JPH04280376A - image display device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image display device that restores and displays encoded image data, and more particularly to an image display device that can perform image searches efficiently.

0002

[Prior Art] Image data, especially halftone image and color image data, has a larger amount of information than character and numerical data.
Since the amount of data (data amount) is orders of magnitude larger, in order to store the data or transmit it at high speed and quality, it is necessary to encode the tone values of each image with high efficiency.

[0003] Also, when searching an image database, in order for the user to more quickly recognize the outline of the image,
Efficient image retrieval is possible using so-called hierarchical restoration (this hierarchy is classified according to image resolution), which gradually improves the image quality from coarse images that can be displayed quickly to high-quality images. It is strongly desired to realize a possible image restoration method. When performing such a search, there is an extremely strong demand for an image display that allows users to search for a desired image with the same feeling of finding a desired page while turning the pages of a book, like searching for a page in a book.

By the way, as a highly efficient compression method for image data, for example, adaptive discrete cosine transform (Adaptive Discrete Cosine Transform) is used.
e Discrete Cosine Transform
rm, hereinafter abbreviated as ADCT) encoding method is known.

[0005] This ADCT encoding method uses, for example, 256
Two-dimensional original image data sampled into ×256 pixels or 512 ×512 pixels, etc. is divided into multiple blocks of, for example, 8 × 8 pixels (or 16 × 16 pixels), and the image signal (level) of each block is divided. Adjustment value or density value, etc.) to 2
2D Discrete Cosine Transform (2D Discrete Co
sine Transform (hereinafter abbreviated as two-dimensional DCT) into coefficients of a spatial frequency distribution, quantize the coefficients using a threshold adapted to visual perception, and encode the determined quantized coefficients using a statistically determined Huffman table. It is something that becomes.

Here, a basic block diagram of a circuit that divides the image signal 23 into blocks of 8×8 pixels and encodes and stores the image signal of each block by ADCT is shown in FIG. 11. The encoding operation will be explained below.

First, a pixel signal 23 of 64×64 pixels is divided into 8×8 blocks each consisting of 8×8 pixels, as shown in FIG. 12, and is input to a two-dimensional DCT transform unit 24. Note that each block value shown in FIG. 11 indicates a gradation value.

The two-dimensional DCT transform unit 24 orthogonally transforms the input image signal of each block by two-dimensional DCT to obtain two-dimensional DCT coefficients F(u
, v), (0≦u≦7, 0≦v≦7) and output to the linear quantization unit 25. Two-dimensional DCT coefficient F(u,v)
A smaller value of u or v represents a spectral component with a lower spatial frequency and a lower resolution, and a larger value of u or v represents a spectral component with a higher spatial frequency and a higher resolution. Further, F(0,0) indicates a DC component.

The linear quantization unit 25 linearly quantizes the input two-dimensional DCT coefficients F(u,v) with reference to a quantization matrix 26 composed of threshold values determined through visual experiments. As a result of this quantization, as shown in FIG. 14, the two-dimensional DCT coefficient F(u,v) below the threshold becomes 0, and
A quantized coefficient is generated in which only the C (DC) component (“5”) and a few AC alternating current components have values.

Next, the quantized coefficients arranged two-dimensionally as shown in FIG. 14 are 1, 2, 3,
. . 61, 62, 63, 64 are zigzag scanned in the order in which they are assigned, thereby being converted into a one-dimensional data string and input to the variable length encoding unit 27.

[0011] Regarding the DC component, the variable length encoding unit 27 performs variable length encoding on the difference between the DC component and the DC component of the previous block. Furthermore, for the AC component, the values of effective coefficients (coefficients whose value is not 0) and the run length (zero run length) of invalid coefficients (coefficients whose value is 0) up to that point are variable-length encoded. And then D
Each of the C and AC components is encoded using a code table 28 consisting of a Huffman table created based on statistics for each image, and stored in a code data storage section 29. At this time, the code data is such that the data necessary to restore a coarse image is in the first layer, the data to be added sequentially to restore a coarser image is in the second layer, the third layer, and so on. Stored in the form of hierarchical code data. That is,
The two-dimensional DCT coefficients F(u,v) are divided and hierarchically stored in order from those corresponding to the lowest spatial frequency components of order (u,v).

