JPH08340552A - Digital image thinning method and interpolation method, and transmission apparatus using the same - Google Patents

Abstract

(57) [Abstract] [Purpose] When transmitting digital image data, a thinning method and an interpolation method for a digital image, which can perform thinning processing and interpolation processing by a simple calculation taking human visual characteristics into consideration, and the same. It is to provide a transmission device using. A data thinning section 81 uses four color difference data in advance as the color difference data as the brightness data of a digital image obtained by reducing the average value of the four brightness data extracted by the CPU 7 during the thinning process. The determined color difference data is typically assigned as the color difference data of the reduced digital image. On the other hand, the data interpolation unit 82 refers to the four pieces of brightness data selected by the CPU 7 when performing the interpolation process, and creates the brightness data of the enlarged digital image.
As for the color difference data, one color difference data is duplicated to create color difference data of a digital image for one frame.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image thinning method and an interpolation method, and a transmission apparatus using the same, and more specifically, a method of thinning a digital image by different methods for luminance data and color difference data. The present invention relates to a decimation method for performing the interpolation, an interpolation method for performing an interpolation process, and a transmission apparatus using them.

[0002]

2. Description of the Related Art Conventionally, when transmitting digital images,
In order to reduce the amount of data to be transmitted, digital image data is thinned out and transmitted. Therefore, when the digital image data is received, it is necessary to interpolate the digital image data for one frame. At this time, as a thinning and interpolation method for the digital image to be applied, for example, the nearest neighbor method or Japanese Patent Laid-Open No. 04-330858.
The method etc. which are disclosed by the gazette are mentioned.

FIG. 6 is a reference diagram for facilitating the understanding of the conventional digital image thinning processing by the nearest neighbor method.
FIG. 6A is a memory map diagram showing a part of the YUV data 61 of one frame of the digital image. The YUV data 61 is Y ₀₀ , Y ₀₁ , ...
And as U and V data representing color difference, U ₀₀ , U ₀₁ , ...
And V ₀₀ , V ₀₁ , ... And respectively. Figure 6
(B) is a partial frame 62 of Y data representing luminance,
FIG. 7 is a schematic diagram of partial frames 63 and 64 of U and V data representing color differences. These partial frames 62, 63
And 64 are extracted from the YUV data 61 in units of 4 pixels, and the partial frame 62 is (Y ₀₀ ,
Y ₀₁ , Y ₁₀ , Y ₁₁ ) and the partial frame 63 is (U ₀₀ ,
U ₀₁ , U ₁₀ , U ₁₁ ) and the partial frame 64 is (V ₀₀ ,
V ₀₁ , V ₁₀ , V ₁₁ ). FIG. 6C is a schematic diagram of the reduced partial frame 65. Reduced partial frame 6
5 is created by thinning out the partial frames 62, 63, and 64 to 1/4, and Y '
And a ₀₀ and U _'00 and V' _00.

First, when thinning a digital image into quarters, partial frames 62, 63 and 64 are extracted from the YUV data 61 for each YUV data. Then, from a partial frame 62 Y _00, from the partial frame 63 U _00, assigned to V ₀₀ from the partial frame 64, creating a reduced portion frame 65 therefrom. Such thinning processing is performed by YU for one digital image frame.
Repeated with respect to the V data, reduced digital image data that has been thinned to 1/4 is created.

FIG. 7 is a digital image by the conventional nearest neighbor method.
FIG. 6 is a reference diagram for facilitating understanding of the interpolation processing of FIG. Figure
7 (a) is YUV data of one frame of digital image 7
2 is a memory map diagram showing a part of FIG. YUV data
Reference numeral 71 denotes Y'as Y data representing luminance.₀₀, Y ’₀₁,…
And U'as V and V data representing the color difference₀₀,
U ’ ₀₁, ... and V '₀₀, V '₀₁,, and have respectively
ing. FIG. 7B shows a partial frame of Y data representing brightness.
Section 72 and partial frame of U and V data representing color difference
73 is a schematic view of the frames 73 and 74. FIG. These, partial frame
72, 73 and 74 selected from YUV data 71
The individual Y ', U'and V'data referenced
It is obtained as a result of interpolation processing. FIG. 7C shows
Regarding all YUV data 71 of one frame of digital image
Expanded YUV data obtained as a result of interpolation processing
It is a memory map figure which shows a part of 75. YUV day
Data 75 is Y data.₀₀, Y₀₁,,, U and
U as V data₀₀, U₀₁, And V₀₀, V₀₁,…When
have.

