JP2004102904A - Filtering device and filtering method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform filtering at high speed in digital image processing. <P>SOLUTION: The number of pixels of an original image 41 is reduced to obtain a reduced image 42. The number of pixels is reduced by a bilinear system or a bi-cubic system, for instance. The pixel data of the reduced image 42 is filtered (by low-pass filter processing, for instance) to obtain the pixel data of a filtered image 43. The number of pixels of the filtered image 43 is restored into the number of pixels of the original image 41 to obtain a restored image 44. Restoration of the number of pixels is performed by, for example, interpolating the image data. The image data of the restored image 44 is further filtered to obtain the filtered image data 45. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to digital image processing.
[0002]
[Prior art]
Conventionally, in digital filter processing, a filter having a relatively large size (for example, a 30 × 30 matrix) is used in order to make the filter effect remarkable. However, if filter processing is performed using such a filter, the amount of calculation is enormous, which causes a reduction in image processing speed (for example, see Patent Document 1).
[0003]
On the other hand, if a filter having a small size (for example, a 3 × 3 matrix) is used to increase the image processing speed, the amount of calculation is reduced, but the reference area of the filter is reduced, and the filter effect, for example, the soft focus effect is sufficiently exhibited I can't let it.
[0004]
[Patent Document 1]
JP-A-9-251532
[Problems to be solved by the invention]
The present invention has been made in view of the above problems, and has as its object to perform a filtering process so that an image processing speed is high and a filtering effect can be sufficiently exhibited.
[0006]
[Means for Solving the Problems]
A filter processing device according to the present invention includes: an image reduction unit configured to obtain a reduced image by reducing the number of pixels of an original image; a filter processing unit configured to filter pixel data forming the reduced image to obtain filtered image data; A pixel number restoring unit for restoring the number of pixels of the filtered image corresponding to the processed image data to the number of pixels of the original image to obtain a restored image. This makes it possible to obtain an image with a high filtering effect by high-speed processing.
[0007]
The image reducing unit divides the original image into a plurality of regions and selects one pixel from each of the regions, or divides the original image into a plurality of regions and calculates an average value of a plurality of pixel data included in each of the regions. It is preferable to obtain a reduced image by any of the above. On the other hand, it is preferable that the pixel number restoring unit obtains a restored image by interpolating the filtered image data.
[0008]
It is preferable that the pixel data forming the restored image is further filtered to remove noise. Preferably, the filtering is a low-pass filtering. Thereby, an image with a high blurring effect can be obtained at high speed. Further, it is preferable that the number of pixels of the reduced image can be set in a plurality of stages. Thereby, it is possible to adjust the filter effect according to the purpose.
[0009]
A filtering method according to the present invention includes an image reducing step of reducing the number of pixels of an original image to obtain a reduced image; a filtering step of filtering pixel data forming the reduced image to obtain filtered image data; A pixel number restoring step of restoring the number of pixels of the filtered image corresponding to the processed image data to the number of pixels of the original image to obtain a restored image.
[0010]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
[0011]
FIG. 1 is a flowchart of a filtering process according to an embodiment of the present invention. In the present embodiment, the filtering is a low-pass filtering. The original image 41 is an image corresponding to image data obtained by a digital camera, for example.
[0012]
In the present embodiment, the original image 41 is an image of 2048 × 1536 pixels, and is converted into a reduced image 42 of, for example, 320 × 240 pixels by reducing the number of pixels. Here, the number of pixels is reduced, for example, by a bilinear method, that is, by dividing the original image 41 into a plurality of regions and setting an average value of a plurality of pixel data included in each region as one pixel data of the reduced image. Done. The reduction in the number of pixels is performed, for example, by dividing the original image 41 into a plurality of regions, selecting one pixel from each of the regions, and making one pixel of the reduced image. When one pixel is selected from each of the plurality of regions to obtain one pixel of the reduced image, interpolation may be performed based on information on surrounding pixels, which is a so-called bicubic method.
[0013]
A known 3 × 3 matrix low-pass filter process is performed on each pixel data of the reduced image 42 a predetermined number of times, and filtered image data that is pixel data of the filtered image 43 is obtained. The filtered image 43 is a blurred image of 320 × 240 pixels whose contour is not clear.
[0014]
Next, the number of pixels of the filtered image 43 is restored to the number of pixels of the original image 41, and the filtered image 43 becomes a restored image 44 of 2048 × 1536 pixels. Here, the restored image 44 is obtained, for example, by interpolating the filtered image data. Thereafter, the pixel data of the restored image 44 is further subjected to a low-pass filtering process a predetermined number of times to remove noise, and processed image data 45 is obtained.
[0015]
Here, the low-pass filter processing is performed after the number of pixels of the image is reduced, as described below, even in the case of using an identical matrix (for example, 3 × 3), in an image having a large number of pixels. This is because an image can be blurred more effectively than performing low-pass filtering.
[0016]
One pixel 47 in the reduced image 42 in which the number of pixels is reduced corresponds to an area 46 of several tens of pixels in the luminance image 41 before the number of pixels is reduced. Therefore, when the low-pass filter is applied after the number of pixels is reduced, the reference area for the 3 × 3 matrix becomes very large as compared with the case where the number of pixels is not reduced. Therefore, even if a low-pass filter is applied to a matrix having a small number of reference pixels, such as a 3 × 3 matrix, a low-pass filter is applied to a matrix having a very large reference pixel (for example, a matrix of several tens × several tens). The same blurring effect as at the time is obtained. Also, since the number of reference pixels in the matrix is small, the processing speed is very high.
[0017]
FIG. 2 shows a flowchart of the filtering process. When the filtering process is started, in step 300, the number of pixels in the reduced image 42 is determined. Here, in the present embodiment, the number of pixels in the reduced image 42 is determined by being selected from a plurality of preset stages. For example, the preset number of pixels in a plurality of stages is 640 × 480, 480 × 320, and 320 × 240. In step 310, the number of pixels of the original image 41 is reduced to the number of pixels determined in step 300, and the reduced image 42 is obtained.
[0018]
Next, in step 320, the pixel data constituting the reduced image 42 is subjected to a low-pass filtering process a predetermined number of times to obtain a filtered image 43. In step 330, the number of pixels of the filtered image 43 is restored to the same number of pixels as the original image, and a restored image 44 is obtained. In step 340, the restored image 44 is further subjected to a low-pass filtering process a predetermined number of times. With these processes, the filtering process ends, and processed image data 45 having a high blurring effect is obtained.
[0019]
As described above, in the filter processing according to the present embodiment, the number of pixels of the image is once reduced, and then the low-pass filter processing is performed on the image data. It is possible.
[0020]
In the present embodiment, the filtering process is a low-pass filtering process, but may be performed by another filtering process according to the purpose.
[0021]
【The invention's effect】
As described above, according to the present invention, since the filtering process is performed after the number of pixels of the original image is reduced in the filtering process, an image having a high filtering effect can be obtained at a high processing speed.
[Brief description of the drawings]
FIG. 1 is a flowchart of a filter process according to an embodiment of the present invention.
FIG. 2 is a flowchart illustrating a processing routine of a filter process according to an embodiment of the present invention.
[Explanation of symbols]
41 Original image 42 Reduced image 43 Filtered image 44 Restored image 45 Processed image data

