JP2000293682A - Picture processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain picture processing such as picture synthesis with high quality by automatically operating proper low pass filter processing even to a face whose reduction rate is changing without making it necessary to preliminarily operate low pass filter processing to a texture picture. SOLUTION: In a picture processor 14 which operates picture processing for adhering a texture picture to one part of an original picture inputted from a picture inputting device 10, a calculating means which calculates a distance on the texture picture coordinates corresponding to a distance between the adjacent points among points on the two-dimensional coordinates of the adhesion region to which the texture picture is adhered, a filter parameter selecting means which automatically selects a low pass filter parameter for each point based on the calculated distance, and a picture processing means which adheres the texture picture filter-processed by using the selected low pass filter parameter to the adhesion region are realized by software.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is applied to an image processing apparatus, particularly to image synthesis for synthesizing a texture image with a part of an original image such as a real image or CG image, and to three-dimensional image processing for mapping a texture image. And a suitable image processing apparatus.

[0002]

2. Description of the Related Art In an image synthesizing apparatus for synthesizing a texture image with a part (pasting area) of an original image such as a real image or a CG image, the original texture image has a higher resolution than the synthesized texture image. In this case (in the case where the texture is reduced and synthesized as compared with the original texture), moire occurs when the texture image is synthesized as it is. Therefore, generally, in order to suppress the occurrence of moiré, a low-pass filter process or the like is performed on a texture image in advance, and synthesis is performed using a texture image in which high-frequency components of the texture image are cut and blurred. For the same reason, in a three-dimensional image processing apparatus that performs mapping of a texture image, texture mapping is performed using a texture image that has been subjected to a low-pass filter process or the like in advance.

[0003]

However, as described above, when combining or mapping an image on which a texture image has been subjected to low-pass filtering in advance,
Although it is possible to suppress the occurrence of moire, there are the following problems in obtaining a high-quality image.

[0004] For example, if moire is observed while looking at an actually synthesized image, the band limit is stronger (on the high frequency side).
If a low-pass filter is selected, or if the texture is too blurry, the low-pass filter must be selected by trial and error, for example, by selecting a low-pass filter with a weak band limit.

[0005] In addition, when a texture is combined with an original image by projective transformation or combination, or when texture mapping is performed, the reduction ratio of the texture is changed (for a distance of one pixel on the combined image, Since the distance on the texture coordinates varies depending on the place to be synthesized as shown in FIG. 8 and the like described later, in a single low-pass filter processing, the reduction Moiré occurs in a place where the reduction ratio is large, and conversely, if the filter parameter is matched to a place where the reduction ratio is large, the place where the reduction ratio is small may be too blurred.

If the low-pass filter processing on the texture image is the same in both the X and Y directions, for example, when the texture is synthesized on the side wall when the building is viewed from an angle close to the front, the X and Y directions are different. When the reduction ratio in the Y direction is significantly different, moiré occurs when the reduction ratio is set to a smaller value, and the texture pattern may be blurred when the reduction ratio is set to a smaller value.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned conventional problems, and it is not necessary to perform a low-pass filter process on a texture image in advance. An object of the present invention is to provide an image processing apparatus capable of automatically performing appropriate low-pass filter processing even on a surface having a significantly different rate and performing image processing such as high-quality image synthesis and texture mapping.

[0008]

According to the present invention, there is provided an image processing apparatus for performing image processing for pasting a texture image to a part of an original image. Calculating means for calculating the distance on the texture image coordinates, corresponding to the distance between adjacent points, and filter parameter selecting means for automatically selecting a low-pass filter parameter for each point based on the calculated distance, The object is achieved by providing a configuration including an image processing unit that pastes a texture image filtered using the selected low-pass filter parameter into the paste region.

