KR101365828B1 - Apparatus and Method for array-histogram equalization - Google Patents

Abstract

The present invention relates to an alignment histogram equalization apparatus and method that considers brightness values of neighboring pixels in pixel alignment and performs low pass filtering after alignment. The present invention includes a display controller 20 for controlling the brightness of the image output to the display panel 10; And a control unit 30 including an alignment histogram equalizer 32 for equalizing histogram equality of pixel brightness of an input image. The alignment histogram equalizer 32 includes pixel brightness of an input image and a periphery of the pixel. The sorting priority is determined according to the brightness average value of the pixels. According to the present invention as described above, there is an advantage that can be prevented from degradation due to excessive histogram expansion that occurs in the general histogram equalization technique.

Sort Histogram, Mask, Filter, Degradation

Description

Apparatus and Method for array-histogram equalization

1 is a block diagram showing the configuration of a specific embodiment of an alignment histogram equalizer according to the present invention;

2 is a flowchart illustrating a specific embodiment of an alignment histogram equalization method according to the present invention.

3 is a detailed flowchart showing in detail a pixel alignment method and an equalization method in the alignment histogram equalization method according to the present invention;

4A is a table showing an example in which a brightness average value of pixels and brightness before alignment and surrounding pixels of the pixel are calculated.

4B is a table showing an example where brightness is equalized for aligned pixels.

5A is a graph illustrating a general histogram distribution of an input image.

5B is a graph showing the histogram distribution of alignment histogram equalized pixels.

6 is a flow chart showing another embodiment of the alignment histogram equalization method of the present invention.

7 is a detailed flowchart illustrating a low pass filtering method according to another embodiment of the present invention in detail.

8 is a table illustrating an example of calculating a lower limit and an upper limit for pixels aligned for low pass filtering according to another embodiment of the present invention.

9 is a schematic diagram showing an example of a mask used in another embodiment of the present invention.

10A to 10B are exemplary views illustrating examples of input images.

11A to 11B are exemplary views illustrating output images obtained by equalizing the input image of FIGS. 10A and 10B according to a specific embodiment of the present invention.

12A to 12B are exemplary views illustrating output images obtained by equalizing the input images of FIGS. 10A and 10B according to another embodiment of the present invention.

Description of the Related Art [0002]

10 display panel 20 display controller

30 control unit 32 alignment histogram equalizer

34: filter part 40; The storage unit

The present invention relates to an alignment histogram equalization apparatus and method that considers brightness values of neighboring pixels in pixel alignment and performs low pass filtering after alignment.

Various methods for correcting a video image to provide a higher quality image have been recently used. Representative examples include histogram equalization and gamma correction.

However, in the gamma technique, it is necessary to find an optimal gamma (γ) value suitable for an image image. Since the human person directly judges the output image, there is a problem in using it and the accuracy is also changed by individual differences. .

On the other hand, in the histogram equalization technique, the histograms are calculated by reading the brightness of all the pixels of the image, and after equalizing the histograms uniformly, the leveled brightness values are input again to the image. Therefore, there is a fear that excessive expansion of the histogram occurs, and in this case, there is a problem that the image may deteriorate.

Accordingly, the present invention has been made to solve the above-mentioned conventional problems, and an object of the present invention is to obtain a histogram of a constant function type that is more than a histogram equalization technique, in order of pixel brightness and average brightness of pixels. An alignment histogram equalization apparatus and method for changing a brightness value using an alignment coefficient such that the number of pixels having one brightness value in the entire dynamic region after the alignment is the same.

Another object of the present invention is to provide an alignment histogram equalization neglect and method for correcting an equalized image by low pass filtering in order to prevent degradation of the equalized image.

According to a feature of the present invention for achieving the above object, the present invention includes a display controller for controlling the brightness of the image output to the display panel; And a control unit equipped with an alignment histogram equalizer for equalizing histogram equality of pixel brightness of the input image, wherein the alignment histogram equalizer determines alignment priority according to the pixel brightness of the input image and the brightness average value of neighboring pixels of the pixel. do.