To interpolate a digital image four times,
The Y data Y ′ ₀₀ selected from the YUV data 71 is referred to, and the Y data (Y ₀₀ , Y ₀₁ , Y ₁₀ ,
The partial frame 72 is generated by setting Y ′ ₀₀ as Y ₁₁ ). That is, Y ₀₀ = Y ₀₁ = Y ₁₀ = Y ₁₁
= Y 'is _00. Similar processing is performed on U data and V data to generate partial frames 73 and 74. Such interpolation processing is repeated for one frame of YUV data, and the interpolation processing is performed four times to generate YUV data 75 which is an enlarged digital image.

FIG. 8 is a flowchart showing a digital image thinning process according to the adaptive digital image enlargement / reduction method disclosed in Japanese Patent Laid-Open No. 04-330858. Hereinafter, with reference to FIG. 8, a digital image thinning process according to an adaptive digital image enlargement / reduction method will be described.

First, a low-pass filter is designed by the FIR digital filter method (step S8).
1). Next, a cubic spline function for interpolating the impulse response of the designed low-pass filter is obtained (step S82). This creates an optimal interpolation filter for creating the reduced digital image data. When digital image data is input to this interpolation filter, the digital image data is filtered (step S83), and reduced digital image data is generated (step S84).

FIG. 9 is an enlarged view of adaptive digital image data disclosed in Japanese Patent Laid-Open No. 04-330858.
It is a flowchart which shows the interpolation process which concerns on the reduction method.
Hereinafter, the interpolation process related to the adaptive enlargement / reduction method of digital image data will be described with reference to FIG.

First, a high-pass filter is designed by the FIR digital filter method (step S9).
1). Next, a cubic spline function for interpolating the impulse response of the designed high-pass filter is obtained (step S92). This creates an optimal interpolation filter for creating the enlarged digital image data. When digital image data is input to this interpolation filter, the digital image data is filtered (step S93), and enlarged digital image data is generated (step S94).

[0011]

Since the nearest neighbor method is easy to operate as apparent from the above, it can be easily realized on a small-scale hardware. However, as the calculation of the thinning process, reduced digital image data is created from the extracted four Y, U, and V data by using one predetermined Y, U, and V data. Further, as the calculation of the interpolation processing, the digital image data which is magnified four times is created from the referenced one piece of Y ′, U ′ and V ′ data. This causes a problem that the image quality of the reduced digital image data and the enlarged digital image data is significantly deteriorated.

In the adaptive digital image enlargement / reduction method, the image quality of the reduced digital image data and the enlarged digital image data is improved as compared with the nearest neighbor method, but the design of the FIR filter and the cubic spline are used. Since the calculation such as the calculation of the function is complicated, there is a problem that the calculation time is required for a small-scale hardware.

Further, according to the above adaptive digital image data enlarging / reducing method, the same processing as the Y data representing the luminance is performed on the U and V data representing the color difference which is difficult to be visually recognized by human eyes. Therefore, the characteristics of the YUV data format of the digital image are not taken into consideration.

Therefore, an object of the present invention is to perform a thinning-out method and an interpolation method for digital image data capable of executing thinning-out processing and interpolation processing of digital image data by simple calculation and creating high-quality digital image data, and those methods. It is to provide a transmission device using.

[0015]

According to a first aspect of the present invention, there is provided a method for thinning out digital image data in which each pixel is separated into luminance data and color difference data. Is divided into a plurality of partial images including a plurality of pixels, and in each of the divided partial images, the average value of the luminance data of the plurality of pixels is assigned as a representative value of the luminance data of the corresponding partial image, and the divided partial images are divided. In the partial image, the color difference data of one of the plurality of pixels is assigned as a representative value of the color difference data of the partial image.

According to a second aspect of the present invention, there is provided a method for interpolating reduced digital image data composed of a plurality of pixels to extend each pixel into a plurality of pixels, the luminance data being extended. Refer to the target pixel and its neighboring pixels, assign the brightness data of the target pixel for expansion to the pixel that should be displayed first, and refer to the other pixels The average value of the luminance data of is assigned, the color difference data is referred to the pixel to be expanded, and for all the pixels to be displayed,
The feature is that color difference data of pixels to be expanded is assigned.

According to a third aspect of the present invention, there is provided an apparatus for transmitting digital image data, comprising thinning means for thinning out digital image data, interpolation means for obtaining digital image data for one frame, and digital means. The thinning means includes a transmitting / receiving means for transmitting / receiving image data and a control means for controlling transmission / reception of digital image data.
Digital image data for a frame is divided into a plurality of partial images each including a plurality of images, and in each of the divided partial images, the average value of the luminance data of a plurality of pixels is compared with the luminance data of the corresponding partial image. In each of the divided partial images assigned as a representative value, digital image data to be assigned and transmitted as a representative value of the color difference data of the partial image corresponding to any of the color difference data of a plurality of pixels is created, and the interpolation means Refers to the pixel to be expanded and its neighboring pixels for the brightness data, allocates the brightness data of the pixel to be expanded to the pixel to be displayed first, and The average value of the luminance data of the referred pixel is assigned to the pixel, and the color difference data is displayed by referring to the pixel to be expanded. For all pixels can, by assigning the color difference data of the pixel to be an extension of the target,
It is characterized in that the received digital image data is interpolated.

