JP2552381B2 - Image signal compression coding device - Google Patents

Description

TECHNICAL FIELD The present invention relates to an image signal compression encoding apparatus, and more particularly to an image signal compression encoding apparatus for performing compression encoding when recording image data forming one screen on a recording medium.

BACKGROUND ART When storing digital image data such as image data taken by a digital electronic still camera in a memory, compression encoding by various methods is performed in order to reduce the amount of data and the storage capacity of the memory. Has been done. In particular, two-dimensional orthogonal transform coding is widely used because it can perform coding at a large compression rate and can suppress image distortion due to coding.

In such a two-dimensional orthogonal transform coding, image data forming one screen is divided into a predetermined number of blocks on the screen, and the image data in each block of the divided screen is two-dimensional orthogonal. To be converted. The orthogonally transformed image data, that is, the transform coefficient, is compared with a predetermined threshold value, and the portion below the threshold value is truncated (coefficient truncation). As a result, the conversion factor below the threshold is
It is processed as 0 data. Next, the transform coefficient subjected to coefficient truncation is divided by a predetermined quantization step value, that is, a normalization coefficient, and quantization by a step width, that is, normalization is performed. This gives the value of the conversion factor,
That is, the dynamic range of the amplitude can be suppressed.

In such a two-dimensional orthogonal transform coding, when the above-mentioned normalization coefficient is normalized with a constant value and coded, the amount of coded data differs depending on the image data, and the data is recorded in the memory. It was inconvenient.

That is, when the data is normalized by using a certain normalization coefficient and coded, the amount of coded data in the image data including many high frequency (AC) components is large, and the low frequency (DC) is large.
The image data including many components has a small amount of encoded data. The difference in the amount of encoded data is 5 to
Since it can reach up to 10 times, it is inconvenient when recording in a fixed capacity memory.

Therefore, for example, one screen is divided into a plurality of blocks, the degree of inclusion of high-frequency components in the divided block-unit image data, so-called activity, is calculated, and the normalization coefficient is set for each block based on this. That is, the number of bits per pixel is set (bit allocation).

However, for example, when an image is recorded by a digital electronic still camera equipped with such an image signal compression encoding device, a main subject (for example, a person) is displayed on the screen.
If there is a high activity area around (background), the bit allocation at the time of image data compression is set to be high in that area. Therefore, the bit distribution in the area of the main subject located in the central portion of the screen becomes relatively low, and there is a problem that the image quality is deteriorated due to the compression processing of the image data.

The present invention has been made in view of such circumstances, and an image signal capable of setting bit allocation according to an area on a screen when compressing and coding image data forming one screen An object is to provide a compression encoding device.

DISCLOSURE OF THE INVENTION According to the present invention, digital image data constituting one screen is divided into a plurality of blocks on the screen, each block is subjected to two-dimensional orthogonal transformation, and the orthogonally transformed image data is given. An image signal compression encoding device for normalizing based on the normalization coefficient and encoding the normalized image data is output from the block activity calculating means for calculating the activity of each block and the block activity calculating means. Total activity calculating means for calculating the total value of the block activity for one screen and distribution to each area including a specific area on the screen for increasing the bit allocation during the compression encoding processing of the image data for one screen To calculate the number of bits to be processed
Bit allocation calculation means, the number of bits calculated by the first bit allocation calculation means and distributed to each area, the activity for each block belonging to each area calculated by the block activity calculation means, and the total activity calculation means. Second bit allocation calculating means for calculating the number of bits allocated to each block belonging to each area based on the total value of the activity of each block unit, and the first bit allocation calculating means is a specific area. Is calculated by weighting so that the bit allocation is higher than in other areas.

Description of Embodiments Embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1A and 1B show the configuration of an embodiment of an image signal compression encoding apparatus according to the present invention.

This device has a blocking unit 12. The blocking unit 12 is composed of a frame buffer, and still image data for one frame captured by an electronic still camera is input through the input terminal 10 and stored. Blocking unit 12
The image data for one frame stored in is divided into a plurality of blocks, read in block units, and sent to the two-dimensional orthogonal transformation unit 14. The two-dimensional orthogonal transformation unit 14 performs two-dimensional orthogonal transformation on the image data input in block units. Well-known techniques such as discrete cosine transform and Hadamard transform are used as the two-dimensional orthogonal transform.

The image data in block units, which has been two-dimensionally orthogonally transformed by the two-dimensional orthogonal transformation unit 14, is arranged vertically and horizontally, low-order data is arranged in the upper left portion, and becomes higher-order data in the lower right direction. The output of the two-dimensional orthogonal transformation unit 14 is sent to the normalization unit 16.