In this case, the present invention may further include a storage unit for storing the pixels arranged in the order of the pixel brightness of the input image, and for storing the average value of the brightness of the surrounding pixels for each pixel.

The control unit may include a filter unit that corrects the brightness of the alignment histogram equalized pixel to a brightness average value of surrounding pixels.

In this case, the low pass filtering is performed within a limited range having a lower limit and an upper limit, wherein the lower limit is a minimum equalization brightness that the input brightness of the pixel can have by alignment histogram equalization, and the upper limit is The input brightness of the pixel may be the maximum equalization brightness value that can be obtained by alignment histogram equalization.

On the other hand the present invention (A) receiving an image to be displayed; (B) arranging pixels of the input image according to input brightness values and brightness average values of surrounding pixels; (C) alignment histogram equalization of the aligned pixels; And (D) outputting the equalized image by applying the equalized brightness.

Here, the step (B) is performed by performing a first order alignment according to the input brightness of the image, and second ordering the pixels having the same input brightness based on the brightness average value of neighboring pixels.

In the step (C), the equalized brightness values may be sequentially set by dividing the aligned pixels by S (alignment coefficients).

In this case, the alignment coefficient S may be a value obtained by dividing the total number of pixels by the number of dynamic regions.

And (E) low pass filtering the aligned equalized pixels; The low pass filtering may be to correct the brightness value of the pixel to a brightness average value of neighboring pixels.

The peripheral pixel brightness average may be calculated by a mask of I × I.

In addition, the mask may be weighted according to the distance from the pixel.

And the low pass filtering is performed within a limit range having a lower limit and an upper limit, wherein the lower limit is a minimum equalization brightness value that the input brightness of the pixel can have by alignment histogram equalization, and the upper limit is the input brightness of the pixel. May be the maximum equalization brightness value that can be obtained by the alignment histogram equalization.

In this case, the step (E) may include: determining whether (E1) the peripheral pixel brightness average value is within the alignment histogram equalization range; (E2) correcting pixel brightness to an average value of surrounding pixel brightness when the surrounding pixel brightness average value falls within an alignment histogram equalization range as a result of the determination in the step (E1); (E3) if it is determined in the step (E1) that the peripheral pixel brightness average value is not within the alignment histogram equalization range, the method may include modifying the pixel brightness to a threshold value beyond the average value of the peripheral pixel brightness. .

The low pass filtering may be performed on all pixels of the alignment histogram equalized, or may be performed on pixels whose difference between the input brightness and the equalization brightness among the alignment histogram equalized pixels is a predetermined value or more.

On the other hand, the present invention comprises the steps of (a) receiving an image to be displayed; (b) arranging pixels of the input image according to input brightness values and brightness average values of neighboring pixels; (c) alignment histogram equalization of the aligned pixels; (d) low pass filtering the aligned equalized pixels; And (e) outputting an equalized image by applying the filtered brightness. The low pass filtering includes an alignment histogram equalization method of correcting a brightness value of the pixel to a brightness average value of neighboring pixels.

In this case, the low pass filtering is performed within a range of equalization brightness values that the input brightness of the pixel may have by alignment histogram equalization.

According to the present invention having the configuration as described above, there is an advantage that can be prevented from degradation due to excessive histogram expansion that occurs in the general histogram equalization technique.

Hereinafter, specific embodiments of the alignment histogram equalization apparatus and method according to the present invention as described above will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing the configuration of a specific embodiment of an alignment histogram equalizer according to the present invention.

As shown in the drawing, the histogram equalizer according to the present invention includes a display panel 10 for displaying an equalized video image. The display panel 10 includes an LCD panel, a PDP, and a glass panel of a CRT, and refers to a screen of an image output apparatus divided by pixel.

As described above, the display panel 10 is divided into units of pixels, and each pixel forms a sampling point when each pixel samples a 2D image.