[0018]

According to the first aspect of the present invention, when a thinning process is performed for separating a digital image into luminance data and color difference data and creating reduced digital image data, a plurality of pixels including luminance data and color difference data are used. Is extracted. Regarding the brightness data, the average value of the plurality of brightness data is typically assigned as the brightness data of the reduced digital image. Regarding the color difference data, predetermined color difference data is typically assigned from the plurality of color difference data as the color difference data of the reduced digital image. Therefore, it is possible to create reduced digital image data by a simple calculation. Further, since the thinning processing is performed in consideration of the human visual resolution characteristics, it is possible to create high-quality reduced digital image data.

According to the second aspect of the invention, the digital image separated for the luminance data and the color difference data is interpolated to generate digital image data for one frame. Regarding the luminance data, by referring to the luminance data of the pixel to be expanded and the pixels adjacent to it, the value of the pixel to be expanded is directly assigned to the pixel to be displayed first, The average value of the referred pixels is assigned to the pixels of the other pixels. Regarding the color difference data, the color difference data of the pixel to be expanded is referred to, and this is assigned to all the color difference data to be displayed.
Therefore, the reduced digital image data can be expanded to digital image data for one frame by a simple calculation. In addition, since the resolution characteristics of human vision are taken into consideration, high quality enlarged digital image data can be created.

In the invention of claim 3, when the transmission device transmits the reduced digital image data, the control means extracts a plurality of pixels including the luminance data and the color difference data. The thinning means typically assigns the average value of the extracted plurality of brightness data to the brightness data as the brightness data of the reduced digital image. Regarding the color difference data, the reduced digital image data is created by assigning the predetermined color difference data as the color difference data of the reduced digital image as a representative from the plurality of extracted color difference data. When the thinning process of one frame of the digital image is completed, the control means transmits the reduced digital image data from the transmitting / receiving means. On the other hand, when the transmission / reception unit of the transmission device receives the digital image, the control unit selects a pixel to be expanded including the luminance data and the color difference data used for the interpolation processing. Further, with respect to the brightness data, the control unit also selects pixels adjacent to the pixel to be expanded. The interpolation means refers to the brightness data of the selected pixel with respect to the brightness data,
The value of the pixel to be expanded is directly assigned to the pixel to be displayed first, and the average value of the referred pixels is assigned to the other pixels. Regarding the color difference data, the color difference data of the pixel to be expanded is referred to, and this is assigned to all the color difference data to be displayed. When the control means creates digital image data for one frame, the control means outputs this digital image data. Therefore, it is possible to create high-quality digital image data by a simple calculation and transmit it.

[0021]

1 is a block diagram showing a detailed configuration of a digital image transmitting apparatus according to an embodiment of the present invention. FIG.
In the transmission device 1, the transmission device 1 uses an ADC (Analog-Di)
digital converter 2, digital decoder 3, first frame memory 4, digital encoder 5, DAC (Digital-Analog Con).
verter) 6, CPU 7, ROM (Read
Only Memory 8 and RAM (Random)
The access memory 9 includes a second frame memory 10, a modem 11, and a data bus 12.

The ADC 2 samples a video signal transmitted from an image input device (not shown) such as a camera installed in the preceding stage of the ADC 2 and quantizes it to convert it into a digital video signal. Here, the video signal is a signal of three primary colors represented by red, green and blue.
The digital decoder 3 separates the digital video signal into luminance data representing the luminance of the image and color difference data which is a signal obtained by subtracting the luminance data from each of the three primary color signals. The digital decoder 3 creates so-called YUV data. The YUV data has Y data that is luminance data and U and V data that is color difference data.
The first and second frame memories 4 and 10 are composed of a storage device having a capacity capable of storing at least one frame of YUV data. The digital encoder 5 performs a process reverse to that of the digital decoder 3 and synthesizes the YUV data into three primary color signals. DAC6 is
The processing opposite to that of the ADC 2 is performed to convert the input digital three primary color signals into analog three primary color signals. DAC
An image display device (not shown) such as a monitor is connected to the subsequent stage of 6 and displays an image based on the three primary color signals. The CPU 7 controls the operation of the transmission device 1. In addition,
The control performed by the CPU 7 will be described in detail later. R
The OM 8 includes a data thinning unit 81 and a data interpolation unit 82. The data thinning unit 81 creates the thinned digital image data transmitted by the transmission device 1 under the control of the CPU 7. On the other hand, the data interpolation unit 82 uses C
Based on the instruction from the PU 7, the digital image data interpolated from the digital image data received by the transmission device 1 is created. The RAM 9 includes the data thinning unit 81 described above.
Alternatively, it is used as a work area of the thinning-out or interpolation processing performed by the data interpolation unit 82, and temporarily stores the digital image data created by the thinning-out or interpolation processing. The modem 11 is connected to an external device via a transmission line and performs modulation / demodulation of a signal transmitted between the transmission device 1 and the external device. The data bus 12 includes first and second frame memories 4 and 10, a CPU 7 and a ROM.
8, the RAM 9, and the modem 11 are connected to each other so that data can be transmitted and received bidirectionally.