[0012] In this way, code data stored in layers is sequentially restored from a lower layer to a higher layer, that is, from a coarse image to a finer image. By the way, as mentioned above, all the pixels of the sampled image are divided into a plurality of blocks each consisting of an appropriate number of pixels, and for each block, the numerical sequence consisting of the sampled values is orthogonally transformed using two-dimensional DCT, etc. , obtaining a number of coefficients equal to the number of blocks and quantizing the coefficients according to a predetermined rule is generally called transform coding or block coding.

A method of restoring an image from this hierarchical coded data will be explained with reference to circuit block diagrams shown in FIGS. 16 and 17. FIG. 16 is a diagram showing the configuration of a conventional apparatus for restoring an image from hierarchical code data (hierarchical code data).

In the prior art, first, the hierarchical code data of a large number of images stored in the code data storage unit 110 shown in FIG.
Sequential selection is made in order of hierarchy from the code data of the hierarchy. The first layer encoded data of the first image selected by the image selection section 120 is output to the image restoration section 130, and is restored into an image signal by the image restoration section 130. The restored image signal is then displayed on the image display section 140.

As described above, after the first layer encoded data of the first image has been restored to an image signal, the image selection section 120 then restores the second layer encoded data of the first image in the encoded data storage section 110. Code data is selected, and an image more detailed than the previous image is restored using the code data of the first layer and the code data of the second layer using the same procedure as described above.

Here, the image restoration section 13 shown in FIG.
0, the image restoration unit 130
Explain the operation. The image restoration unit 130 first restores the coded data of the first layer of the first image. That is, first, the code data of the first layer of the first image is read from the code data storage unit 110 into the variable length decoding unit 131 in order from the first block, and the variable length decoding unit 131 refers to the decoding table 132. However, fixed length data (two-dimensional DCT
coefficient). 2 restored by the variable length decoding unit 131
The dimensional DCT coefficients are dequantized by the dequantization unit 133 using the quantization matrix 134, and the quantized coefficients obtained by the dequantization are stored in the coefficient storage unit 135. Each value in the coefficient storage section 135 is initially given zero. Therefore, only the coefficients restored to non-zero values by the variable length decoding unit 131 and quantized by the dequantization unit 133 are newly written into the coefficient storage unit 135. After all code data of the first layer of the first image is read and variable length decoding is completed, each coefficient (quantized coefficient) of the first layer of the first image stored in the coefficient storage unit 135 is subjected to inverse DCT. The conversion unit 136 calculates and converts it into gradation data (image signal), which is displayed on the image display unit 140.

After the first layer encoded data of the first image is restored as described above, all the subsequently transmitted encoded data of the second, third, . . . layers are sequentially restored. Then, after restoring the code data of all layers for the first image, the second image is sequentially restored starting from the first layer using the same procedure.

In this way, images equal to the number of layers are displayed for each of the first, second, etc. images from coarse images to finer images, and all layers are displayed. After finishing, the display moves to the first layer of the next image.

[0019]

[Problems to be Solved by the Invention] As mentioned above, conventionally, each image is transform-encoded, and the coded data obtained by the transform-encoder is hierarchically layered so that finer images are obtained sequentially from coarser images. I remembered it. By sequentially decoding each image in the order of the first layer, second layer, etc., each image is sequentially restored and displayed from a coarse image to a finer image.

However, in this restoration/display, the next image is restored/displayed on the same screen after the previous image has been restored/displayed to the required level, so each image is presented independently. It was like that. For this reason, there have been drawbacks such as the inability to realize a natural image search that is familiar to humans, such as searching while turning the pages of a book.

SUMMARY OF THE INVENTION An object of the present invention is to realize an image display device that allows a user to efficiently search for images as naturally as if turning the pages of a book.

[0022]

[Means for Solving the Problems] FIG. 1 is a diagram illustrating the principle of the present invention. The present invention is based on an image display device that restores and displays image data of a plurality of encoded images.