Accordingly, the display panel 10 includes a plurality of rows and a plurality of columns of pixels. In order to clarify the explanation of the present invention, the display panel 10 includes pixels of n 'rows and m' columns. Assume that you do.

The display panel 10 is connected to a display controller 20 that adjusts the brightness of the display panel 10. In this case, the display controller 20 adjusts the brightness by dividing each pixel separately. Therefore, the brightness of the pixel may be known through the display controller 20.

The display controller 20 is connected to a controller 30 which equalizes the brightness of the input image and outputs the same through the display controller 20.

The controller 30 includes an alignment histogram equalizer 32 that equalizes the brightness of the input image.

The alignment histogram equalizer 32 aligns pixels of the input image according to a predetermined rule, and sequentially assigns brightness to the aligned pixels. In this case, the pixel sorting order is arranged in order of pixel brightness of the input projection, and in the case of pixels having the same brightness, the sorting priority is determined according to the brightness average value of the neighboring pixels. The equalization brightness is set to have one brightness value as much as the alignment coefficient S. The alignment coefficient is a value obtained by dividing the number of all pixels of the display panel 10 by the size of the dynamic region. Thus, the same equalization brightness is set for the same number of pixels.

In addition, the control unit 30 further includes a storage unit 40 for storing the aligned pixels. In addition to the aligned pixels, the storage unit 40 may further store the alignment coefficient and the equalized pixel brightness.

Meanwhile, the alignment histogram equalizer according to another embodiment of the present invention may further include a filter unit 34 in the control unit 30.

The filter unit 34 is a part that performs low pass filtering on pixels for which equalization brightness is set. Here, low-pass filtering means replacing the brightness of the pixel with the brightness average value of the surrounding pixels.

However, the low pass filtering according to the present invention has a limited range in replacing the brightness value of the pixel. That is, the brightness of the pixel is substituted within a limited range having a lower limit and an upper limit. In this case, the limit range refers to a range of values that the pixel may have by the above-described alignment histogram equalization, which will be described in detail in describing the operation of the present invention.

The filter unit 34 may be provided with a mask for live low pass filtering, and the mask may be a square mask of I × I. Specific examples of the mask will also be specifically examined in explaining the operation of the present invention.

Hereinafter, the operation of the alignment histogram equalization technique according to the present invention will be described in detail according to the equalization method.

2 is a flowchart illustrating a specific embodiment of an alignment histogram equalization method according to the present invention, and FIG. 3 is a detailed flowchart illustrating a pixel alignment method and an equalization method in detail in the alignment histogram equalization method according to the present invention. FIG. 4A is a table showing an example in which the pixel and brightness before alignment and the brightness average value of the surrounding pixels of the pixel are calculated. FIG. 4B is a table showing an example in which the brightness is equalized for the aligned pixels. 5 is a graph showing a histogram distribution of an input image, and FIG. 5B is a graph showing a histogram distribution of aligned histogram equalized pixels.

As shown in FIG. 2, in order to perform alignment histogram equalization according to a specific embodiment of the present invention, first, an image to be equalized is received (S10).

Pixels of the input image are aligned according to a predetermined rule (S20), which will be described in detail with reference to FIG. 3.

That is, first, the pixels are first sorted in the order of the brightness of the pixels (hereinafter referred to as 'input brightness') (S22). At this time, the sorting is sorted in ascending or descending order according to the input brightness. In the present specification, for convenience of description, the description will be given by way of example in descending order. Therefore, since the display panel 10 has a brightness of 0 to 255, the pixels are divided into 256 grades. Of course, when there are no pixels having a specific input brightness, it may be divided into classes of 256 or less.

Next, pixels having the same input brightness are again sorted in descending order by the brightness average value of neighboring pixels (S24).

Here, the brightness average value of neighboring pixels means an average value of input brightness of pixels in contact with the pixel. It may simply mean an average of eight pixels in contact with the pixel, or it may be extended to mean an average of 24 pixels closest to the pixel (of course, it may further mean an average of adjacent X pixels). ).