FIG. 2 is a reference diagram for facilitating the understanding of the thinning process of YUV data controlled by the CPU 7.

FIG. 2A is a diagram showing a memory map of the YUV data 21 corresponding to one frame stored in the first frame memory 4. In FIG. 2A, YU
The V data 21 is each Y, U and V data separated by the digital decoder 3, that is, Y _{00 to} Y _nm and U _{00 to} U.
_nm and V _{00 to} V _nm , respectively. Also, Y,
The U and V data are in order to simplify the description below.
It is assumed that the images are stored in the first frame memory 4 in the order of scanning by the image input device or the image display device.

FIG. 2B shows a partial frame 22 of Y, U and V data extracted from the first frame memory 4.
FIG. 23 shows 23 and 24. In FIG. 2 (b),
The partial frame 22 is a partial frame of Y data,
_{(Y 00, Y 01, Y} 10, Y 11) has a. The partial frame 23 is a partial frame of U data, and (U ₀₀ ,
U ₀₁ , U ₁₀ , U ₁₁ ). The partial frame 24 is a partial frame of V data, and is (V ₀₀ , V ₀₁ , V ₁₀ ,
V ₁₁ ). These partial frames 22, 23 and 24 are extracted in units of 2 pixels × 2 pixels. In addition,
In FIG. 2B, the CPU 7 first extracts Y from the YUV data 21 stored in the first frame memory 4,
Partial frames 22, 23 and 24 of U and V data
Is illustrated.

FIG. 2 (c) is a diagram showing reduced partial frames 25, 26 and 27 of YUV data. The reduced partial frame 25 is a reduced partial frame of Y data, the reduced partial frame 26 is a reduced partial frame of U data, and the reduced partial frame 27 is a reduced partial frame of V data.
Each of these is reduced to one pixel. It is to be noted that FIG. 2C shows reduced partial frames 25, 26 obtained as a result of thinning out the partial frames 22, 23, 24 of the Y, U and V data which the CPU 7 first extracted, which will be described later. And 27 are shown.

FIG. 2D is a diagram showing a reduced frame 28 obtained by thinning the YUV data 21 into quarters.
The reduced frame 28 is the thinned Y, U and V.
The _{data, Y '00 ~Y' (n} / 2) (m / 2), U '00 ~U'
_{(n / 2) (m /} 2) and _{V '00 ~V' (n /} 2) has a _{(m / 2).}

FIG. 3 is a flow chart showing the thinning process of YUV data controlled by the CPU 7. Figure 1 below
A YUV controlled by the CPU 7 with reference to FIGS. 2 and 3.
The data thinning process will be described in detail. In the following description, Y, U extracted by the CPU 7 first
The thinning process for the partial frames 22, 23 and 24 of V data and V data will be described in detail.

An analog three primary color signal for one frame of image sequentially transmitted from an image input device (not shown) such as a camera is input to the ADC 2 and converted into a digital three primary color signal by the ADC 2. This digital three primary color signal is
Input to the digital decoder 3. Digital decoder 3
Separates the digital three primary color signals into YUV data and outputs these YUV data to the first frame memory 4. At this time, the CPU 7 gives an address to each YUV data output from the digital decoder 3, and sequentially stores the YUV data in the first frame memory 4.

When the CPU 7 stores the YUV data for one frame in the first frame memory 4 (see FIG. 2A), the CPU 7 stores the image from the YUV data 21 in order to create the partial frame 22 from the first frame memory 4. The two Y data items correspond to the first and second scanning lines in the horizontal scanning direction decomposed by the input device and are decomposed into the first and second scanning lines in these scanning lines, respectively ( Y ₀₀ , Y ₀₁ , Y ₁₀ , Y ₁₁ ) are extracted, and the partial frame 2
Create 2. Similar to this procedure, partial frames 23 and 24 are also created for U and V data (step S31, see FIG. 2B). The CPU 7 stores the created partial frames 22, 23 and 24 in a predetermined address position of the RAM 9 (step S32).

