JPH057584A - Image expander - Google Patents

Abstract

PURPOSE:To achieve a prevention of a drop in quality of image and speeding up of an image expansion by adding products of pixel data arranged vertically and horizontally in one way and a specified coefficient to generate pixel data positioned between two center pixel data while 7 pixel data is made after interpolation in the same way as above. CONSTITUTION:In an image memory 3, a command of a reading address is provided with a Y counter and an X counter 2 to read out pixel data stored at the address. A selector 4 distributes the pixel data sequentially among a plurality of registers 51-54. Here, an X-way interpolator 6 performs an X-way interpolation calculation based on pixel data while the pixel data are sent to a plurality of line buffers 81-84 via a selector 7. Moreover, an assigning signal from an X counter 1 is sent to the line buffers 81-84 and the image data are read out from the head of the lines to be sent to a Y-way interpolator 9, which interpolates the pixel data in the direction Y to be sent to a display memory 12.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image enlarging apparatus, and more particularly to an image enlarging apparatus suitable for enlarging a medical image obtained by a medical image diagnostic apparatus such as an X-ray CT apparatus or an MRI apparatus.

[0002]

2. Description of the Related Art When a medical image or various digital images obtained by a medical image diagnostic apparatus or the like is enlarged and displayed on a CRT display device or the like, it is necessary to automatically interpolate between pixels of an enlarged image. .. In general, the tile interpolation method or the linear interpolation method is used to calculate new pixel data.

[0003]

However, the conventional tile interpolation method or linear interpolation method has a problem that the image quality of an enlarged image is poor and the contour is not smooth or unclear. is there.

In view of the above, the present invention enables high-speed enlargement with almost no deterioration in image quality.
An object is to provide an image enlarging device.

[0005]

In order to achieve the above object, in the image enlarging device according to the present invention, a predetermined coefficient is assigned to each of at least four pixel data arranged adjacent in one direction in the vertical and horizontal directions. The pixel data located between the two pixel data in the center is interpolated by multiplying and then adding, and the pixel data after the interpolation are arranged at least 4 adjacent to each other in other vertical and horizontal directions. It is characterized in that each pixel data is multiplied by a predetermined coefficient and then added to interpolate pixel data positioned between the two pixel data in the center. Vertical and horizontal 4
Since interpolation calculation is performed from individual pixel data, so-called cloud-shaped ruler spline function can be used for interpolation, the contour is smooth, and the image can be enlarged with almost no deterioration in image quality. Further, since one of the vertical and horizontal directions is interpolated and then the other is interpolated, and the respective interpolation calculation only requires multiplication and addition, it is possible to perform high-speed processing.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the drawings. In this embodiment, it is assumed that the image memory 3 shown in FIG. 1 stores image data to be enlarged. The Y counter 1 and the X counter 2 are reset by the controller 13 and count-up signals are sequentially sent to them. A read address is designated by the Y counter 1 and the X counter 2 in the image memory 3, and the pixel data stored at the address is read. As a result, the data of one pixel is sequentially read from the image memory 3 for each line in the X direction.

The selector 4 sequentially distributes the pixel data to each of the registers 51 to 54. Therefore, the controller 13 outputs the select signal to the selector 4 and the register select signals to the registers 51 to 54. In this way, four pixel data adjacent to each other on the X line are stored in the four registers 51 to 54. When all the four adjacent pixel data are stored in the registers 51 to 54 in this way, they are sent to the X-direction interpolator 6, and the interpolation calculation in the X-direction is performed from the four pixel data. ..

Each pixel data output from the interpolator 6 including new pixel data obtained by the interpolation calculation is sent to the line buffers 81 to 84 via the selector 7. Each of the line buffers 81 to 84 has a capacity for storing pixel data for one line in the X direction (including the pixel data interpolated and increased). The pixel data of each line is sequentially filled in each of the line buffers 81 to 84 by the select signal from the controller 13 to the selector 7 and the buffer select signal from the controller 13 to the line buffers 81 to 84. In this way, in the line buffers 81 to 84, pixel data for four lines of the enlarged image stored in the image memory 3 are interpolated and enlarged in the X direction.

A signal designating the X direction pixel position from the X counter 11 is sent to the line buffers 81 to 84, the X direction address is designated from the beginning by this signal, and four pixel data having the same X direction address are designated. (4 adjacent lines in the Y direction) are read from the beginning of each line and sent to the Y direction interpolator 9. The interpolator 9 interpolates the pixel data in the Y direction (interpolates between lines) by the four pixel data, and the pixel data output from the interpolator 9 including the new pixel data obtained by the interpolation. Are sent to the display memory 12. X counter 1
1 and the Y counter 10 generate a signal for designating a write address of the display memory 12, send an X direction address signal to the line buffers 81 to 84, and send a Y direction address signal to the Y direction interpolator 9. By sending
The address of the pixel data output from the Y-direction interpolator 9 and the write address of the display memory 12 are matched. As a result, interpolation enlargement between lines has been performed.

