing apparatus and a method thereof capable of forming a high-quality image close to the above.

[0006]

The image processing apparatus of the present invention has the following configuration as one means for achieving the above-mentioned object.

That is, a detecting means for detecting a pixel having a density difference of a predetermined value or more with an adjacent pixel as a tilted pixel, an extracting means for extracting a tilted pixel continuous by a predetermined number or more, and the extracting means. And a data conversion unit for binarizing the density value of the extracted tangential pixel based on the threshold value. .

For example, the extraction means is characterized by extracting tilted pixels in which at least three pixels are consecutive.

For example, the threshold value determining means uses the average of the maximum density value and the minimum density value of the tilted pixels extracted by the extracting means as a threshold value.

For example, the data conversion means converts a tilted pixel having a density value larger than the threshold value to the maximum density value, and a tilted pixel having a density value smaller than the threshold value to the minimum density value. It is characterized by being binarized.

Alternatively, the data converting means converts a tilted pixel having a density value larger than the threshold value into a density value obtained by adding a predetermined ratio of a difference between the density value and the maximum density value,
The tilted pixel whose density value is smaller than the threshold value is binarized by converting the density value into a density value obtained by subtracting a predetermined ratio of the difference from the minimum density value.

Further, the threshold value determining means uses the average density value of all tilted pixels extracted by the extracting means as a threshold value.

Further, the threshold value determining means is characterized in that the threshold value is the density value of the pixel located at the center in the array of tilted pixels extracted by the extracting means.

For example, when the number N of tilted pixels extracted by the extracting means is an odd number, the threshold value determining means sets the density value of the (N + 1) / 2th pixel to be the threshold value and is an even number. In this case, the average of the density values of the N / 2th and (N / 2 + 1) th pixels is used as the threshold value.

[0015]

With the above-described structure, it is possible to perform binarization within a range of the inclination for an array of a predetermined number or more of pixels having a predetermined inclination. Therefore, it is possible to properly form an edge even in a portion where a gentle density change continuously occurs.

Further, the density value to be converted can be calculated by the ratio of the original density value and the maximum density value and the minimum density value of the tilted pixel, so that the edges are smoothed and a more natural image is formed. It becomes possible.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described in detail below with reference to the drawings.

<First Embodiment> FIG. 1 is a block diagram showing the arrangement of an image processing apparatus according to this embodiment. In FIG. 1, 101R, 101G, and 101B are image input units to which RGB signals are input. 102R, 102G, 1
Reference numeral 02B is an A / D converter that converts each RGB signal that is an analog signal into an 8-bit digital signal. Reference numerals 103R, 103G, 103B are frame memories for storing one frame of RGB signals converted into digital data by the A / D converters 102R, 102G, 102B, respectively. The frame memory 103R,
103G and 103B can be written and read, and may be of any type as long as the sampling time is high.

Reference numeral 105 denotes a CPU, which will be described in detail later, but performs edge enhancement processing and image processing for improving various image quality in this embodiment based on a program stored in the ROM 106. It should be noted that the CPU 105 transmits the frame memories 103R, 103G,
Address is set to 103B and data is read out. 104
Is an address generator for generating an address for writing the digitally converted image data in the frame memories 103R, 103G, 103B. 106 is RO
M, which stores the image processing program and the like in this embodiment. Reference numeral 107 denotes a RAM, which is used as a work memory for temporarily storing the data in the frame memories 103R, 103G, 103B, the data calculated during image processing, and the like. An output I / F unit 108 outputs the image processing result of the apparatus of this embodiment. Although the output unit 108 is based on Centronics in this embodiment, the output unit 108 is not limited to this. 109 is CPU1
05 to frame memories 103R, 103G, 103B
Address bus line for accessing the data stored in 110, 110 is a data bus line for reading and writing the data of the address designated by the address bus line 109 as described above, and 111 is for outputting the image processing result. It is a bus line.

The image processing in this embodiment is characterized by the conversion processing of digital data. Hereinafter, the conversion process in this embodiment will be described in detail.