Next, the data thinning section 81 of the ROM 8
In order to create the reduced partial frame 25 of Y data, the average value of each Y data is assigned as the reduced partial frame 25. _{That, Y '00 = (Y 00} + Y 01 + Y 10 + Y 11)
By assigning / 4, the reduced partial frame 25 of Y data is created (step S33). And CP
The U7 stores the reduced partial frame 25 at a predetermined address position in the RAM 9.

Next, the data thinning section 81 of the ROM 8
In order to create the reduced partial frame 26 of U data, U which is the U data decomposed first by the image input device from the U data of the partial frame 23 as the reduced partial frame 26.
Assign ₀₀ . That is, by assigning the U _'00 = U _00, creating a reduced portion frame 26 (step S34). Then, the CPU 7 stores the reduced partial frame 26 in a predetermined address position of the RAM 9.

Next, the data thinning section 81 of the ROM 8
In order to create the reduced partial frame 27 of V data, V which is the V data decomposed first by the image input device from the V data of the partial frame 24 as the reduced partial frame 27.
Assign ₀₀ . That is, by assigning the V _'00 = V _00, creating a reduced portion frame 27 (step S35). Then, the CPU 7 stores the reduced partial frame 25 in a predetermined address position of the RAM 9.

As described above, the RAM 9 stores the reduced partial frames 25, 26 and 27 of the Y, U and V data created by the data thinning section 81. The CPU 7 controls the three reduced partial frames 25,
A predetermined address value is given to 26 and 27 and stored in the second frame memory 10 (step S36).

The CPU 7 determines whether or not the creation of the reduced partial frame of the YUV data for one frame is completed by referring to the address value of the YUV data stored in the first frame memory 4 (step S37). The CPU 7 extracts a new partial frame based on this determination result (step S38). At this time, the new partial frame extracted by the CPU 7 is the first and second scanning lines in the horizontal scanning direction, and the third and fourth scanning lines are decomposed by the image input device. YU
It is created from V data. The CPU 7 creates a reduced partial frame using this partial frame. After that, CPU
In step 7, the same procedure as the operation in steps S32 to S37 is repeated to create a reduced partial frame using the YUV data included in the first and second scanning lines in the horizontal scanning direction. Next, the CPU 7 creates a reduced partial frame using the YUV data included in the third and fourth scanning lines. After that, one frame of YUV
The above steps S32 to S37 are repeated until a reduced partial frame of data is created.

As described above, one frame of YUV
When the creation of the reduced partial frame of data is completed, the reduced frame 28 of YUV data for one frame is stored in the second frame memory 10 (FIG. 2).
(See (d)). When the CPU 7 confirms that the creation of the reduced frame 28 of the YUV data for one frame is completed by the above-described determination method (step S37), the CPU 7
7 transmits the reduced frame 28 from the modem 11 to the external device via the transmission path.

Next, the operation of each part when the transmission device 1 functions as an interpolation device will be described in detail.

FIG. 4 is a reference diagram for facilitating understanding of the YUV data interpolation processing controlled by the CPU 7. FIG. 4A is a diagram showing a memory map of the reduced frame 41 of the YUV data for one frame stored in the second frame memory 10. In FIG. 4A, the reduced frame 41 is created by an external device and is transmitted to the modem 1 via a transmission line.
Y ', which is the individual Y, U and V data input in 1.
_{00 ~Y '(n / 2)} (m / 2), U' 00 ~U '(n / 2) (m / 2) and _{V' 00 ~V '(n /} 2) have a _{(m / 2)} are doing. In the present embodiment, the reduced frame 28 and the reduced frame 41 that have been subjected to the thinning process will be described as being the same.

FIG. 4B shows partial frames 42, 4 of Y, U and V data created in the work area of the RAM 9.
FIG. 3 shows 3 and 44. In FIG. 4B, the partial frame 42 is a partial frame of Y data, and (Y
₀₀ , Y ₀₁ , Y ₁₀ , Y ₁₁ ). Partial frame 4
3 is a partial frame of U data, which is (U ₀₀ , U ₀₁ , U
₁₀ , U ₁₁ ). The partial frame 44 is a partial frame of V data, and has (V ₀₀ , V ₀₁ , V ₁₀ , V ₁₁ ). In addition, what is illustrated in FIG.
The CPU 7 first creates Y from the Y, U and V data for one frame stored in the first frame memory,
U and V data subframes 42, 43 and 44
Is illustrated. FIG. 4C is a diagram showing a memory map of YUV data 45 obtained as a result of the interpolation processing described later.

FIG. 5 is a flow chart showing the YUV data interpolation processing controlled by the CPU 7. The YUV data interpolation processing controlled by the CPU 7 will be described in detail below with reference to FIGS. 1, 4 and 5. In the following description, the interpolation process for the partial frames 42, 43, and 44 of the Y, U, and V data first created by the CPU 7 will be described in detail.