That is, in the present invention, a corresponding low-pass filter is selected using the low-pass filter parameter reflecting the distance between adjacent points on the original image, that is, the distance on the texture image corresponding to the unit distance. Filter processing, it is possible to perform an appropriate low-pass filter processing for each point (pixel) of the pasting area according to the reduction ratio when the texture image is pasted. Image processing can be performed.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an image synthesizing apparatus according to an embodiment of the present invention will be described in detail with reference to the drawings.
The image synthesizing apparatus of the present embodiment has already been disclosed by
As a basic function, it has the same function as that of the image synthesizing apparatus which performs projection normalization conversion and synthesizes the image as proposed in -274649.

FIG. 1 is a block diagram showing a schematic configuration of an image synthesizing apparatus according to an embodiment of the present invention. The image synthesizing device of the present embodiment includes an image input device 10, an instruction input device 1,
2, an image processing device 14, a display device 16, and an image output device 18.

[0012] The image input device 10 is used for real images and CG images.
An image scanner for inputting an original image such as an image, a texture image to be synthesized (pasted) onto the image, and the like, and a storage device such as a hard disk capable of storing and reading these image data are configured. The instruction input device 12 includes a keyboard, a mouse, and the like, and can input various information.

The image processing device 14 comprises a computer. The device 14 has a distance (basic distance) between adjacent points for a point (pixel) on a two-dimensional coordinate of a synthesis area where a texture image is synthesized. Calculating means for calculating a distance on the texture image coordinates, a filter parameter selecting means for automatically selecting a low-pass filter parameter for each point based on the calculated distance, and a selected low-pass filter parameter Each of the synthesizing means for synthesizing the texture image subjected to the filtering process with the synthesizing area in the synthesizing area is realized by software.

Next, the operation of this embodiment will be described with reference to the flowchart shown in FIG.

First, an original image (FIG. 3) and a texture image (FIG. 4) are input by the image input device 10 (steps 1 and 2). Here, for convenience, the texture image is shown in an enlarged manner.

Next, a synthesis area (pasting area) for synthesizing the texture image in the original image is designated by the instruction input device 12 (step 3). Here, 4 shown in FIG.
The case where the squares P1 to P4 are designated is shown. Then, for the rectangular area of the texture image to be synthesized with the synthesis area, the upper left coordinate value corresponding to P1 of the synthesis area and the width and height of the rectangle are similarly specified. Here, it is assumed that all the regions of the texture image shown in FIG. 4 have been designated.

Next, the following processing is performed in the image processing device 14. First, the coordinate values (x, y) of the original image are subjected to projective normalization transformation for all the pixels in the synthesis area designated by the original image (for this coordinate transformation, for example, “computer geometry and geographic information processing”). bit separate volume Kyoritsu Publishing (198
6) is performed according to the following equations (3) and (4) including c and d defined by the following equations (1) and (2), and the obtained normalized coordinate values α and β are used. , The following (5), (6)
Conversion to coordinate values (x _t , y _t ) on the texture image is performed by the equation (step 4).

[0018]

(Equation 1) Here, x, y: x for combining region, y coordinates x _t, y _t: coordinate x1~x4 on transformed texture image, y1 to y4: x quadrangle vertices of synthesis region, y
Coordinates x ₀ ′, y ₀ ′: x, y coordinates of the upper left corner of the square of the texture to be synthesized Wt, Ht: width and height of the square of the texture to be synthesized α: x coordinate after projective normalization transformation β: Y coordinate after projective normalization transformation

Next, for all pixels in the synthesis area,
The distance (pixel unit) on the texture image corresponding to the distance between adjacent points (pixels) on the original image is calculated separately in the x and y directions on the texture image (step 5). FIG.
9 shows an example of calculating a distance on a texture image corresponding to a unit distance between a certain point of interest P (x, y) in a synthesis area and an adjacent point. A point P on the original image shown in FIG. 6A and four points adjacent to the point P in the upper, lower, left and right directions are shown in FIG.
Assuming that the distance dx in the x direction and the distance dy in the y direction are half of the length of the arrow shown in FIG. The average distance between adjacent points. The fact that the distance in the x direction is dx means that when the texture image is synthesized in the synthesis area, the texture image is reduced to 1 / dx in the same direction.