The search direction instructing means 1 instructs, for example, the search direction of a plurality of images stored in advance in the storage means in a predetermined order, that is, whether it is a forward search or a backward search.

The restoring means 2 performs the above-mentioned encoding and inverse transformation processing on the image data encoded by various orthogonal transformations in the order according to the search direction indicated by the search direction indicating means 1, thereby producing an image for display. Restore to data. At this time,
The encoded image data is stored in a hierarchical manner so that each image is displayed in stages from a coarse image to a finer image so that, for example, an image database can be searched efficiently.

The transforming means 3 transforms the display image data that has been restored by the restoring means 2 into the search direction indicating means 1.
Various transformations are performed depending on the search direction indicated by . This modification is, for example, a process in which the display image data of the restored previous image is displayed after being reduced in the horizontal or vertical direction. In the modification, for example, as described in claim 4,
The transformation in which the previous image is displayed in a reduced size is divided into multiple steps so that the reduction ratio gradually increases, and each time the transformation process is completed, the created image data for transformation display of the previous image is is output to display control means 4, which will be described later. In this case, for example, when the reduction direction of the previous image is the horizontal direction and the search direction instructed by the search direction means 1 is the forward direction, the deformation means 3, as described in claim 5, The transformation process is performed so that the previous image, which is being reduced and displayed in the horizontal direction, is displayed moving from left to right or right to left on the screen, while the search direction specified by the search direction instruction means 1 When is in the opposite direction, the deformation process is performed so that the previous image, which is being reduced in the horizontal direction, is moved and displayed in the opposite direction to that in the forward search. Furthermore, when the reduction direction of the previous image is the vertical direction and the search direction instructed by the search direction instruction means 1 is the forward direction, the transformation means 3 is configured to The transformation process is performed so that the previous image, which is being reduced and displayed in the vertical direction, is displayed moving from top to bottom or from bottom to top on the screen, and on the other hand,
When the search direction instructed by the search direction instructing means 1 is the opposite direction, the transformation process is performed so that the display is moved in the opposite direction from the forward search.

The display control means 4 also displays an image obtained by the display image data of the next image newly restored by the restoration means 2 and a previous image obtained by the display image data created by the transformation means 3. Control is performed so that the deformed image is displayed in combination with the deformed image. In this composite display, for example, as described in claim 2, the transforming means 3
The deformed image of the previous image obtained by the display image data created by is controlled to be superimposed and displayed on the new image obtained by the display image data restored by the restoring means 2.

Furthermore, the display control means 4 can display a previous image obtained from the image data created by the transformation means 3 and a new image obtained from the image data newly restored by the restoration means 2. Control is performed so that at least one image is displayed with a border. Further, as described in claim 8, the luminance of the transformed image of the previous image obtained by the image data created by the transforming means 3 and the data of the new image obtained by the image data newly restored by the restoring means 1 are different. Control to be displayed differently.

[0028]

[Operation] First, the user instructs the image search direction (forward or reverse direction) via the search direction instruction means 1. In this case, the forward search corresponds to, for example, searching in the order of storage of encoded image data.
The backward search corresponds to searching for images in the opposite order to the above-mentioned forward search.

The restoration means 2 sequentially inputs the encoded image data of each image in an order corresponding to the search direction instructed by the search direction instruction means 1. Then, when encoded image data of an arbitrary image is input, an inverse transformation of the encoding is performed and image data (display image data) is restored from the encoded data. At this time, if the encoded data is stored (stored) in layers such as the first layer, second layer, etc. so that the code data is restored and displayed from a coarse image to a finer image, the first layer The image data is restored in hierarchical order, such as the encoded data of the second layer, the encoded data of the second layer, and so on. When the restoration and display of the first image (first image) is completed during forward search, restoration of the image data of the next second image is started.

On the other hand, in the backward search, after the restoration and display of the second image is completed, restoration of the first image is started. Next, the operation during forward search will be explained first.