The normalization unit 16 cuts the coefficient of the image data that has been two-dimensionally orthogonally transformed by the two-dimensional orthogonal transformation unit 14 based on the setting output of the normalization coefficient setting unit 22, that is, transform coefficients, and then performs normalization. To do. The coefficient truncation is to compare the orthogonally transformed transform coefficient with a predetermined threshold value and to cut off a portion below the threshold value. The normalization is the product of a predetermined quantization step value of the transform coefficient subjected to the coefficient truncation, that is, the normalization coefficient α obtained from the quantization table 54 and the weighting coefficient T set by the parameter setting unit 52. αT
Quantization is performed by dividing by. Normalization coefficient α
Is set by the normalization coefficient setting unit 26 based on a value obtained by summing up the activities in block units calculated by the total activity calculating unit 24, as will be described later.

The image data in block units output from the blocking unit 12 is also sent to the block activity calculation unit 20. The block activity calculation unit 20 calculates the activity on a block-by-block basis, that is, the degree to which high-frequency component image data is included in the block.

The calculation of the activity B-act (i, j) in block units is performed by dividing the divided blocks into, for example, 8 blocks as shown in FIG.
If it consists of × 8 pixels, the pixel data is
Then (where i, j = 0 ... 7), the expression Is calculated by here, Is.

That is, according to this equation, 8 × 8 that form a block
The average value DC (i, j) of the pixel data is calculated, the absolute value of the difference between each pixel data and the average value DC (i, j) is added, and the activity B-act (i, j) in block units is calculated. Ask for.

If you use the above formula to find the activity,
Since (i, j) is obtained by adding each pixel data and dividing the added data by 64 which is the number of pixels of the block, it can be configured only by an adder and a shifter for shifting data. Also, B-act (i, j) is the obtained DC
It is calculated by the absolute value conversion circuit and the adder using (i, j). Therefore, a multiplier and a divider are not required in the calculation of activity B-act (i, j).

As described above, the block activity calculation unit 20
Calculates the activity B-act (i, j) in block units and outputs it to the total activity calculation unit 24. The total activity calculation unit 24 sends the activity B-act (i, j) in block units sent from the block activity calculation unit 20.
Is added to calculate the total value T-act of the activities. The total activity calculation unit 24 outputs the total value T-act of activities to the normalization coefficient setting unit 26, the bit factor calculation unit 44, and the bit allocation calculation unit 46.

The normalization coefficient setting unit 26 refers to the quantization table 54 based on T-act, which is the total value of the activity B-act (i, j) in block units, which is calculated by the total activity calculation unit 24 for the normalization coefficient α. And set.

The normalization coefficient α is weighted by the weighting coefficient T set by the parameter setting unit 52 and output to the normalization unit 16.

The image data normalized by the normalization unit 16 is two-dimensionally Huffman coded by the two-dimensional Huffman coding unit 18 and output to the AC code amount counting unit 50 and the multiplexer 64.

The parameter setting unit 52 selects the image quality mode (compression ratio of image data) according to the application, outputs the weighting coefficient T necessary for setting the normalization coefficient α to the normalization coefficient setting unit 26, and also the image quality and the image. A parameter for specifying the number of recorded sheets is output to the bit factor calculation unit 44.

The quantization table 54 stores the activity of the entire screen, that is, the total value T-act of the activity in block units.
2 is a table showing the relationship between the normalization coefficient α and the normalization coefficient α.

On the other hand, the DC quantization unit 28 calculates a difference value between blocks for the low frequency (DC) component of the block-by-block image data output from the blocking unit 12. Huffman encoder 40
Performs Huffman coding on the difference value output from the DC quantizer 28. The DC code amount counting unit 42 counts the code amount output from the Huffman encoding unit 40 and outputs the total number to the bit factor calculation unit 44.

Further, the area designating section 70, as shown in FIG. 3 and FIG. 4, is a viewfinder of the viewfinder 32 when the object field 90 is photographed by the electronic still camera 30 incorporating the image signal compression coding apparatus according to the present invention. On the screen 34 displayed on the display unit 80, the area 38 (inside the dotted frame 36) where the bit allocation should be increased during the image data compression processing is designated. Area 38 in this case
Is designated by designating the coordinate data indicating the center position O of the frame 36 displayed on the finder display section 80 on the screen 34 with a keyboard or the like. For example, if the horizontal and vertical sides of the frame 36 are 2p and 2q, respectively, as shown in FIG. 5, the center position O of the frame 36 in the Cartesian coordinate system is the origin (0,0), and The coordinates of the vertices A, B, C, D are A (p,
q), B (-p, q), C (-p, -q), D (p, -q).