Claims (7)

Image reduction means for reducing the number of pixels of an original image to obtain a reduced image, filter processing means for filtering pixel data constituting the reduced image to obtain filtered image data, and a filter corresponding to the filtered image data A filter processing apparatus comprising: a pixel number restoring unit that restores the number of pixels of a processed image to the number of pixels of the original image to obtain a restored image. The image reduction unit divides the original image into a plurality of regions and selects one pixel from each of the regions, or divides the original image into a plurality of regions and a plurality of pixel data included in each of the regions. 2. The filter processing apparatus according to claim 1, wherein the reduced image is obtained by taking one of the following average values. The filter processing apparatus according to claim 1, wherein the pixel number restoring unit obtains the restored image by interpolating the filtered image data. 2. The filtering device according to claim 1, wherein the filtering process is further performed on the pixel data forming the restored image. The filtering device according to claim 1, wherein the filtering process is a low-pass filtering process. The filter processing device according to claim 1, wherein the number of pixels of the reduced image can be set in a plurality of stages. An image reduction step of reducing the number of pixels of an original image to obtain a reduced image, a filtering step of filtering pixel data forming the reduced image to obtain filtered image data, and a filter corresponding to the filtered image data A pixel number restoring step of restoring the number of pixels of the processed image to the number of pixels of the original image to obtain a restored image.

Images