Next, for all the pixels in the synthesis area,
Using the distance calculated as described above, appropriate filter parameters are separately selected in the x and y directions (step 6). Here, to explain the appropriate filter parameters,
Assuming that the distance dx calculated for a certain pixel P (x, y) is 5, the band limitation in the x direction is 1 /
Select a low pass filter parameter designed to be 5. Here, with respect to the design of the filter parameters, a filter design method disclosed in Japanese Patent Application Laid-Open No. H10-187965 can be used. According to this, it is possible to design a filter parameter that can perform arbitrary band limitation. Actually the bandwidth limit is 1 / 2,1 / 3,1 / 4,1 / 5 ... 1
/ 10,1 / 12,1 / 14,1 / 16,1 / 18,1 / 20,1 / 22,1 / 24,1 / 27,1 / 30,1
Filter parameters such as / 35, 1/40, 1/45, 1/50 ... are prepared separately, and based on the calculated distance dx, a filter parameter close to it is automatically selected by the program. It has become. At this time, it is better that the number of filters to be prepared is large. However, if the number of filters to be selected is large, the synthesis time becomes long.

Next, filter processing is performed using the filter parameters selected as described above, and processing for synthesizing a texture image with the synthesis area is performed (step 7,
8).

Here, an example of the filtering process will be specifically described. If the bandwidth of the pixel P (x, y) is actually limited to ２, the filter parameter is set to c.
And _-10 to c _10, the pixel P (x, y) into RGB color data of R
(0,0), G (0,0), B (0,0) and the neighboring point P '(x + n, y + m)
Is represented by R (n, m), G (n, m), and B (n, m), and the color data of P (x, y) after filtering is represented by fR, fG, fB. Can be calculated by the following formula (7). Further, fG and fB can be similarly calculated. In this example, the color data is stored in RGB,
The same calculation can be performed for other color data such as YK and Lab.

[0023]

(Equation 2)

The procedure of the filtering process executed in step 7 will be described below with reference to FIG. Here, for convenience of explanation, among the prepared filter parameters, filter parameters actually selected are f1, f2, f
Suppose there were five of 3, f4, f5. Also, x,
If the combination of the filter parameters in the y direction is expressed as (f1, f2) when the x direction is f1 and the y direction is f2, it is assumed that there are 11 filter combinations in the whole pixel as shown below. I do.

(F1, f1) (f1, f2) (f1, f3) (f2, f1) (f2, f2) (f2, f3) (f2, f4) (f3, f2) (f3, f3) (f3 , f4) (f3, f5)

Since the low-pass filter can be processed independently in each of the x and y directions, the upper combination (f1,
It is assumed that filter processing is performed in order from f1) to the right. First,
Filter processing is performed on the texture image T in FIG. 7A in the x direction using the filter parameter f1 to create an image T 'shown in FIG. 7B. Next, a filtering process is performed on this image T 'in the y direction using the filter parameter f1 to create an image T1' shown in FIG. Then, the image T1 'is synthesized with the pixels of the pixels of the synthesis area for which the combination of the filter parameters (f1, f1) is selected.

Next, similarly, for the image T ', y
Filter processing is performed using the other filter parameters f2 and f3 in the direction, and the image T shown in FIGS.
2 'and T3' are created, and synthesis is performed on each of the pixels in the pixels of the synthesis area for which the combination of the filter parameters (f1, f2) and (f1, f3) is selected.

Hereinafter, although not shown, similarly, for each of the images that have been subjected to the filter processing using the filter parameters f2 and f3 in the x direction of the texture image, y is obtained by using the filter parameters having a combination. Filter processing is performed in the direction, and each created image is combined with each pixel in the combination area in which the combination of the corresponding filter parameters is selected.

FIG. 8 is an enlarged view of an image obtained as a result of synthesizing the texture image of FIG. 4 with the synthesizing area of FIG. For comparison, FIG. 9 shows an image synthesized without performing the low-pass filter processing, and FIGS. 10, 11, and 12 show images synthesized by performing a low-pass filter processing with a different band limitation on the texture image in advance.