In this embodiment, since the data is processed in the scanning direction, the processing is carried out on a line-by-line basis. Therefore, the image data of vertical N × horizontal M pixels (N ≧ 1, M ≧ 3) stored in the frame memories 103R, 103G, and 103B is read into the RAM 107 for each line (M pixels). Then, with respect to the data read in the RAM 107, a pixel whose density difference between adjacent two pixels is α or more is detected as having an inclination, and a pixel having the inclination (hereinafter referred to as an inclined pixel) is β or more in the same direction. When the pixels are continuous, the continuous pixels are extracted. Hereinafter, the density value in this embodiment is 2
The description will be made assuming that 56 gradations are expressed and the respective threshold values α and β are α = 5 and β = 3. Of course, α and β are not limited to this example, and if α ≧ 1 and β ≧ 3, they are within the applicable range of this embodiment.

When the tilted pixels are extracted, the maximum density value and the minimum density value are updated and stored in the RAM 107. Then, an average value of the maximum density value and the minimum density value stored in the RAM 107 is set as a threshold value Th1, and a pixel having a density equal to or higher than the threshold value Th1 becomes the maximum density value and the threshold value T1.
Pixels having densities smaller than h1 are binarized by converting them to the minimum density value.

The data conversion processing in this embodiment described above will be further described with reference to FIG.

In FIG. 2, 201R to 209R are part of the R data obtained by sampling in the scanning direction, and the density values of 201R to 209R are "9".
8 ”,“ 100 ”,“ 105 ”,“ 90 ”,“ 80 ”,
These are "72", "65", "70", and "73".

As a result of performing the conversion process according to the present embodiment on each of these data, five pixels 203R to 207R are extracted as an array of tilted pixels having a density difference of "5" or more. At this time, “105” as the maximum density value and “65” as the minimum density value are stored in the RAM 107, respectively, and the threshold value Th1 for binarization is calculated by the following equation.

Th1 = (maximum density value + minimum density value) / 2
= (105 + 65) / 2 = 85 Therefore, the threshold Th1 is “85”.

As a result, the pixels 203 to 207R are replaced with the pixels indicated by black circles. For example, since the pixel 203R matches the maximum density value "105", the density value does not change and remains "105". Since the pixel 204R is larger than the threshold Th1, the pixel 204R is replaced with 204R 'having a density value of "105". Since the pixel 205R is smaller than the threshold value Th1, the pixel 206R has a density value of “65”, and the pixel 206R
Similarly, R is also replaced with 206R ′ having a density value of “65”. Since the pixel 207R matches the minimum density value "65", the density value remains "65" without conversion.

Therefore, finally, the pixels 204R to 206R indicated by hatching in the drawing disappear by being replaced with 204R 'to 206R'.

In this way, the above processing is repeated for all data for one line in the scanning direction, and when the processing for one line is completed, the conversion result is output to the output I / F unit 108 or the frame memory 103R. , Move to the next line. Then, the same processing is repeated for N (N ≧ 1) lines.

In the above description, R data is taken as an example, but of course the same processing is performed for each of G data and B data. Then, when the processing for all R, G, B is completed, the entire conversion processing in this embodiment is completed.

As described above, according to this embodiment, the density difference between the pixel of interest and the pixel adjacent to it in the scanning direction is α.
When (α ≧ 1) or more, it is detected as a tilted pixel, and a pixel whose tilt in the same direction is continuous for β (β ≧ 3) pixels or more is extracted, and the maximum density value in the extracted tilted pixels is extracted. By using the average of the minimum density values as a threshold value, the pixels whose density is greater than or equal to the threshold value are binarized with the maximum density value, and the pixels with a density less than the threshold value are binarized with the minimum density value. Also, it is possible to obtain a sharp high-quality image with the image blur removed at the edge portion.

<Second Embodiment> The second embodiment according to the present invention will be described below.

The device configuration of the second embodiment is similar to that of the first embodiment shown in FIG.

In the second embodiment, the process up to the determination of the threshold value is the same as in the first embodiment described above, so the explanation is omitted.
In the conversion process in the second embodiment, for a pixel whose density is greater than the threshold value, x% of the difference from the maximum density value
Is added to the density value of the original pixel, and the density value is replaced with that value. On the contrary, for those smaller than the threshold value, a value corresponding to x% of the difference from the minimum density value is subtracted from the density value of the original pixel, and the density value is replaced with the value.

The conversion process in the second embodiment will be described in detail below with reference to FIG. 3, but in FIG.
The description will be made assuming that x described above is 80%.

In FIG. 3, 301R to 309R are part of the R data obtained by sampling in the scanning direction, and the density values of 301R to 309R are "9".
8 ”,“ 100 ”,“ 105 ”,“ 90 ”,“ 80 ”,
These are "72", "65", "70", and "73".