Also in the external device, the reduced frame 41 is created in the same procedure as the thinning process. This reduced frame 41 is input to the modem 11 via the transmission line,
It is stored in the second frame memory 10 from the modem 11 (see FIG. 4A). At this time, the CPU 7 assigns an address to each Y ′, U ′, and V ′ data of the reduced frame 41 and sequentially stores them in the second frame memory 10.

When the reduced frame 41 is stored in the second frame memory 10, the CPU 7 stores the partial frames 42 and 43.
As an initial setting of Y ′, U ′ and V ′ data to be referred to when creating Y and Y, (Y ′ ₀₀ , Y ′ ₀₁ , Y ′
_10, 'and _11), U' Y _00, selects the V _'00 (Figure 4
See (a), step S51).

The CPU 7 controls the partial frame 42 of the Y data.
Of the reduced frame 41 (Y _'00 ,
_{Y '01, Y' 10,} Y '11) referring to. Data interpolator 8
2 first assigns Y ′ ₀₀ as Y ₀₀ of the partial frame 42. That is, Y ₀₀ = Y _'00. Next, as Y ₀₁ of the partial frame 42, Y ′ ₀₀ and Y ′ ₀₁ are referenced to calculate the average value of the two, and this average value is assigned. That is, Y ₀₁ = (Y _'00 + Y' ₀₁ ) / 2. next,
As Y ₁₀ of the partial frame 42, Y ′ ₀₀ and Y ′ ₁₀ are referred to, the average value of both is calculated, and this average value is assigned. That is, Y ₁₀ = (Y _'00 + Y' ₁₀ ) / 2.
Next, as Y ₁₁ of the partial frame 42, Y ′ ₀₀ and Y ′ ₀₁
And Y ′ ₁₀ and Y ′ ₁₁ are calculated and the average value is calculated, and this average value is assigned. That is, Y ₁₁ = (Y '
_{00 + Y '01 + Y'} 10 + Y '11) / 4. As a result, the partial frame 42 of Y data is created (step S52). The CPU 7 stores the created four Y data at a predetermined address position of the RAM 9.

Next, the data interpolating unit 82 refers to the U _'00 of the reduced frame 41 to create a partial frame 43 of U data. The data interpolation unit 82 assigns U ′ ₀₀ to U ₀₀ , U ₀₁ , U ₁₀ and U ₁₁ of the partial frame 43. That is, U ₀₀ = U ₀₁ = U ₁₀ = U ₁₁ = U ′ ₀₀ . As a result, the partial frame 43 of the U data is created (step S53). CPU7
The created four U data are stored in a predetermined address position of the RAM 9.

Next, the data interpolating unit 82 refers to the V _'00 of the reduced frame 41 to create a partial frame 44 of V data. The data interpolation unit 82 assigns V ′ ₀₀ to V ₀₀ , V ₀₁ , V ₁₀ and V ₁₁ of the partial frame 44. That is, V ₀₀ = V ₀₁ = V ₁₀ = V ₁₁ = V ′ ₀₀ . As a result, the partial frame 44 of V data is created (step S54). CPU7
The created four V data are stored in a predetermined address position of the RAM 9 (step S54).

When the operations of steps S52 to S54 are completed, the CPU 7 gives a predetermined address value to each of the created YUV data and then stores the YUV data in the first frame memory 4 (step S55).

By the operations of the steps S52 to S55, the first and second scanning lines in the horizontal scanning direction, which correspond to the first and second scanning lines in the vertical scanning direction, respectively, 4 This means that YUV data for pixels has been created.

The CPU 7 determines whether or not YUV data for one frame has been created by referring to the address values of the Y ', U'and V'data stored in the second frame memory 10 (step S56). . The CPU 7 updates the Y ', U', and V'data to be referred to in order to create a new partial frame, based on this determination result (step S57). Next, the Y ′ data selected by the CPU 7 to create the YUV data is (Y ′ ₀₁ , Y ′ ₀₂ ,
Y is _'11, Y' _12), U 'data U' is _01,
V 'data V' is _01. Additionally, the following Y ', U' data 'Y CPU 7 selects the updating data' which and V is _{(Y '02, Y' 03} , Y '12, Y' 13), U '
The data is _U'02 and the _V'data is _V'02 . In addition, when four YUV data corresponding to the third and fourth scanning lines in the horizontal scanning direction and corresponding to the first and second scanning lines in the vertical scanning direction are created. Y CPU 7 selects the 'data, (Y' _{10,
Y '11, Y' 20,} ' a _21), U' Y data U _'10
In it, V 'data V' is _10. After that, step S
At 57, the Y'U 'and V'data are updated with the above-described periodicity. In addition, as for the Y ′ data at the right end and the lower end in FIG. The processing performed by the CPU 7 in such a situation will be described below.

