KR20100118262A - Apparatus and method of digital image processing and program recording medium for the same - Google Patents

Abstract

PURPOSE: A digital image processing apparatus and a method thereof using a low-pass filter and a recording medium therefor are provided to obtain the processed image processed in a digital image and reduce the number of filtering by pixel. CONSTITUTION: A pixel value of a first pixel is filtered among the image consisting of a plurality of pixels(S602). The difference of the pixel value of the second pixel and pixel value of the first pixel is obtained and compares with the contrast value(S603). The difference of the pixel value of the pixel value of the second pixel in the return value of the first pixel and the first pixel difference does not bigger than the contrast value is increased and it to the return value of the second pixel(S604).

Description

Apparatus and method of digital image processing and program recording medium for the same

The present invention relates to a digital image processing apparatus and a recording medium therefor, and more particularly, to an image processing apparatus and method capable of reducing a time required for image processing using a low pass filter.

The digital image processing apparatus encompasses a device including a digital signal processing unit (DSP) for processing an electrical image signal, and includes a digital camera, a digital camcorder, a mobile phone incorporating these functions, a personal digital assistant (PDA), The present invention may also be applied to digital devices such as a portable multimedia player (PMP).

During the digital image processing, the original image is filtered to pass through a low pass filter (LPF), a high pass filter (HPF), and a band pass filter to obtain an image suitable for the purpose. . In particular, the low pass filter is a filter that passes the low frequency region and blocks the high frequency region. The low frequency region of the image means a region where brightness difference is small from the surrounding region, and thus, when the original image passes the low pass filter, a smoothing or blur effect occurs, thereby creating a blurry image having a soft feeling.

In some cases, a convolution operation may be used as a method of implementing the low pass filter (LPF), the high pass filter (HPF), the band pass filter, and the like. Filtering through a convolution operation is to perform a convolution operation between pixel values of a plurality of pixels included in a mask and element values of the mask. Such an image processing process is to reduce the size of the original image by convolutional operation and then to enlarge the original image. In this case, there is a problem that a considerable time is required for the interpolation process to enlarge and correct the aspect ratio by replenishing pixels in the step of enlarging an image. In addition, since a pixel value obtained by performing a convolution operation for each pixel included in the mask must be obtained, a long time is required only for the convolution operation in the case of a large image.

Among the low pass filters, the convolution operation is used, such as a mean filter, Gaussian filter, and bilateral filter. A typical low pass filter that does not use the convolution operation is a median filter. (median filter). In the case of the median filter or the bidirectional filter, an edge corresponding to a boundary or contour of an image is preserved in a certain amount, and a portion corresponding to a pixel having a similar brightness value such as a surface is strongly smoothed. However, when using such a filter, if the image is reduced and enlarged, the same result as that of a kind of low pass filter is obtained, which causes blurring of unwanted edges.

Figure 1 shows the result of processing the median filter, Figure 2 shows the result of processing the two-way filter.

In fact, the result of processing the median filter of the original image of FIG. 1 (a) as it is is shown in FIG. 1 (b). 1 (c) is an image. In the same manner, FIG. 2 (a) shows the result of the bidirectional filtering process of the original image as it is, and the original image is reduced in half (1/2), subjected to the bidirectional filter, and then enlarged twice. to be. Comparing Figs. 1 (b) and (c), and Figs. 2 (b) and (c), it can be seen that only the blurring lines of toxic borders are blurred in Fig. 1 (c) and Fig. 2 (c). Ring phenomenon has occurred.

The present invention relates to a digital image processing apparatus and method for solving a problem occurring when using a conventional low pass filter, and a recording medium therefor.

In particular, the present invention provides an image processing apparatus and method for reducing the time required for image processing using a low pass filter and eliminating unwanted blurring.

The digital image processing method according to the present invention comprises the steps of: (a) deriving a result value of the first pixel by filtering the pixel value of the first pixel of the image consisting of a plurality of pixels; (b) obtaining a difference between the pixel value of the second pixel and the pixel value of the first pixel and comparing it with the control value; And (c) adding the difference between the pixel value of the second pixel and the pixel value of the first pixel to the result value of the second pixel when the difference is not greater than the control value. It includes;

The filtering may be performed by a low pass filtering process including a mean filter, a median filter, a Gaussian filter, a bilateral filter, and the like.