In FIG. 9, moiré occurs because the low-pass filter processing is not performed. FIG. 10 is an image obtained by synthesizing a texture obtained by performing filter processing by selecting a filter parameter so as to be appropriate in the leftmost synthesis area, but moire occurs on the right side. FIG. 12 is an image obtained by synthesizing a texture obtained by performing a filter process by selecting a filter parameter so as to be appropriate in the rightmost synthesis area. Therefore, moire does not occur. Is quite blurred. FIG. 11 shows an image obtained by synthesizing a texture image obtained by performing a filtering process by selecting a filter parameter approximately in the middle between FIG. 10 and FIG. Then the texture pattern is blurred. However, the difference is not clear in the attached drawing read by the scanner because the resolution is low.

Compared to the respective images shown in FIGS.
In the image of FIG. 8 synthesized using the image processing apparatus of the present embodiment, an image is obtained in which no moiré is generated and the texture pattern is not blurred and crushed even in the left area. As described above, according to the present embodiment, a high-quality composite image can be easily obtained.

Although the present invention has been specifically described above, the present invention is not limited to the above-described embodiment, and can be variously modified without departing from the gist thereof.

For example, in the above embodiment, the image synthesizing apparatus has been described. However, it is needless to say that the present invention is not limited to this, and may be a three-dimensional image processing apparatus for mapping a texture image to an original image.

In the above-described embodiment, the image subjected to the filtering process is created by using the selected filter parameters, and the image is synthesized. However, it is not always necessary to create the texture image subjected to the filtering process. At the time of image synthesis, necessary pixels on the texture image may be subjected to filter processing for each pixel using filter parameters corresponding to each pixel, and then synthesized.

[0035]

As described above, according to the present invention, it is not necessary to perform a low-pass filter process on a texture image in advance, and the present invention can be applied to a surface where the reduction ratio changes or a surface where the reduction ratio in the xy direction greatly differs. Also, appropriate low-pass filter processing is automatically performed, and image processing such as high-quality image synthesis and texture mapping can be performed.

[Brief description of the drawings]

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

FIG. 2 is a flowchart showing the operation of the embodiment;

FIG. 3 is an explanatory diagram showing an example of an original image.

FIG. 4 is an explanatory diagram showing an example of a texture image.

FIG. 5 is an explanatory diagram showing the original image in which a combination area is specified;

FIG. 6 is a diagram illustrating a principle of calculating a distance on a texture image corresponding to a distance between adjacent points of a synthesis area.

FIG. 7 is an explanatory diagram showing characteristics of a low-pass filter process according to the present invention.

FIG. 8 is an explanatory diagram showing an example of a composite image according to the present invention.

FIG. 9 is an explanatory diagram showing a corresponding composite image when low-pass filter processing is not performed.

FIG. 10 is an explanatory diagram showing a corresponding composite image when low-pass filter processing has been performed in advance.

FIG. 11 is an explanatory diagram showing a corresponding composite image when another low-pass filter processing is performed in advance.

FIG. 12 is an explanatory diagram showing a corresponding composite image when another low-pass filter process is performed in advance.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Image input device 12 ... Instruction input device 14 ... Image synthesis device 16 ... Instruction device 18 ... Image output device

Claims (3)

[Claims] An image processing apparatus for performing an image processing for pasting a texture image to a part of an original image, wherein a point on a two-dimensional coordinate of a pasting area where a texture image is pasted corresponds to a distance between adjacent points. Calculating means for calculating a distance on the texture image coordinates, filter parameter selecting means for automatically selecting a low-pass filter parameter for each of the points based on the calculated distance, and using the selected low-pass filter parameter. Image processing means for pasting the texture image subjected to the filtering process into the pasting area. 2. The image processing apparatus according to claim 1, wherein the image processing is image synthesis or mapping of a texture image. 3. The method according to claim 1, wherein the calculating means calculates a distance on the texture image corresponding to a distance between two adjacent points around the point of interest, and calculates an average value of the distance for the point of interest. An image processing apparatus characterized in that the distance is between adjacent points.