For example, when referring to an average of eight pixels, if the target pixel is P (n, m), the peripheral pixel average value is

.

Of course, including the brightness of the target pixel

It can also be set.

4A and 4B show that the input pixels are arranged in the manner as described above.

As shown in FIG. 4A, when pixels having different brightnesses are input, brightness averages of neighboring pixels are calculated for each pixel. In this case, when the pixel is a pixel located at the edge portion, the average is calculated by considering only the pixel in contact with the pixel.

As such, an example in which the brightness average of the surrounding pixels is calculated is shown in FIG. 4A.

The pixels are then rearranged according to the input brightness and the brightness average of the surrounding pixels as aligned in FIG. 4B. In this case, P (1,3) and P (1,1) have the same input brightness value of 25, but are arranged above P (1,3) because the peripheral pixel brightness average values are 20 and 125.

Afterwards, as shown in FIG. 3, equalization brightness is set in units of alignment coefficients at the aligned pixels (S30). The alignment coefficient is a value obtained by dividing the total number of pixels by the number of dynamic regions (and thus having a fixed value according to the display panel 10), so that the number of pixels to be set for each dynamic region is the same. to be.

To this end, first, the brightness coefficient is initialized (S32). The brightness coefficient is a function that is sequentially converted from the initial value to the final value of the dynamic range, and has a value from 0 to 255 in the general display panel 10.

Subsequently, equalization brightness is set to the upper S of the aligned pixels as the brightness coefficient value (S34).

In addition, since the brightness coefficient is increased, the dynamic range is increased by '1', and the increase of the brightness coefficient is 1 (S36).

Thereafter, it is determined whether the brightness coefficient is greater than or equal to the maximum value (S38). In this case, the maximum value means the final value of the dynamic region.

As described above, an example in which equalization brightness values are specified for all pixels is shown in Fig. 4B.

As such, when the equalization brightness is set, the controller 30 outputs the equalized image to the display panel 10 through the display controller 20.

As such, the alignment histogram equalization means that the histogram of the equalized pixel brightness is equalized to form a constant function, as shown in FIG. 5.

5A illustrates a histogram of a general input image. 5b shows a histogram of an alignment histogram leveled image according to the present invention.

As shown, it can be seen that an input image having an uneven histogram distribution is uniformly distributed over the entire dynamic region by alignment histogram equalization.

On the other hand, to look at the alignment histogram equalization method according to another embodiment of the present invention in detail.

6 is a flowchart illustrating another embodiment of an alignment histogram equalization method according to the present invention, FIG. 7 is a detailed flowchart illustrating a low pass filtering method according to another embodiment of the present invention, and FIG. 9 is a table illustrating an example of calculating a lower limit and an upper limit of pixels aligned for low pass filtering, and FIG. 9 is a schematic diagram illustrating an example of a mask used in another embodiment of the present invention.

In another embodiment of the present invention, the image is received in the same manner as in the specific embodiment (S100), the pixels are aligned (S200), and then the alignment histogram equalization (S300) is performed. However, in another exemplary embodiment of the present invention, low pass filtering is performed to smoothly process the alignment histogram equalized image (S400).

In the low pass filtering, the brightness of the pixel is replaced by the brightness average value of the neighboring pixels. The low pass filtering according to the present invention has a limited range in replacing the brightness value of the pixel.

The theoretical background of the low pass filtering technique is the application of the POSC technique. In order to use the POSC technique, all features of the original image must be represented by a closed convex set. At this time, two convex set conditions are used. The first is to project using a low pass filter, and the second is to use boundary conditions. Hereinafter, a specific processing method will be described below.

Specifically, as shown in FIG. 7, first, it is determined whether the average value of the surrounding pixel brightness falls within the limited range (S410).

Here, the limit range refers to a range determined by the minimum and maximum values that the input brightness value may have.

To this end, when the alignment histogram equalization according to another embodiment of the present invention, the lower limit value and the upper limit value are calculated. As shown in FIG. 8, the lower limit value and the upper limit value are calculated after setting equalization brightness to the pixel.