In the filtering, the mask may be a 2X2 or 3X3 mask.

Also, the step (c) may include the step of deriving a result value of the second pixel by filtering the pixel value of the second pixel when the difference is greater than the control value; It is characterized by that.

The present invention may be implemented as a recording medium recording a program for executing the digital image processing method on a computer.

In accordance with another aspect of the present invention, a digital image processing apparatus includes: a filter processor configured to filter a pixel value of a first pixel of an image including a plurality of pixels to derive a result value of the first pixel; A calculation unit for obtaining a difference between the pixel value of the second pixel and the pixel value of the first pixel; A comparison unit comparing the difference obtained by the calculation unit with a control value; And a difference between the pixel value of the second pixel and the pixel value of the first pixel by adding the difference value between the pixel value of the first pixel and the result value of the first pixel when the difference is not greater than the control value. Processing result deriving unit; It includes.

In the filter processing unit, the filtering may be performed by low pass filtering, and the filtering may be performed using a 2X2 or 3X3 mask.

The filter processor may be further configured to derive a result value of the second pixel by filtering the pixel value of the second pixel when the difference is greater than or equal to the control value as a result of the comparison.

According to an exemplary embodiment of the present invention, an image reduction and enlargement step for image processing is unnecessary, and the number of filtering for each pixel is reduced, so that a digital image processed image can be obtained in a short time. By eliminating the step, there is a feature that prevents unintended blurring of the edges.

A digital camera having a digital image processing unit of a digital image processing method and apparatus according to the present invention will be described as an embodiment. However, the digital image processing apparatus according to the present invention is not limited to a digital camera, but may be applied to any device capable of digital image processing such as a camera phone, a personal digital assistant (PDA), a portable multimedia player (PMP), a camcorder, and the like.

3 is a block diagram showing a configuration of a digital camera according to an embodiment of the present invention.

Referring to FIG. 3, the digital camera includes an optical unit 11 for inputting an optical signal from a subject, a driver 12 for driving the optical unit 11, and an optical signal input through the optical unit 11. The imaging element 13 which converts into an electrical signal is provided. In addition, the video signal processing is performed on the video signal provided from the analog signal processing unit 14 and the analog signal processing unit 14 which perform signal processing such as noise reduction processing and digital signal conversion processing on the electrical signal provided from the imaging element 13. A digital signal processor (DSP) 20 is provided. The image data may be input to the digital signal processor 20 in real time, but may be temporarily stored in a buffer memory (not shown) and then supplied to the digital signal processor 20 as necessary. The digital camera includes a recording unit 15 storing image data and predetermined information, and a display unit 16 displaying an image. In addition, a storage unit 17 for storing various algorithms and data, and an operation unit 18 for inputting a user's operation signal may be provided. In addition, the digital camera includes a control unit 19 for controlling each component as a whole according to a user's manipulation signal or an input image.

The operation of the DSP 20 will be described in more detail with reference to FIG. 4.

Referring to FIG. 4, the DSP 20 according to an exemplary embodiment of the present invention includes a filter processor 21, a calculator 22, a comparator 23, and a process result derivator 24.

The filter processing unit 21 derives a result value of the arbitrary pixel by filtering the pixel value of an arbitrary pixel of the original image composed of a plurality of pixels input to the DSP. In more detail, a result value of the first pixel, which is a filtering result of the pixel value of the first pixel, may be obtained. Also, a result value of the second pixel obtained by filtering the pixel value of the second pixel may be obtained according to the results of the calculation unit 22 and the comparison unit 23 described later. In this case, the first pixel and the second pixel are arbitrarily designated among several pixels included in the original image, and the pixel value is a brightness value of the designated pixel. The calculation unit 22 will be described first with respect to the method for setting the pixel.