When the area designating unit 70 designates the coordinate data O (x, y) of the center position O of the frame 36, the position information is obtained by the area computing unit 7.
Entered in 2.

In the area calculation unit 72, the center position O (x, y) of the frame 36 is designated,
Coordinates A (p + x, q + y), B of vertices A, B, C, D after movement of the frame 36
(-Q + x, q + y), C (-p + x, -q + y), D (p + x,
-Q + y) and the number of blocks A-blk (n) contained in the area defined by these vertices A, B, C, D and the area other than this designated area. The calculated coordinate data of each apex is within the range included in the display screen 34 in a frame 36.
Is allowed to be moved, that is, the designation of the area 38 is permitted.

Note that n is a number given to each area when the screen 34 is divided into a plurality of areas.

The calculation output of the area calculation unit 72 is input to the area storage unit 74, the area bit allocation calculation unit 78, and the finder control unit 80.

The area storage unit 74 stores the position information for specifying the area 38, which is the calculation output of the area calculation unit 72.

Further, the finder control unit 80 controls to move the frame 36 in the finder display unit 82 based on the position information output from the area calculation unit 72.

The position information read from the area storage unit 74 is input to the image quality weighting count setting unit 76.

The image quality weighting coefficient setting unit 76 sets the image quality weighting coefficient W (n) for the area on the screen 34 designated by the area designating unit 70 and each area other than this area. Here, the total value of the image quality weighting factors W (n) is set to be 1, for example.

The image quality weighting coefficient setting unit 76 outputs the image quality weighting coefficient W (n) to the area bit allocation calculation unit 78.

The area bit allocation calculating unit 78 receives the number of blocks A-blk from the area calculating unit 72 and the image quality weighting coefficient setting unit 76, respectively.
(N) and the image quality weighting coefficient W (n) are taken in and the following equation (1)
The number of bits A-bit (n) distributed to each area is calculated by.

A-bit (n) = A-blk (n) .W (n) .T-bit / T-blk ...
(1) In the above equation (1), T-bti is the target bit number assigned to one screen, and T-blk is the number of blocks included in one screen.

The number of bits A-bit (n) obtained by the area bit allocation calculating unit 78 is input to the bit factor calculating unit 44.

In the bit factor calculation unit 44, the number of bits S (i,
A bit factor f (n) that is a coefficient for calculating j) is calculated.

f (n) = A-bit (n) / T-bit = W (n) .A-blk (n) / T-blk (2) Bit factor f (n) calculated by the bit factor calculator 44 Is input to the bit allocation calculation unit 46. The bit allocation calculation unit 46 includes a block activity calculation unit 20,
The activity B-act (i, j) of each block, which is the output from the total activity calculating unit 24, and the activity T-act of the entire screen are also input.

The bit allocation calculation unit 46 calculates the number of bits S (i, j) allocated to each block of one screen by the following equation.

S (i, j) = f (n) .T-bit.B-act (i, j) / T-act = A-bit (n) .B-act (i, j) / T-act ...
(3) On the other hand, the AC code amount counting unit 50 counts the AC code amount output from the two-dimensional Huffman encoding unit 18, and outputs it to the encoding stop determination unit 48.

The encoding stop determination unit 48 compares the AC code amount output from the AC code amount counting unit 50 with the bit number S (i, j) output from the bit allocation calculation unit 46, and the AC code amount is the bit number S.
When it exceeds (i, j), the determination signal for stopping the coding of the block is output to the two-dimensional Huffman coding unit 18.

In the above configuration, when the photographer shoots the field of view 90 using the electronic still camera 30, an area where the bit allocation should be increased during image data compression processing on the display screen 34 displayed on the finder display unit 80 of the finder 32. Coordinate data O (x, y) indicating the center position O of the frame 36 for designating 38 is designated by the area designating unit 70.

Position information indicating the coordinate data O (x, y) is output from the area designating unit 70 to the area calculating unit 72. In the area calculation unit 72, each of the frames 36 that specifies the position on the display screen 34 of the area 38 specified based on the position information of the center position O of the frame 36 on the display screen 34 of the finder display unit 80 that has been input. The coordinates of the vertices A, B, C, D and the number of blocks A-blk (n) included in the region defined by these vertices A, B, C, D and the region other than the designated region are obtained.