The first image data that has been restored by the restoration means 2 is output to the display control means 4 and also to the transformation means 3. The deforming means 3 performs a deforming process on the restored and completed first image data, such as displaying it in a reduced size, for example, horizontally or vertically. This deformation process is performed multiple times, for example, while changing the reduction ratio stepwise (linearly or non-linearly). The transformed image data subjected to the transformation processing is sequentially outputted to the display control means 4 after the processing is completed.

In this manner, while the transforming means 3 is sequentially outputting the transformed image data of the previous image to the display controlling means 4, the restoring means 2 outputs the image data of the second image, which is the next new image. and outputs the restored image data to the display control means 4.

The display control means 4 combines the transformed image data of the first image data inputted from the transforming means 3 and the newly restored second image data inputted from the restoring means 2, and displays the resultant image as an image. .

The above-described operations are sequentially repeated for the second image, the third image, and so on. Therefore, the deforming means 3 sequentially reduces the image data of the previous image that has been restored and completed by the restoring means 2, for example, in the horizontal or vertical direction at a predetermined ratio (which may be non-linear) multiple times. By doing so, it becomes possible to display an image as shown in FIG. 7 or 8. Note that in FIGS. 7 and 8, image A is the previous image, and image B is the new image.

On the other hand, at the time of backward search, the restoring means 2
The encoded image data is sequentially restored in the reverse order of the forward search, such as the nth image, the n-1st image, . . . the second image, and the first image. Therefore, the restored and completed image data input to the transformation means 3 are also in the reverse order of the forward search. Therefore, similar to the forward search described above, the deformation means 3 sequentially reduces the image data of the previous image at a predetermined ratio in the horizontal or vertical direction multiple times. It becomes possible to display an image as shown in 10. Note that in FIGS. 9 and 10, image B is the previous image and image A is the new image.

As can be seen by comparing FIG. 7 and FIG. 9 or FIG. 8 and FIG. 10, the display position (movement direction) of the reduced image of the previous image is reversed between the forward search and the backward search. This makes it possible to perform an image search as if turning the pages of a book in order or in reverse.

[0037]

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 2 is a block diagram showing the system configuration of an image display device according to an embodiment.

In the figure, the code data storage unit 210 stores each pixel obtained by sampling each image into M×M (for example, M=256 or M=512) pixels as described above. The image signal (gradation value) of
,n) (m=0,1,...N-1;n=0,1,...
・The data encoded by sequentially performing the two-dimensional DCT, linear quantization, and variable length coding on N-1) is transferred to each layer of the first layer, second layer, ... It is divided and stored. As the code data storage section 210, various storage means such as a semiconductor memory, a magnetic storage device, an optical disk, a magneto-optical disk, and a CD-ROM can be used.

[0039] The following various methods are applied to the encoding using the two-dimensional DCT. ■ The DC component is DPCM (Differentia) with the previous block.
ADCT encoding method performs two-dimensional Huffman encoding of zero run length and non-zero coefficient values by zigzag scanning the AC component. ■ Low-frequency components are encoded using inter-frame/intra-frame adaptive prediction DPCM, and high-frequency components are directly encoded using two-dimensional DCT. ■ Both DC and AC components are two-dimensional DCT
A method of encoding.

The image restoration section 220 first receives the code data of the first layer of the first image from the code data storage section 210. The code data storage unit 210 stores hierarchical code data of a plurality of images in a predetermined order.
It is expressed as an image.

[0041] In the forward search, the first image,
An image search is performed in the storage order, such as the second image, etc., and in the reverse search, it is possible to search in the reverse order of the storage order, for example, the 10th image, the 9th image, etc.

The user's instruction for the forward search or backward search is given to the code data storage unit 210 via the search direction detection unit 280. Here, an example of the configuration of the search direction detection section 280 is shown in FIG.

In the example shown in the figure, the search direction detection section 280
A mouse cursor position detection unit 281 detects the position (coordinates) of the mouse cursor on the display screen from mouse movement amount information output from a mouse (not shown), and a mouse cursor output from the mouse (not shown). A mouse button input detection unit 282 detects a mouse button click operation from button press information, and when a click operation detection signal is input from the mouse button input detection unit 282, the mouse cursor position is read from the mouse cursor position detection unit 281. , determines whether the mouse cursor position at the time of the click operation is within a predetermined display area that instructs forward or reverse search, and determines that forward or reverse search has been performed. , and an area determination unit 283 that outputs the search instruction signal to the code data storage unit 210.