The filter processor 21 filters the pixel value of the first pixel using a low pass filter such as a mean filter, a median filter, a Gaussian filter, a bilateral filter, and the like. To obtain a result value of the first pixel. Since the digital image processing method according to the present invention is for obtaining a smoothed and blurred image, the filter used may be used in addition to the types listed above as long as the low pass filter derives the smoothing and blurring effect. In addition, the present invention can be applied as a means when the high pass filter that has a sharpening effect by emphasizing the outline and boundary of the image is composed of the difference between the original image and the low pass filtered image.

The calculator 22 calculates a difference between the pixel value of the second pixel and the pixel value of the first pixel. Here, the first pixel and the second pixel refer to any pixel set in a mask including a plurality of pixels. The size of the mask is not limited, but the larger the mask, the greater the number of pixels to be processed and the faster the image processing speed according to the present invention, while the image quality is degraded due to excessive smoothing. Therefore, it is necessary to set a range of mask to increase the image processing speed but maintain the image quality at the conventional image processing level. According to an embodiment of the present invention, a mask including a total of 4 pixels or 2X2 or a mask including a total of 9 pixels of 3X3 is preferable.

Once the pixel set as the first pixel is set once in one mask, the pixel is fixed to the first pixel until the digital image processing of the pixels included in the mask is completed. The first pixel may be arbitrarily set among 4 pixels constituting the 2X2 mask or arbitrarily set among 9 pixels constituting the 3X3 mask. Preferably, the first pixel of the 9 pixels constituting the 3X3 mask may be selected. It is good to set as.

The second pixel means a pixel adjacent to the first pixel and is different from the first pixel. In an embodiment, three pixels are referred to as the second pixel in the case of the 2X2 mask, and eight pixels are referred to as the second pixel in the case of the 3X3 mask in which the center pixel is the first pixel.

The comparison unit 23 compares the difference between the pixel value of the second pixel and the pixel value of the first pixel included in the original image, which is the calculation result of the calculation unit 22, with the control value. The control value is a value that can be arbitrarily designated by the user. The control value can be set when the product is released from the manufacturer or can be changed by firmware upgrade. The range to which the digital image processing apparatus and method according to the present invention can be applied is set by the check value. The correlation between the range of the contrast value and the digital image processing result according to the present invention will be described with reference to FIG. 5.

The processing result deriving unit 24 performs an operation when the difference between the pixel value of the second pixel and the pixel value of the first pixel included in the original image is not greater than the control value as a result of the comparison by the comparing unit 23. Specifically, the difference between the pixel value of the second pixel and the pixel value of the first pixel obtained by the calculation unit 22 is added to the result value of the first pixel obtained by the filter processing unit 21 to be the result value of the second pixel. The result of the second pixel can be obtained in a faster time than the image processing by filtering.

In addition, when the difference between the pixel value of the second pixel and the pixel value of the first pixel included in the original image is greater than the reference value as a result of the comparison by the comparator 23, the filter processor 21 may determine the pixel value of the second pixel. Filter by to derive the result of the second pixel. It also includes a case where the difference between the pixel value of the second pixel and the pixel value of the first pixel is equal to the control value.

5 illustrates an image processing process using a convolution operation according to an embodiment of the present invention.

Referring to FIG. 5, F [y] [x], which is a pixel value of a first pixel having a coordinate of [y] [x], is designated among pixels constituting the mask when the 2X2 mask is referred to. Similarly, convolution is performed to obtain R [y] [x], which is the result of the first pixel.

Where y and x are the pixel coordinates of the image, R is the pixel value of the processed image, F is the pixel value of the original image, M is the mask, p is the position of the pixels contained in the mask, and M (p) is at position p The mask value, F (P), means a pixel value of the original image at position p.

Specify the pixel value F [y] [x + 1] of the second pixel with coordinates [y] [x + 1] adjacent to the first pixel with coordinates of [y] [x], and then specify the absolute value of the difference. Find the value.

If this difference is compared with the control value (T) and the difference is not greater than the control value (T), the brightness of the first and second pixels is considered to be similar and a single, rather than edge, such as an outline or boundary in the actual image. It is a part corresponding to the surface of, so it can be blurred. In this case, as shown in Equation 2, the pixel value F [y] [x + 1] of the second pixel and the pixel value of the first pixel in the filtered result value R [y] [x] of the first pixel obtained in Equation 1 The difference of F [y] [x] is added to obtain the result value R [y] [x + 1] of the second pixel which is the processed result.