By repeating such an operation, the enlarged image stored in the image memory 3 is finally interpolated and enlarged and stored in the display memory 12.

Since the interpolators 6 and 9 have the same structure, only the interpolation in the X direction will be described in detail with reference to the interpolator 6 shown in FIG. As shown in FIG. 2, the X-direction interpolator 6 has 4
Coefficient memories 21 to 24 and four multipliers 31 to 34
And an adder 41. As shown in FIG. 3, four pixel data Mi−1, Mi, Mi + 1, Mi + 2 which are adjacent to each other in the X direction on one X line are stored in the registers 51 to 5.
It is assumed that the interpolation data M (t) stored in each of the data No. 4 is interpolated from these data in the X direction from the point i to the point i + 1. Here, interpolation using a cloud-shaped ruler spline function is performed using the formula M (t) = α1 (t) Mi-1 + α2 (t) Mi + α3 (t) Mi + 1 + α4 (t) Mi + 2. Assumed to be performed. However, 0 ≦ t ≦ 1. The weighting factor α (t) is α1 (t) =-{t (t-1) ² } / 2 α2 (t) = {(t-1) (3t ² -2t-2)} / 2 α3 (t) = {t (3t ² -4t-1)} / 2 α4 (t) = {t ² (t-1)} / 2. At the left end of the image, −1 ≦ t <0, and the weighting factor α
(T) is α1 (t) =-{t (t ² -4t + 5)} / 10 α2 (t) = {(t + 1) (3t ² -10t + 10)} / 10 α3 (t) = {t (t + 1) (5-3t)} / 10 α4 (t) = {t ² (t + 1)} / 10. The same is true at the right end.

The above weighting factors are stored in advance in the coefficient memories 21 to 24 of FIG. The number of interpolation points is obtained in advance from the enlargement magnification when enlarging the image, and each weighting coefficient is stored in the coefficient memories 21-24. For example, when the magnification is 4 times, three points are interpolated between Mi and Mi + 1.
Interpolation coefficients are obtained with t = 0.25, t = 0.5, and t = 0.75. Mi itself has a weighting coefficient (α1 = 0, t = 0,
α2 = 1, α3 = 0, α4 = 0) is obtained. These are written in advance in the coefficient memories 21 to 24 via the controller 13, the multipliers 31 to 34 perform multiplication of these weighting factors and the pixel data, and the addition result is added by the adder 41. , It means that the interpolation calculation by the cloud-shaped ruler-like spline function shown in the above equation is performed.

Although the weighting coefficient is written in the coefficient memories 21 to 24 in advance here, the weighting coefficient is calculated every time by the controller 13 and is multiplied by the multiplier 31 of the interpolators 6 and 9.
It is also possible to output directly to ~ 34. In addition, the magnified image data is not stored in the image memory 3,
It is also possible to directly use the one that is sequentially output from each line from the X-ray CT apparatus or the MRI apparatus. Furthermore, if the processing of the interpolators 6 and 9 is configured to be performed at high speed, the image data enlarged by interpolation can be directly displayed on the CRT display device or the like without being temporarily stored in the display memory 12.

[0014]

As described in the above embodiments, according to the image enlarging apparatus of the present invention, interpolation by a so-called cloud ruler-like spline function can be performed, the image quality is hardly deteriorated, and the contour is smooth. The image can be enlarged without impairing the image quality. Also, vertical
Since each horizontal interpolation calculation only requires multiplication and addition, it can be processed at high speed.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a block diagram showing a part of the embodiment in detail.

FIG. 3 is a diagram showing pixel data.

[Explanation of symbols]

 1, 10 Y counter 2, 11 X counter 21, 22, 23, 24 Coefficient memory 3 Image memory 31, 32, 33, 34 Multiplier 4, 7 Selector 41 Adder 51, 52, 53, 54 Register 6 X direction interpolation Device 81, 82, 83, 84 Line buffer 9 Y direction interpolator 12 Display memory 13 Controller

Claims (1)

Claim: What is claimed is: 1. At least four pixel data arranged adjacently in one direction in the vertical and horizontal directions are multiplied by a predetermined coefficient, respectively, and then added to obtain the two central pixels. First interpolation means for creating pixel data located between the data and at least four pieces of pixel data after the above interpolation, which are arranged adjacent to each other in other vertical and horizontal directions, are multiplied by a predetermined coefficient. A second interpolating means for producing pixel data positioned between the two pixel data at the center by adding the pixel data at the center.