The last Y CPU 7 is selected when the interpolation process 'data, for example in the 1st and the 2nd and scan lines according to a horizontal direction scanning, exceptionally Y' according to the horizontal scanning direction _{0 (m / 2)} , U ′ data is U ′ _{0 (m / 2)} , and V ′ data is V ′ _{0 (m / 2)} . Therefore, for U and V data, a partial frame is created in the same procedure as in the normal interpolation process. for that reason,
The procedure for creating a partial frame of Y data will be described below. At this time, the data interpolation unit 82 allocates Y ′ _{0 (m / 2)} as all the Y data of the partial frame to be created. After that, the final interpolation processing in the horizontal scanning direction is performed using the above method.

Further, the Y'data selected by the CPU 7 in the final interpolation processing in the vertical scanning direction is, for example, in the first and second scanning lines in the vertical direction,
Exceptionally, Y ' _{(n / 2) 0} , U'data is U' _{(n / 2) 0} , and V'data is V ' _{(n / 2) 0} . Therefore, for U and V data, a partial frame is created in the same procedure as in the normal interpolation process. for that reason,
The procedure for creating a partial frame of Y data will be described below. At this time, the data interpolation unit 82 allocates Y ′ _{(n / 2) 0} as all the Y data of the partial frame to be created. After that, the final interpolation processing related to the vertical scanning is performed using the above method.

Thereafter, the CPU 7 sequentially selects the Y ', U', and V'data to be referred from the second frame memory 10 in the same manner as above, and creates the YUV data. The CPU 7 repeats the operations in steps S51 to S54 until one frame of YUV data is created. 1
When the creation of the YUV data for one frame is completed, the YUV data for one frame is stored in the first frame memory 4. When the CPU 7 confirms that the creation of the YUV data for one frame is completed by the determination method described above (step S56), it outputs the YUV data to the digital encoder 5. The digital encoder 5 synthesizes YUV data into three primary color signals and
Output to 6. The DAC 6 converts the three primary color signals, which are digital signals, into analog signals. An image display device (not shown) installed in the subsequent stage of the DAC 6 displays an image by the three primary color signals which are analog signals.

As described above, according to this embodiment, when the frame of the color digital image is thinned out, the CP
U7 extracts four Y data from the frame and calculates the average value of them. The CPU 7 assigns the average value obtained as a result as Y data of the reduced frame. Regarding U and V data, four U and V data are extracted from the frame, and one predetermined U and V data is representatively assigned from these. This can simplify the calculation during the thinning process.

When performing the interpolation process, the CPU 7
The four Y ′ data are selected from the reduced frame 41, and the data interpolating unit 82 refers to these and performs the calculation as described above to create the Y data. In this way, during the interpolation processing, the interpolation processing is performed in order to reduce the difference in luminance between the Y data adjacent in the horizontal scanning direction and the vertical scanning direction of the image.
Further, Y data is created by utilizing the correlation between the image in the horizontal scanning direction and the vertical scanning direction. This makes it possible to create high-quality digital images. Also,
For U and V data, the data interpolator 82 uses C
PU7 refers to one piece of U'and V'data from the selected frame and creates U and Y data. As a result, the calculation during the interpolation processing can be simplified.

Further, when an image obtained from the image input device is displayed on the image display device, as is apparent from the above, the image input device (not shown) such as a camera is provided with three primary color signals which are analog signals. (Input video signal) to ADC2
To enter. The ADC 2 converts the three primary color signals that are analog signals into digital signals and outputs the digital signals to the digital decoder 3. The digital decoder 3 separates the three primary color signals into YUV data. This YUV data is temporarily stored in the first frame memory 4. Digital encoder 5
Receives YUV data from the first frame memory 4, synthesizes the YUV data into three primary color signals, and outputs the signals to the DAC 6.
The DAC 6 converts the three primary color signals, which are digital signals, into digital signals and outputs the digital signals to an image display device (not shown).
The image display device displays an image based on the three primary color signals.

In this embodiment, the RAM 9 and the second
Although it has a separate function from the frame memory 10,
It is also possible to configure both with the same memory and have both functions function in this memory.

[0056]

According to the invention of claim 1, the fact that the resolution characteristic with respect to the human color difference is sufficiently lower than the resolution characteristic with respect to the luminance is utilized to obtain the luminance data as follows.
The average value of a plurality of brightness data is assigned as the brightness data of the reduced digital image. As for the color difference data, predetermined color difference data is assigned from the plurality of color difference data as color difference data of the reduced digital image. This makes it possible to create a reduced digital image by a simple calculation.