As a result of performing the conversion process according to the second embodiment on each of these data, five pixels 203R to 207R are extracted as an array of tilted pixels having a density difference of "5" or more. At this time, “105” as the maximum value and “65” as the minimum value are stored in the RAM 107, respectively, and the threshold value Th2 for binarization is calculated by the following equation.

Th2 = (maximum density value + minimum density value) / 2
= (105 + 65) / 2 = 85 Therefore, the threshold Th2 is "85".

Therefore, the pixel 303R has a maximum density value of "10".
5 ”, the density value does not change and remains“ 105 ”.

On the other hand, since the density value of the pixel 304R is "90", which is larger than the threshold value Th2, the density value to be converted is calculated according to the following equation.

(Conversion density value) = (original density value) + (maximum density value−original density value) × (x / 100) = 90 + (105−90) × (80/100) = 102 Therefore, the pixel 304R is 304 with a density value of "102"
Is replaced by R '. Further, the density value of the pixel 305R is “80”, which is smaller than the threshold Th2, so the density value to be converted is calculated according to the following equation.

(Conversion density value) = (Original density value) − (Original density value−Minimum density value) × (x / 100) = 80− (80−65) × (80/100) = 68 Therefore, the pixel 305R Is 305R 'with a density value of "68"
Is replaced by Similarly, the pixel 306R is also replaced with 306R 'having a density value of "67" according to the above equation. Further, since the pixel 307R coincides with the minimum density value "65", the density value is not converted and remains "65".

Although the value of x has been described as "80" in the second embodiment, if the value is from "0" to "100", it depends on the characteristics of the image to be processed and the characteristics of the apparatus. , May be set appropriately. For example, since the edge tends to be stronger as the value of x is larger, the value of x may be adjusted depending on the type of image, the magnitude of density change, and the like. For example, in the case of a natural image, by using a value close to “0” as x, the edge can be naturally emphasized while alleviating a sharp change in density at the edge point. In the case of a CG image, by using a value close to “100” as x,
It becomes possible to obtain clearer processing results.

As described above, according to the second embodiment,
In addition to the effect obtained in the first embodiment described above, it is possible to perform optimum image processing on the image to be processed by changing the parameters according to the characteristics of the image to be processed.

<Third Embodiment> The third embodiment of the present invention will be described below.

The device configuration of the third embodiment is similar to that of the above-described first embodiment shown in FIG.

In the data conversion processing in the third embodiment,
It is characterized by the method of determining the threshold value. In the above-described first embodiment, the average of the maximum density value and the minimum density value of the extracted β or more tilted pixels is used as the threshold value, but in the third embodiment, all the tilted pixels similarly extracted. The average density value of is used as the threshold value.

The conversion process in the third embodiment will be described in detail below with reference to FIG. In FIG. 4, 40
1R to 409R are a part of R data obtained by sampling in the scanning direction, and the density values of 401R to 409R are "98", "100", "105", and "9", respectively.
0 "," 80 "," 72 "," 65 "," 70 "," 7 "
3 ”.

As a result of performing the conversion process according to the third embodiment on each of these data, five pixels 403R to 407R are extracted as an array of tilted pixels having a density difference of "5" or more. At this time, “105” as the maximum value and “65” as the minimum value are stored in the RAM 107, respectively, and the threshold value Th3 for binarization is calculated by the following equation.

Th3 = (sum of density values of 403R to 407R) / extracted tilted pixel number = (105 + 90 + 80 + 72 + 65) / 5 = 82 Therefore, the threshold value Th3 is “82”.

Then, data conversion is performed so that pixels having a density value equal to or greater than the threshold value Th3 are replaced with maximum density values and pixels having a density value smaller than the threshold value Th3 are replaced with minimum density values. That is, since the pixel 403R coincides with the maximum density value “105”, the density value does not change and remains “105”. Further, since the density value of the pixel 404R is larger than the threshold value Th3, the density of the pixel 40R is set to 404R ′ of the density value “105” and the pixel 40
Since the density value of 5R is smaller than the threshold value Th3, the density value of 5R is replaced with 405R 'having a density value of "65", and similarly, the pixel 406R is replaced with 406R' having a density value of "65". Further, since the pixel 407R coincides with the minimum density value “65”, the density value is not converted and remains “65”.