,

,

When the difference between the pixel value of the second pixel and the pixel value of the first pixel and the contrast value T are greater than or equal to the contrast value T, the pixel value of the first pixel and the pixel of the second pixel are equal. The difference in values is large, which means that the brightness of the pixels differs by more than a certain level. That is, since the image is likely to be a part of the contour of the image, a thin line such as hair, and a boundary part, the blurring of the boundary line is crushed and blurred or the thin line disappears when blurring. Therefore, in this case, it is preferable to express by convolutional processing for each pixel as shown in Equation (3).

The same process is repeated for the plurality of pixels [y + 1] [x] and [y + 1] [x + 1] adjacent to the first pixel among the pixels included in the mask, and the result value R [y + 1] [ x] and R [y + 1] [x + 1]. This process is the same as the process of using the pixel value F [y] [x] of the pixel having the coordinates of [y] [x]. When the digital image processing of the plurality of pixels included in the given mask is completed, another mask may be designated and processed, and finally, the result of digital image processing on the entire original image may be obtained.

In the conventional image processing method using a convolution operation, a result having a low pass filtering effect may be obtained by reducing the image and then performing the convolution operation for each pixel included in the mask and then expanding the image. In this case, the time required for image processing was considerable because the image was reduced and enlarged and the convolution processing was performed for each pixel.

However, the image processing method according to an embodiment of the present invention removes a process of reducing and enlarging an image, performs a convolution operation on a pixel at least once, and then obtains a pixel value difference value of pixels included in a mask, and the difference is not large. It is possible to achieve high speed digital image processing by obtaining a conventional low pass filtering effect by a simple calculation process that adds the difference value.

A digital image processing process using a median filter, which is another embodiment of the present invention, is also similar to the embodiment through the convolution operation. However, based on the 2X2 mask, among the pixels constituting the mask, F [y] [x], which is the pixel value of the first pixel, is arbitrarily designated, and the resultant R [y] [x is obtained by performing median filter processing (Mid). ] Only differs from Equation 4. Therefore, the following process will be omitted.

는 마스크와 동일한 위치의 원본 영상의 픽셀값이며,

은 픽셀에 대한 마스크값이다.Where y and x are the pixel coordinates of the image, R is the pixel value of the processing result,

Is the pixel value of the original image at the same position as the mask.

Is the mask value for the pixel.

6 is a flowchart illustrating an embodiment of a digital image processing method according to the present invention.

First, a first pixel and a second pixel adjacent to the first pixel are arbitrarily designated among pixel values constituting a mask in the original image input to the DSP. (S601) Then, the pixel value of the first pixel is filtered to process the first pixel. (S602) The filtered result may be stored in the internal storage of the DSP or separately in a separate memory.

The difference between the pixel value of the first pixel and the pixel value of the second pixel is compared with the contrast value (S603). The contrast value may be arbitrarily set by a user, and the manufacturer may set a contrast value specified for each image.

   If the difference between the pixel values of the second pixel and the pixel value of the second pixel is not greater than the reference value as a result of the comparison in step S603, that is, if the brightness difference between the first pixel and the second pixel is not large, the first step obtained in step S602 is performed. The result value of the second pixel is obtained by adding the difference of the pixel value of the first pixel from the pixel value of the second pixel to the result value of one pixel. (S604) In this step, the image processing effect through the conventional low pass filter is applied to the pixel. It is a simple and fast implementation in an approximate way that adds the difference between the brightness values.

If the difference between the pixel values of the second pixel and the pixel value of the first pixel is greater than or equal to the control value in step S603, that is, the brightness of the first pixel and the second pixel differs by a predetermined level or more, the same image as in step S603. When the processing is performed, the boundary lines of the image are crushed or blurred, and the thin lines disappear. Therefore, in this case, the pixel value of the second pixel is filtered separately to obtain a result value of the second pixel.

The same method is repeated for the other pixel values included in the mask to obtain the result. After image processing is performed on the pixels included in a given mask, another mask may be designated to perform image processing, and finally, an image processed result of the entire original image may be obtained.