According to the second aspect of the invention, the fact that the resolution characteristic with respect to human color difference is sufficiently lower than the resolution characteristic with respect to luminance is utilized, and the luminance data is predetermined from a plurality of luminance data. In this way, the brightness data of the expanded digital image is created. Therefore, the brightness data is interpolated to reduce the difference in brightness between the Y data adjacent in the horizontal scanning direction and the vertical scanning direction of the image. At the same time, Y data is created by utilizing the correlation between the image in the horizontal scanning direction and the vertical scanning direction. Therefore, a high quality digital image can be created. Regarding color difference data, a predetermined number of color difference data to be expanded are duplicated to create color difference data of an expanded digital image. This makes it possible to create a high-quality magnified digital image with a simple calculation.

According to the third aspect of the present invention, the control means causes the thinning means to thin the digital image data when transmitting the reduced digital image data, and causes the interpolating means to perform the thinning processing when the digital image data is received. Interpolate digital image data. Therefore, the reduced or expanded digital image data can be created by a simple calculation, and thus the transmission device for the digital image data can be configured with a small-scale hardware. Moreover, since the resolution characteristics for human color difference are sufficiently lower than the resolution characteristics for luminance, thinning and interpolation processing of digital images are used to create high-quality reduced and enlarged digital images. can do.

[Brief description of drawings]

FIG. 1 is a block diagram showing a detailed configuration of a transmission device according to an embodiment of the present invention.

FIG. 2 is a reference diagram for facilitating understanding of thinning processing in the transmission apparatus shown in FIG.

FIG. 3 is a flowchart showing a thinning process in the transmission device shown in FIG.

4 is a reference diagram for facilitating understanding of an interpolation process in the transmission apparatus shown in FIG. 1. FIG.

5 is a flowchart showing an interpolation process in the transmission device shown in FIG.

FIG. 6 is a reference diagram for facilitating understanding of a conventional digital image thinning process according to the nearest neighbor method.

FIG. 7 is a reference diagram for facilitating understanding of conventional digital image data interpolation processing according to the nearest neighbor method.

FIG. 8 is a flowchart showing thinning processing according to an adaptive digital image data enlargement and reduction method.

FIG. 9 is a flowchart showing an interpolation process according to an adaptive digital image data enlargement and reduction method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Transmission apparatus 2 ... ADC 3 ... Digital decoder 4 ... 1st frame memory 5 ... Digital encoder 6 ... DAC 7 ... CPU 8 ... ROM 81 ... Decimation program storage 82 ... Interpolation program storage 9 ... RAM 10 ... 2nd frame Memory 11 ... Modem 12 ... Data bus

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location H04N 1/46 H04N 1/40 D 7/24 1/46 Z 7/13 Z

Claims (3)

[Claims] 1. A method for thinning out digital image data in which each pixel is separated into luminance data and color difference data, wherein one frame of digital image data is divided into a plurality of partial images each including a plurality of pixels. In each of the divided partial images, the average value of the luminance data of multiple pixels is assigned as the representative value of the luminance data of the corresponding partial image, and one of the multiple pixels in each divided partial image is assigned. The method of thinning out a digital image, wherein the method is assigned as a representative value of the color difference data of the partial image corresponding to the color difference data. 2. A method for interpolating reduced digital image data composed of a plurality of pixels to expand each pixel into a plurality of pixels, wherein luminance data is a pixel to be expanded. , And pixels adjacent to it, the brightness data of the pixel to be expanded is assigned to the pixel to be displayed first, and the brightness data of the referenced pixel is assigned to other pixels. An average value is assigned, with respect to the color difference data, the pixel to be expanded is referred to, and the color difference data of the pixel to be expanded is allocated to all the pixels to be displayed. Interpolation method for digital images. 3. An apparatus for transmitting digital image data, comprising thinning means for thinning the digital image data, interpolation means for obtaining the digital image data for one frame, and the digital image data. And a control unit for controlling the transmission and reception of the digital image data, wherein the thinning unit converts one frame of digital image data into a plurality of partial images each including a plurality of images. In each of the divided partial images, the average value of the luminance data of multiple pixels is assigned as the representative value of the luminance data of the corresponding partial image, and one of the multiple pixels in each divided partial image is assigned. Color difference data of the digital image data to be assigned and transmitted as the representative value of the color difference data of the corresponding partial image. The interpolation means refers to a pixel to be expanded and a pixel in the vicinity thereof for luminance data, and a pixel to be expanded for the pixel to be displayed first. Luminance data, assign the average value of the luminance data of the referenced pixels to other pixels, refer to the pixel to be expanded for color difference data, and refer to all the pixels to be displayed. The digital image transmission device is characterized in that the color difference data of the pixel to be expanded is assigned and the received digital image data is interpolated.