In the case of P (1,4), since the input brightness is 1 and the equalization brightness value of the input brightness is only 0, both the lower limit and the upper limit are zero.

However, in the case of P (1,3) and P (1,4), the input brightness is 25, and the equality brightness values of the input brightness 25 are 0 and 1, so the lower limit is 0 and the upper limit is 1.

As another example, if the alignment coefficient is 100 and the number of pixels having an input brightness value of 0 is 310, the equalization brightness value has a value from 0 to 4. In this case, the lower limit is 0 and the upper limit is 4.

As a result of the determination in step 410, if the average brightness of the surrounding pixels is within the limit, the equalization brightness of the pixels is set as the brightness average of the surrounding pixels (S420).

However, as a result of the determination in step 410, when the average value of the surrounding pixel brightness is outside the limit, the limit value outside the average value of the surrounding pixel brightness is set to the equalization brightness of the pixel (S430).

In this way, all of the n'xm 'pixels are low pass filtered repeatedly. Of course, in order to shorten the processing time, it is also possible to compare the input brightness and the equalization brightness among the pixels so as to perform the low pass filtering only in the case of a pixel having a high brightness difference (only when a certain value is greater than a predetermined value).

The filtering may be performed using a mask of I × I. The mask is preferably a square mask. When the I value is relatively large, the mask may be weighted differently according to the distance to the target pixel to calculate the average.

An example of such a mask is shown in FIG. 9.

In Fig. 9 a 3x3 mask is shown as an example.

In this case, if the target pixel to be filtered is P (n, m), the pixels adjacent to the target pixel are P (n-1, m-1), P (n-1, m),... , P (n + 1.m + 1)

Hereinafter, an example in which an input image is aligned histogram or the like according to another embodiment of the present invention will be described with the accompanying drawings.

10A to 10B are exemplary views illustrating an example of an input image, and FIGS. 11A to 11B are exemplary views illustrating an output image obtained by equalizing the input image of FIGS. 10A and 10B according to a specific embodiment of the present invention. 12A to 12B are exemplary views illustrating output images obtained by equalizing the input images of FIGS. 10A and 10B according to another embodiment of the present invention.

The image shown in FIG. 10A and an enlarged portion of the image shown in FIG. 10A show that the background image is not clear and difficult to distinguish because the brightness of the surrounding background is too large as the input original image.

In contrast, the images shown in FIGS. 11A and 11B are images obtained by performing alignment histogram equalization according to a specific embodiment of the present invention, and it can be seen that objects are clearly distinguished from the images of FIG. 10.

12A and 12B are images obtained by alignment histogram equalization of the images of FIG. 10 according to another embodiment of the present invention, and represent images closer to the original images than those of FIG. 11.

That is, when comparing 11b and 12b, it can be seen that in the case of 11b, the brightness is excessively dispersed from FIG. 10b so that the original image and the heterogeneity are generated and blurring occurs at the interface. However, in the case of FIG. 12B in which such excessive brightness change is corrected, an image closer to the original image is generated and the boundary surface becomes clear.

The rights of the present invention are not limited to the embodiments described above, but are defined by the claims, and various changes and modifications can be made by those skilled in the art within the scope of the claims. It is self-evident.

In the apparatus and method for alignment histogram equalization according to the present invention as described in detail above, the following effects can be expected.

That is, in using the histogram equalization technique, the input brightness and the average brightness value of the peripheral pixels are used for the alignment order of the pixels. Accordingly, there is an advantage in that pixel brightness may be equalized in consideration of harmony with surrounding pixels.

Accordingly, in the alignment histogram equalizing apparatus and method according to the present invention, there is a long way to prevent the image from being excessively deteriorated due to a breakdown of the balance with surrounding pixels.

In addition, by performing low pass filtering, the present invention can not only improve image quality but also prevent degradation of image quality due to excessive histogram expansion because it is matched with surrounding pixels within the range allowed by the alignment histogram equalization technique. In addition, the contrast of the background part of the resultant image is also improved.

