JPS58108868A - Picture converter - Google Patents

Abstract

PURPOSE:To obtain an output picture of which top and reverse is discriminated, by utilizing a sign of either positive or negative in the matrix of a determinant appearing in approximation by affine deformation in each extremely small area, and discriminating whether a top or reverse surface is obtained in every extremely small area. CONSTITUTION:When an input picture is converted, a top/reverse decision is made and on the basis of the decision result, two input pictures 1A and 1B are selected as a top-side and a reverse-side picture respectively to obtain an output picture 2; even when one input picture is used, the reverse side 5 is replaced by proper reverse-side information, e.g. specific color. After this conversion, whether the picture faces up or down is discriminated on the basis of the sign of the determinant of a matrix of coefficients of this picture conversion system. This picture conversion system provides geometric transformation of pictures in every extremely small area of the picture by affine conversion.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an image conversion device that performs geometric conversion of an image using a memory or the like.

As an example of geometric transformation, as shown in FIG. 1, an output image 2 is formed from an input image 1, having a transformed image 3 and a background portion 4, which are obtained by wrapping the input image 1 around a cylindrical surface. There is something to do. In such a case, the back side 5 of image 3 after conversion
(K), the input image will appear visually unnatural.

This invention aims to solve this problem, and when converting an input image, it distinguishes between the front side and the back side, and based on this discrimination, one input image 1 as shown in Fig. 1B are selected as the front side image and the back side image, respectively, to obtain both output images 2. Even when using 7 input images, as shown in FIG. The back side information example is replaced with a predetermined color.

As described above, whether the orientation of the surface after conversion is the front side or the back side is determined from the sign or negative of the determinant of the coefficient matrix in the image conversion method of LXK of the present applicant. Jin's image transformation method implements geometric transformation of images by approximating each minute region of both images by affine transformation. First, we will explain this image conversion method and in Fig. 3,
The point (X2,
Suppose that we move to y, y: ) by geometric transformation. Let us write this as y2=ψ1(x2, X:) y:=92(x2, X:). 91. and (X:,X:)K of ψ3
Using the partial differential value (or difference value) at , ψl and ψ3 are approximated to the seventh order as follows.

However, i, j=1. z and α are numbers greater than or equal to 1) where matrix M=
Determinant of (alj) When the value is positive, it means that the orientation of the surface after transformation is the same as the orientation of the surface before transformation, that is, it is the front side, and when the value is negative, the orientation of the surface after transformation is This indicates that it is different from the one before conversion, that is, it is the back side. Therefore, 1
It is possible to distinguish between the front side and the back side depending on the sign of the frame 1.

:' Figure D shows the configuration of an embodiment of Jin's invention. In the same figure, 6 and 7 indicate input image memories each having a field memory or 7-frame memory configuration, and one of the input image memories. The input image memory 6 stores image data corresponding to the front side image, and the other input image memory 7 stores image data corresponding to the back side image. One of these two input image memories 6 and 7 is selected by an input image selection device 8, and the input image from the selected input image memory is supplied to a conversion processing device 9. The conversion processing device 9 performs address calculation and image data interpolation processing based on the address calculation, and the converted image data is written into the output image memory 10.

The conversion processing device 957 receives information on the conversion formula in advance, and
Address control and arithmetic processing for image memory 6.7.10 are performed. sample points of the input image (X
If OI A write address for the output image memory 10 is formed.

Further, from this conversion processing device 9, information regarding the positive or negative sign of the determinant 1m1 of the matrix 1 in each minute region is sent to the input image selection device 8, and when the sign of the determinant 1m1 is positive, Image data is read out from the human-powered image memory 6, and if this is negative, image data is read out from the input image memory 7.

As shown in Figure 1 B, when specifying the back side part 5 and applying nine colors, only seven input image elements are used, and the sign of column 1 is negative. At this time, access from this input image memory is stopped, and processing is performed to write the data of the specified color to the specified address of the output image memory.

Furthermore, as shown in Figure 3, one input image memory-6
Even if a selection device 11 is provided on the data input side and the selection device 11 is switched in response to a control signal indicating the sign of the determinant 1 from the conversion processing device 9, .

As can be understood from the description of the above embodiment, in this invention, by using the sign of the determinant of the matrix that appears during approximation by affine transformation on each minute region),
It is possible to determine whether each micro-region is a front or back surface 1, and it is possible to obtain an output image that distinguishes between the front and the back. According to the present invention, it is also possible to provide the image conversion device with the function of a switcher.

[Brief explanation of drawings]

1 is a route map used to explain a conventional image conversion device, FIG. 1 is a route map used to explain an image conversion device according to the present invention, and FIG. 3 is a route map used to explain an example of an image conversion method. FIG. 3 is a block diagram showing the configuration of one embodiment of the present invention, and FIG. 3 is a block diagram showing the configuration 6 of another embodiment of the present invention. 1.1m, l B-v-・-Human image, 2-------
--Output image, 5... Back side, 6. '7=
......Input image memory -1,9...
... Conversion processing device, 10-... Output image,
memory. Agent Tadashi Sugiura Figure 1-X+ Yl Figure 4! Figure 5

Claims (1)

[Claims] Divide the input image information into predetermined areas, perform calculation processing using a conversion formula for each area, determine the polarity of the calculation output, and convert the input image 1 to other image information according to the judgment output of Jin. An image conversion device that replaces the .