FIG. 4 shows an example of a screen display for instructing a search direction, and a method for instructing an image search direction on the display screen will be described. If the user desires to perform a forward search by operating a mouse, the user moves the mouse cursor 291 into the "forward" display area 292 and then clicks the mouse button. On the other hand, if you wish to search in the reverse direction, move the mouse cursor 291 to the "reverse direction" display area 2.
93, click the mouse button. The mouse button input detection section 282 outputs a signal to the area determination section 283 when detecting the click operation of the mouse button. When the area determination unit 283 receives the signal from the mouse button input detection unit 2823, it reads the mouse position information from the mouse position detection unit 281 and determines the position of the mouse cursor 291 within the display screen. The determination is made that the display position of the mouse cursor 291 is “
The current position information of the mouse cursor 291 is displayed in the "forward" display area 2 to determine whether it is within the "forward" display area 292 or within the "reverse" display area 293.
92 or the coordinates of the "reverse direction" display area 293 on the screen.

When the user instructs ``forward search'' or ``reverse search,'' the search direction instruction signal is output to code data storage section 210 and modification calculation section 260.

The code data storage unit 210 outputs the code data of the relevant layer of the image data to the image data restoration unit 220 in response to the search direction instruction signal applied from the search direction detection unit 280 as described above.

The configuration of this image restoration section 220 is shown in FIG. As shown in the figure, the block configuration of this image restoration section 220 is similar to that of the image restoration section 130 shown in FIG. The hierarchical code data of any image is
The first block in the image is sequentially read into the variable length decoding unit 221, and using the decoding table 222, fixed length data (2
dimensional DCT coefficients). Then, the two-dimensional DCT coefficients restored by the variable length decoding section 222 are dequantized by the dequantization section 223 using a quantization matrix 224, and the resulting quantized coefficients are stored in the coefficient storage section 222.
It is stored in 5. Each value in the coefficient storage section 225 is given zero as an initial value. Therefore, the variable length decoding section 22
Only the quantized coefficients restored to non-zero values by 1 and dequantized by the dequantization unit 223 are newly written into the coefficient storage unit 225. After reading all the code data of the first layer of an arbitrary image and completing variable length decoding, the coefficient storage unit 22
The inverse two-dimensional DCT transform unit 226 performs inverse two-dimensional DCT transform on each coefficient of the first layer of an arbitrary image stored in the image storage unit 230 as image data. write to.

Thereafter, the image restoration section 220 similarly restores the image data of the second, third, . . . , higher layers, and stores the restored image data in the image storage section
Write it down to 30. The image restoration unit 220 also writes the final restored image data of each image to the transformed image storage unit 250.

The image data written to the image storage section 230 is immediately transferred to the image display section 270 by the display control section 240.
The image data written in the transformation image storage unit 250 is displayed on the image display unit 270 via the display control unit 240 after being transformed by the transformation calculation unit 260.

[0050] Here, the first pixel [image A] and the second image [
Taking up the display of image B], the display operation in the case of forward search and the display operation in the case of reverse search will be explained. It is assumed that the code data of each image is stored in the code data storage unit 210 in layers from the first layer to the Nth layer. ■ Search from image A to image B (forward search) In this case, the code data storage unit 210 first outputs the code data of the first layer of image A, which is the first image, and then sequentially
2nd layer, 3rd layer, . . . Nth layer code data, and further the 1st layer, 2nd layer of image B which is the second image, etc.
- Output the coded data of the Nth layer to the image restoration section 220.

[0051] The image restoration unit 220 includes the encoded data storage unit 2
The encoded data input from 10 is sequentially restored to image data, and the image data is written into the image storage section 230. In this case, image data from the first layer to the N-1th layer of image A are written only to the image storage section 230. Then, the display control unit 240 reads out the image data sequentially written into the image storage unit 230 in the writing order, and
The image is output to the image display section 270.