7 to 9 compare the results of the digital image processing method according to the embodiment of the present invention with the conventional image processing results.

7 shows a Gaussian filtering result for pixels included in a 7X7 mask based on the original image (a), (b) shows a conventional Gaussian filtering result, and (c) shows image processing according to an embodiment of the present invention. The result is. Comparing (b) and (c), the result of the processing is similar, but (b) takes 0.047 seconds, while (c) takes 0.015 seconds, and according to the present invention, the image processing is about three times faster. There are advantages to it.

FIG. 8 shows the result of the median filter processing on the pixels included in the 3X3 mask based on the original image (a), (b) shows a result of the conventional median filter processing, and (c) shows one embodiment of the present invention. Image processing result by example. In this case, it can be seen that (b) takes 0.031 seconds (c) takes 0.016 seconds in order to obtain similar image quality processing results, so that the image processing can be performed about twice as fast according to the present invention.

Finally, Figure 9 is a result of the two-way filter processing on the 11X11 mask based on the original image (a) (b) is a conventional two-way filter processing result, (c) is an image processing result according to an embodiment of the present invention to be. Also, for image processing results of similar image quality, (b) takes 0.153 seconds and (c) takes 0.107 seconds faster. From this, as a result of the image processing according to the present invention, it is not necessary to reduce and enlarge an image for conventional image processing, and it is possible to obtain a digital image processed image in a short time since there is no need to perform a calculation by a filter for each pixel. .

As can be seen in Figures 7 to 9, the image processing method according to the present invention is an image processing method using a low pass filter such as a median filter, a Gaussian filter, a bilateral filter, and the like. It is widely available for algorithms. Therefore, the image processing method according to the present invention is particularly useful when a low pass filter should be used in an embedded system which is limited in image processing speed.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential features of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

1 and 2 illustrate image processing results using a median filter and a bidirectional filter according to the related art.

3 is a block diagram showing a configuration of a digital camera as an embodiment of a digital image processing apparatus according to the present invention.

4 is a block diagram showing the configuration of a DSP inside a digital camera according to an embodiment of the present invention.

5 illustrates an image processing process using a convolution operation according to an embodiment of the present invention.

6 is a flowchart illustrating an embodiment of a digital image processing method according to the present invention.

7 to 9 compare the results of the digital image processing method according to the embodiment of the present invention with the conventional image processing results.

Claims (9)

(a) deriving a result value of the first pixel by filtering the pixel value of the first pixel among an image composed of a plurality of pixels; (b) obtaining a difference between the pixel value of the second pixel and the pixel value of the first pixel and comparing it with the control value; And (c) adding the difference between the pixel value of the second pixel and the pixel value of the first pixel to the result value of the second pixel when the difference is not greater than the control value. ; Digital image processing method comprising a. The method of claim 1 And the filtering is low pass filtering. The method of claim 2 The filtering method is a digital image processing method, characterized in that using a 2X2 or 3X3 mask. The method of claim 1 Step (c) is Filtering the pixel value of the second pixel when the difference is greater than or equal to the control value to derive a result value of the second pixel; Digital image processing method characterized in that. A recording medium on which a program for executing the digital image processing method of any one of claims 1 to 4 is recorded on a computer. A filter processor configured to filter a pixel value of the first pixel among the plurality of pixels to derive a result value of the first pixel; A calculation unit for obtaining a difference between the pixel value of the second pixel and the pixel value of the first pixel; A comparison unit comparing the difference obtained by the calculation unit with a control value; And When the difference is not greater than the reference value, the comparison result of the comparison unit adds the difference between the pixel value of the second pixel and the pixel value of the first pixel to the result value of the second pixel. Processing result deriving unit; Digital image processing apparatus comprising a. The method of claim 6 In the filter processing unit And the filtering is low pass filtering. The method of claim 7, In the filter processing unit And the mask is a 2x2 or 3x3 mask. The method of claim 8 The filter processing unit And a result value of the second pixel is derived by filtering the pixel value of the second pixel when the difference is greater than or equal to the comparison value as a result of the comparison.

Images