As described above, according to the third embodiment,
Since the average density of all the extracted tilted pixels is adopted as the threshold value, the threshold value has flexibility according to the value of the extracted tilted pixel. For example, even if the maximum density value, the minimum density value, and the number of pixels in the extracted tilted pixels are the same, the threshold value can be flexibly changed if the amount of change in density is different. Therefore, flexible conversion processing according to the amount of change in density becomes possible, and more optimal image processing becomes possible.

<Fourth Embodiment> The fourth embodiment according to the present invention will be described below.

Since the device construction of the fourth embodiment is the same as that of the first embodiment shown in FIG. 1, the description thereof will be omitted.

In the data conversion processing in the fourth embodiment,
It is characterized by the method of determining the threshold value. In the above-described first embodiment, the average of the maximum density value and the minimum density value of the extracted β or more tilted pixels is used as the threshold value, but in the fourth embodiment, all the tilted pixels similarly extracted. The density value of the pixel located at the center of the array is set as the threshold value. For example, assuming that the number of extracted tilted pixels is A, if A is an odd number, the density value of the (A + 1) / 2th pixel is, and if it is an even number, the A / 2nd and (A / 2 + 1) th pixels are density values. The average density value of the pixels is used as a threshold.

The conversion process in the fourth embodiment will be described in detail below with reference to FIG. In FIG. 5, 50
1R to 509R are a part of R data obtained by sampling in the scanning direction, and the density values of 501R to 509R are “98”, “100”, “105”, and “9”, respectively.
0 "," 80 "," 72 "," 65 "," 70 "," 7 "
3 ”.

As a result of performing the conversion process according to the fourth embodiment on each of these data, five pixels 503R to 507R are extracted as an array of tilted pixels having a density difference of "5" or more. At this time, “105” as the maximum value and “65” as the minimum value are stored in the RAM 107, respectively, and since the extracted tilted pixel number A is an odd number, the pixel number that becomes the threshold Th4 for binarization Is calculated by the following equation.

(Central pixel number) = (A + 1) / 2 = (5
+1) / 2 = 3 Accordingly, the density value of the third pixel in the extracted tilted pixels, that is, the density value “80” of the pixel 505R becomes the threshold Th4.

Then, data conversion is performed so that pixels having a density value equal to or greater than the threshold Th4 are replaced with maximum density values, and pixels having a density value smaller than the threshold Th4 are replaced with minimum density values. That is, since the pixel 503R coincides with the maximum density value “105”, the density value does not change and remains “105”. Further, since the pixel 504R has a density value larger than the threshold value Th4, the pixel 50
Since 5R is the threshold Th4 or more, the density value is “105”.
Of the pixel 406R is smaller than the threshold value Th4, the pixel 406R is replaced with 506R ′ of which density value is “65”. Further, since the pixel 507R matches the minimum density value "65", the density value is not converted and remains "65".

As described above, according to the fourth embodiment,
An appropriate threshold value can be determined when the data conversion process is performed on an image in which an almost constant density change is continuous.

<Other Embodiments> In the first to fourth embodiments described above, the threshold value α for detecting tilted pixels is used.
Was defined as α = 5, but any value may be used as long as it is a value corresponding to the density distribution of the input image. For example, it is appropriate to reduce the value of α for an image with a small change in density and to increase it for an image with a large change in density.

Similarly, the threshold value β of the number of pixels in which the gradient in the same direction is continuous is set to be 3, and the pixels continuous by β pixels or more are extracted, but 3 ≦ β ≦ M (M: horizontal width size of image) is exceeded. If there is no range, the threshold β can be set appropriately.

Further, in the image data for one frame, it is possible to selectively use the above-described first embodiment and second embodiment for each line, and the above-mentioned first to fourth embodiments. It is also possible to combine and execute a plurality of the embodiments. For example, a good result can be obtained by combining the threshold value setting method in the second embodiment so that the average density value of all extracted pixels in the third embodiment is used.

The image processing apparatus of the present invention can be applied to any apparatus that inputs and outputs image data, such as a copying machine, a printer and a facsimile.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0066]

As described above, according to the present invention, it is possible to perform binarization within a range of the inclination for an array of a predetermined number or more of pixels having a predetermined inclination. Therefore,
It is possible to appropriately form an edge even in a portion where a gentle density change continuously occurs, which has been difficult to form an edge in the past.