Therefore, first, the image A is restored and displayed on the image display section 270 in stages from the first layer, second layer, . . . to the Nth layer, that is, from a coarse image to a fine image. . When image A is restored and displayed up to the Nth layer (final layer) in this manner, the image restoration section 220 outputs the image data of the Nth layer of image A to the transformation image storage section 250. Then, the image restoration unit 230
Restoration of the first layer encoded data into image data is started.

When the image data of the Nth layer of image A is written to the image storage unit 250 for transformation as described above, the transformation calculation unit 260 reads out the image data and transforms the image data multiple times. The processing is performed, and each time the transformation processing is completed, it is output to the display control unit 240. At this time, the display control unit 240 also stores the image data of the first layer of the image B restored by the image restoration unit 220 in the image storage unit 23.
Input from 0. Therefore, the display control unit 240 receives the N-th layer deformed image data of the image A that has been deformed by the deformation calculation unit 260 and the first layer image data of the image B from the image storage unit 230. . At this time, the display control section 240 synthesizes the image data of the Nth layer of the image A and the image data of the first layer of the image B, and causes the image display section 270 to display the synthesized image.

FIGS. 7 and 8 show examples of image display when the transformation calculation unit 260 performs transformation processing by affine transformation on the image data of the Nth layer of image A. In this case, image B is displayed so as to be below image A, and image B is displayed so as to appear from below image A.

[0055] Hereinafter, similar processing is performed on the second image [Image B]
and the third image [Image C], the third image [Image C]
and the fourth image [Image D], it is possible to restore and display images as if turning the pages of a right-bound book, allowing the user to search for images in a natural way. It can be performed.

Contrary to the above embodiment, the previous image may be displayed in a reduced size on the left side of the screen in accordance with a left-bound book. In affine transformation, orthogonal (x
, y) coordinate system to the orthogonal (X, Y) coordinate system using the following equation (1), and calculate the position (coordinates) of the pixel after the transformation. Calculate and enlarge/reduce, rotate, etc. the image.

[0057]

[Math 1]

In the same equation, m and n are the coordinate axes x, y
represents the magnification in the direction, and θ is the rotation angle around the origin (0,0). In the example shown in FIG. 7, n=1, θ=0, and the values of m are 0.9, 0.8, 0.6, 0.4.
, 0.1, 0.0, m=
0.7 ((B) in the same figure), m=0.3 ((C) in the same figure)
), the screen display when m=0.0 ((D) in the same figure) is shown.

If n=1 and θ=0, then X=mx Y=y, so the image A transformed by the transformation calculation section 260
The transformed image is an image that has been reduced in the horizontal direction (X direction) of the screen, as shown in (B) and (C) of the same figure.

Further, in the example shown in FIG. 8, in the above equation (1), m=1, θ=0, and the values of n are 0.9, 0.
．． This is an example of a composite screen displayed when affine transformation is performed to sequentially change the value in six steps: 8, 0.6, 0.4, 0.1, and 0.0.

In the above formula (1), m=1, θ=0
Then, X=x Y=ny, so the image A transformed by the transformation calculation unit 260
The modified image of the Nth layer is an image reduced in the vertical direction of the screen, as shown in (B) and (C) of the same figure.

When the deformation calculation unit 260 performs deformation processing on the image data of the Nth layer of image A in which the reduction ratio increases in stages in the horizontal direction, the results are shown in FIGS.
It is clear that if we look at C) and (D) in order,
Image display is performed as if turning the pages of a right-bound book page by page.

[0063] Also, in the deformation calculation section 260, the image A
When the image data of the Nth layer is subjected to transformation processing in which the reduction ratio increases in stages in the vertical direction, the results are as shown in FIGS. 8(A) and 8(B).
, (C), and (D), the image display is performed as if one were turning the pages of a bottom-bound book one by one. Such image display has the advantage that when displaying a newly restored image ([Image B] in FIG. 6), the images can be displayed as is in the order in which the images are restored by the image restoration unit 220.