Claims (17)

A display controller controlling brightness of an image output to the display panel; An alignment histogram equalizer for equalizing histogram equalization of pixel brightness of an input image, and a control unit having low pass filtering for correcting the brightness of the alignment histogram equalized pixel with brightness average values of surrounding pixels. Are configured by: The alignment histogram equalizer, And an order of priority of the alignment according to the pixel brightness of the input image and the brightness average value of the surrounding pixels of the pixel. The method of claim 1, And a storage unit for storing the pixels arranged in the order of pixel brightness of the input image, and storing the brightness average value of the surrounding pixels for each pixel. delete 3. The method according to claim 1 or 2, The low pass filtering is performed within a limited range having a lower limit and an upper limit. Wherein the lower limit value is a minimum equalization brightness value that the input brightness of the pixel can have by alignment histogram equalization, and the upper limit value is an alignment histogram characterized in that the input brightness of the pixel is the maximum equality brightness value that can be obtained by alignment histogram equalization. Equalizer. (A) receiving an image to be displayed; (B) arranging pixels of the input image according to input brightness values and brightness average values of surrounding pixels; (C) alignment histogram equalization of the aligned pixels; And (D) outputting the equalized image by applying the equalized brightness, Step (B) is, And performing first order alignment according to the input brightness of the image, and performing second order alignment of pixels having the same input brightness based on brightness average values of neighboring pixels. delete 6. The method of claim 5, Step (C) is, An alignment histogram equalization method characterized by dividing the sorted pixels by S (alignment coefficients) and sequentially setting equalization brightness values. The method of claim 7, wherein The alignment coefficient S is An alignment histogram equalization method characterized by dividing the total number of pixels by the number of dynamic regions. The method according to any one of claims 5 , 7, and 8 (E) low pass filtering the aligned equalized pixels; The low pass filtering, And correcting the brightness value of the pixel to a brightness average value of neighboring pixels. The method of claim 9, The peripheral pixel brightness average is An alignment histogram equalization method, characterized in that it is calculated by a mask of IxI. 11. The method of claim 10, In the mask, And a weighted value according to a distance from the pixel. The method of claim 9, The low pass filtering, Is carried out within the limits having a lower limit and an upper limit, Wherein the lower limit value is a minimum equalization brightness value that the input brightness of the pixel can have by alignment histogram equalization, and the upper limit value is an alignment histogram characterized in that the input brightness of the pixel is the maximum equality brightness value that can be obtained by alignment histogram equalization. Equalization method. The method of claim 9, The (E) step, (E1) determining whether the surrounding pixel brightness average value falls within the alignment histogram equalization range; (E2) correcting pixel brightness to an average value of peripheral pixel brightness when the peripheral pixel brightness average value is within the alignment histogram equalization range as a result of the determination in the step (E1); (E3) if the average value of the surrounding pixel brightness is not within the alignment histogram equalization range as a result of the determination in the step (E1), modifying the pixel brightness to a threshold value beyond the average value of the surrounding pixel brightness. Sort histogram equalization method. 13. The method of claim 12, And said low pass filtering is performed on every pixel of the alignment histogram equalized. 13. The method of claim 12, And the low pass filtering is performed on pixels whose difference between input brightness and equalization brightness among the alignment histogram equalized pixels is greater than or equal to a predetermined value. (a) receiving an image to be displayed; (b) arranging pixels of the input image according to input brightness values and brightness average values of neighboring pixels; (c) alignment histogram equalization of the aligned pixels; (d) low pass filtering the aligned equalized pixels; And (e) applying the filtered brightness to output an equalized image; The low pass filtering, And correcting the brightness value of the pixel to a brightness average value of neighboring pixels. 17. The method of claim 16, The low pass filtering, And the input brightness of the pixel is performed within a range of equalization brightness values that can be obtained by alignment histogram equalization.

Images