Further, the density value to be converted can be calculated by the ratio of the original density value and the maximum density value and the minimum density value of the tilted pixel, so that the edges are smoothed and a more natural image is formed. It becomes possible. Further, the edge strength can be adjusted by adjusting the ratio according to the type of image to be processed, the size of the inclination, etc., and more flexible edge formation is possible.

[0068]

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining a data conversion process in the present embodiment.

FIG. 3 is a diagram for explaining a data conversion process in the second embodiment according to the present invention.

FIG. 4 is a diagram for explaining a data conversion process in a third embodiment according to the present invention.

FIG. 5 is a diagram for explaining a data conversion process in the fourth embodiment according to the present invention.

[Explanation of symbols]

 101R, 101G, 101B Image input section 102R, 102G, 102B A / D converter 103R, 103G, 103B Frame memory 104 Address generation section 105 CPU 106 ROM 107 RAM 108 Output I / F section

Claims (16)

[Claims] 1. A detection unit that detects a pixel having a density difference of a predetermined value or more from an adjacent pixel as a tilted pixel, an extraction unit that extracts a tilted pixel that continues for a predetermined number or more, and an extraction unit that extracts the tilted pixel. And a data conversion unit for binarizing the density value of the extracted tangential pixel based on the threshold value. Image processing device. 2. The image processing apparatus according to claim 1, wherein the extracting unit extracts tilted pixels that are continuous with at least three pixels. 3. The image processing apparatus according to claim 1, wherein the threshold value determining unit uses an average of the maximum density value and the minimum density value of the tilted pixels extracted by the extracting unit as a threshold value. 4. The data conversion means converts a tilted pixel having a density value larger than the threshold value to the maximum density value and a tilted pixel having a density value smaller than the threshold value to the minimum density value. The image processing apparatus according to claim 3, which is binarized. 5. The data conversion means converts a tilted pixel having a density value larger than the threshold value into a density value obtained by adding a predetermined ratio of a difference between the density value and the maximum density value to the density value. 4. The image processing apparatus according to claim 3, wherein the tilted pixels smaller than the threshold value are binarized by converting the density value obtained by subtracting a predetermined ratio of the difference between the density value and the minimum density value from the density value. . 6. The image processing apparatus according to claim 1, wherein the threshold value determining unit sets an average density value of all tilted pixels extracted by the extracting unit as a threshold value. 7. The image processing apparatus according to claim 1, wherein the threshold value determining unit sets the density value of a pixel located at the center in the array of tilted pixels extracted by the extracting unit as a threshold value. 8. The threshold value determining means (N +) when the tilted pixel number N extracted by the extracting means is an odd number.
1) The density value of the second pixel is used as a threshold value, and if it is even, the average of the density values of the N / 2th and (N / 2 + 1) th pixels is used as the threshold value. The image processing device according to claim 7. 9. A pixel having a density difference of a predetermined value or more with an adjacent pixel is detected as a tilted pixel, tilted pixels continuous by a predetermined number or more are extracted, and the density value of the extracted tilted pixel is set. An image processing method, wherein a threshold value is determined based on the threshold value, and the density value of the extracted tangible pixel is binarized based on the threshold value. 10. The image processing method according to claim 9, wherein tilted pixels having at least three consecutive pixels are extracted. 11. The image processing method according to claim 9, wherein an average of the maximum density value and the minimum density value of the extracted tilted pixels is used as a threshold value. 12. Binarization is performed by converting a tilted pixel having a density value larger than the threshold value to the maximum density value and a tilted pixel having a density value smaller than the threshold value to the minimum density value. The image processing method according to claim 11, which is characterized in that. 13. A tilted pixel having a density value larger than the threshold value is converted into a density value obtained by adding a predetermined ratio of a difference between the density value and the maximum density value, and the density value is smaller than the threshold value. 12. The image processing method according to claim 11, wherein the tilted pixel is binarized by converting it into a density value obtained by subtracting a predetermined ratio of the difference between the density value and the minimum density value. 14. The image processing method according to claim 9, wherein an average density value of all the extracted tilted pixels is used as a threshold value. 15. The image processing method according to claim 9, wherein a density value of a pixel located at the center in the array of the extracted tilted pixels is used as a threshold value. 16. When the extracted tilted pixel number N is an odd number, the density value of the pixel located at (N + 1) / 2th is set as a threshold value, and when it is an even number, it is set as N / 2nd ( N /
The image processing method according to claim 15, wherein the average of the density values of the 2 + 1) th pixel is used as the threshold value.