It should be noted that, contrary to the above embodiment, the previous image may be displayed in a reduced size on the upper side of the screen. ■ Search for image A from image B (reverse search) In this case, restoration is started from image B first. That is, the code data storage section 210 sequentially outputs the code data of the first layer of image B, the second layer, the third layer, . . . Nth layer to the image restoration section 220. In this case as well, the image restoration unit 220 does not output the restored image data to the transformation image storage unit 250 until the Nth layer code data of image B is restored, and outputs it only to the image storage unit 230. . Therefore, first, image B is restored and displayed hierarchically from the first layer to the Nth layer, that is, from a coarse image to a fine image.

Next, the code data storage section 210 outputs the code data of the first layer, second layer, third layer, . . . Nth layer and higher layers of the image A to the image restoration section 220. To go. When the image restoration unit 220 finishes restoring the Nth layer code data of the image B, the image restoration unit 220 writes the restored Nth layer image data of the image B to the transformation image storage unit 250. When the image data of the Nth layer of image B is written to the transformed image storage section 250, the transformation calculation section 260 performs the same transformation process as that performed on the image data of the Nth layer of image A in the forward search. This is performed on the image data of the Nth layer of image B, and the transformed image data is output to the display control unit 240. At this time, the image data of the first layer of image A newly restored by the image restoration section 220 is also input to the display control section 240. Then, the display control unit 240 displays on the image display unit 270 a composite image of the first layer image data of the image A and the modified image data of the restored image data of the image B.

Examples of the composite image displayed as described above are shown in FIGS. 9 and 10. The image displays shown in FIGS. 9 and 10 are examples in which the same affine transformation as shown in FIGS. 7 and 8 described above is performed on the restored image data of image B. In this display, image A is displayed in front of image B.

By the way, when deforming the previous image two or more times, the ratio of the first deformation is made small (in equation (1),
By decreasing the rate of change in the value of m or n) and increasing the rate of subsequent deformations (Equation (1)
By increasing the rate of change in the value of m or n in ), it is possible to express natural image movement similar to the movement of pages when turning a book.

Furthermore, in addition to the embodiments described above, only the deformed image or both the deformed image and the newly restored image can be displayed with a border, and the border and the deformed image can be deformed at the same rate. Even when very similar images of the same type are successively transformed and restored, two images may be easily distinguishable. Further, as another method for easily distinguishing two images, the brightness of the deformed image may be lowered or the brightness of the restored image may be gradually increased.

Furthermore, the encoded image data to be restored is encoded data using two-dimensional discrete cosine transform, but the encoded image data applied to the present invention is It is not limited to this, and the Hadamard transform (Hadamard transform) using a block coding method is not limited to this.
amard Transform), Karhaunen-Love Transform
sform), Haar Transf
It is also possible to restore image data encoded using other orthogonal transformations such as (orm) or other encoding methods.

Furthermore, the present invention is not limited to an image display device that displays each image step by step from a coarse image to a finer image as in the above embodiment, but also displays each image one screen at a time. However, it is also applicable to an image display device that moves to display the next image. Such an image display device is possible by realizing a high-speed processor that successively restores and displays compressed image data at such a high speed that it does not look unnatural to the human eye.

[0071]

According to the present invention, it is possible to specify the forward or reverse image search direction, and at the same time, the image data that has been restored and displayed can be transformed and the next image can be restored. The new image that has been transformed and the newly restored image are combined and displayed, allowing the user to quickly and efficiently use it with the natural feeling of turning the pages of a book. Image searches can be performed.

[Brief explanation of the drawing]

FIG. 1 is a diagram explaining the principle of the present invention.

FIG. 2 is a block diagram illustrating the configuration of an image display device according to an embodiment.

FIG. 3 is a diagram illustrating a configuration example of a search direction detection section.

FIG. 4 is a diagram showing an example of a screen display for specifying a search direction.

FIG. 5 is a block diagram showing the configuration of an image restoration section.

FIG. 6 is a diagram illustrating affine transformation.

FIG. 7 is a diagram illustrating a first example of image display for forward search.

FIG. 8 is a diagram illustrating a second example of image display for forward search.

9 is a diagram illustrating an image display for backward search corresponding to the image display for forward search shown in FIG. 7. FIG.

10 is a diagram illustrating an image display for a backward search corresponding to the image display for a forward search shown in FIG. 8; FIG.

FIG. 11 is a block diagram of an ADCT encoding circuit.

FIG. 12 is a diagram showing how an original image signal is divided into blocks.

FIG. 13 is a diagram showing values of two-dimensional DCT coefficients of each block obtained by a two-dimensional DCT transform unit.

FIG. 14 is a diagram showing values of quantization coefficients of each block obtained by quantizing two-dimensional DCT coefficients.

FIG. 15 is a diagram illustrating the scanning order of zigzag scanning of quantization coefficients.

FIG. 16 is a block diagram illustrating the configuration of a conventional image restoration device.

FIG. 17 is a block diagram illustrating the configuration of a conventional image restoration section.

[Explanation of symbols]

1 Search direction indicating means 2 Restoration means 3　Transformation means 4 Display control means

Claims (8)

[Claims] Claim 1: An image display device for restoring and displaying encoded image data of a plurality of images, comprising: a search direction instructing means (1) for instructing an image search direction; a restoring means (2) for successively restoring the image data of each image encoded in the order according to the designated search direction into display image data; and the restoring means (2)
); a transformation means (3) for transforming the display image data that has been restored according to the search direction instructed by the search direction instruction means (1); and a display that is newly restored by the restoration means (2). a new image obtained from the image data for display, a new image obtained from the image data for display created by the transformation means (3), and a transformation of the previous image obtained from the image data for display created by the transformation means (3). An image display device comprising: display control means (4) for controlling the display so that the images are combined and displayed. 2. The display control means (4) is configured to convert a transformed image of the previous image obtained by the display image data created by the transformation means (3) into a display image that is restored by the restoration means (2). 2. The image display device according to claim 1, wherein the image display device is controlled to be displayed in a superimposed manner as a superimposed image on a new image obtained from the image data. 3. The transforming means (3) transforms the display image data of the previous image that has been restored by the restoring means (2) so that the previous image is displayed in a reduced size. An image display device according to claim 1 or 2. 4. The deformation means (3) performs the deformation in which the previous image is displayed in a reduced size in a plurality of times so that the reduction rate gradually increases, and each time the deformation process is completed, 4. The image display apparatus according to claim 3, wherein the image data for displaying the modified previous image thus created is outputted to the display control means (4). 5. The deforming means (3) is configured to transform the retrieving means (
When the search direction specified in 1) is the forward direction, the transformation process is performed so that the previous image, which is reduced and displayed in the horizontal direction, is displayed moving from left to right or right to left on the screen. On the other hand, when the search direction instructed by the search direction instruction means (1) is the opposite direction, the previous image that is being reduced in the horizontal direction moves in the opposite direction to the forward search. 5. The image display device according to claim 4, wherein the deformation process is performed so that the image is displayed. 6. The deformation means (3) is configured to reduce the size of the previous image in the vertical direction when the search direction instructed by the search direction instruction means (1) is the forward direction when the reduction direction of the previous image is the vertical direction. The transformation processing is performed so that the previous image, which is reduced and displayed in the direction, is displayed moving from top to bottom or from bottom to top on the screen, while the search instructed by the search direction instruction means (1) 4. When the direction is the opposite direction, the transformation process is performed so that the display is moved in a direction opposite to that in the forward search.
The image display device described. 7. The display control means (3) displays a transformed image of the previous image obtained by the display image data created by the transformation means (2) and a display newly restored by the restoration means (1). Claims 1 and 2, characterized in that at least one of the new images obtained by the image data for use is controlled to be displayed with a border.
3. The image display device according to 3, 4, 5 or 6. 8. The display control means (3) controls the brightness of the deformed image of the previous image obtained by the display image data created by the deformation means (2) and the restoration means (1).
), the new image is controlled to be displayed differently from the brightness of the new image obtained by the newly restored display image data.
8. The image display device according to 6 or 7.