The calculation output of the area calculation unit 72 is input to the area storage unit 74, the area bit allocation calculation unit 78, and the finder control unit 80. The area storage unit 74 stores the position information for specifying the area 38 on the display screen 34 designated by the area designation unit 70.

Further, the finder control unit 80 controls the frame 36 to move to a designated position in the finder display unit 82 based on the position information output from the area calculation unit 72. As a result, it is possible to confirm the area designated by the photographer by the area designation section 70 on the display screen 34 of the finder display section 82 where the bit allocation is increased during the image data compression processing.

On the other hand, one frame of image data input from the input terminal 10 is input to the block unit 12 and stored therein. The image data for one frame stored in this block unit 12 is
It is divided into a plurality of blocks, read in block units, and sent to the two-dimensional orthogonal transformation unit 14. The two-dimensional orthogonal transformation unit 14 two-dimensionally orthogonally transforms the image data input from the block unit 12 in block units. The output signal of the two-dimensional orthogonal transformation unit 14 is input to the normalization unit 16.

The image data in block units output from the block unit 12 is sent to the block activity calculation unit 20, and the block activity calculation unit 20 calculates the activity B-act (i, j) in block units, and the total activity calculation unit Output to 24. The total activity calculating section 24 or the block activity calculating section 20 outputs the activity B-act (i, j) in block units to calculate the total value T-act of the activities, and the calculated value T-act. It outputs to the normalization coefficient setting unit 26, the bit factor calculation unit 44, and the bit allocation calculation unit 46.

The normalization coefficient setting unit 26 refers to the quantization table 54,
Based on the setting output of the parameter setting unit 52, the weighted normalization coefficient α · T is set in the normalization unit 16.

The image data normalized by the normalizer 16 is two-dimensionally Huffman-encoded by the two-dimensional Huffman encoder 18.

On the other hand, the screen 34 set by the image quality weighting factor setting unit 76
Image quality weighting coefficient W (n) of each area, and the number of blocks A- included in each area of the screen 34 calculated by the area calculation unit 72.
The area bit allocation calculating unit 78 calculates the number of bits A-bit (n) allocated to each area based on bnk (n), and outputs it to the bit factor calculating unit 44.

The bit factor calculation unit 44 calculates the bit factor f (n) based on the number of bits A-bit (n) input from the area bit distribution calculation unit 78, and the bit distribution calculation unit
Output to 46.

The bit allocation calculation unit 46 receives the block-by-block activity B-act from the block activity calculation unit 20, the total activity calculation unit 24, and the bit factor calculation unit 44.
(I, j) The activity T-act of the entire screen and the bit factor f (n) are taken in, the number of bits S (i, j) distributed to each block of the single screen is calculated, and the coding stop determination unit Output to 48.

On the other hand, the difference value of the DC component between the blocks of the image data in block units output from the blocking unit 12 is the DC quantization unit.
The Huffman coding unit 40 Huffman-encodes the difference value obtained by the step 28. The image data output by the Huffman encoder 40 and the two-dimensional Huffman encoder 18
The image data output by the multiplexer 64 is selectively output by the multiplexer 64.

The encoding stop determination unit 48 compares the AC code amount output from the AC code amount counting unit 50 with the number of bits S (i, j) output from the bit allocation calculation unit 46, and determines that the AC code amount is the number of bits S.
When it exceeds (i, j), the determination signal for stopping the coding of the block is output to the two-dimensional Huffman coding unit 18.

One screen is designated area 38 (area number n on the display screen of finder display 80 as shown in FIG. 6).
= 1) and its peripheral portion are divided into four areas (area numbers n = 2 to 5) excluding the designated area 38, and bits are individually allocated to blocks belonging to these areas.

Effect As described above, in the present invention, it is possible to specify in advance a region of the field image displayed in the viewfinder field in which the bit allocation is to be increased when the image data is compressed.
In addition, the bit distribution is set to be high for the image data belonging to the designated area during the compression process.

Therefore, according to the present invention, when the main subject in the center of the screen is dark and there is an image of high activity in its periphery, and the main subject is compressed when the image data is compressed with the same normalization coefficient, It is possible to improve the image quality even if the image is markedly deteriorated.

[Brief description of drawings]

FIG. 1A and FIG. 1B are block diagrams showing a configuration of an embodiment of an image signal compression coding apparatus according to the present invention, and FIG. 2 is an explanatory diagram showing a configuration example of image data forming a block of a screen. FIG. 3 is an explanatory view showing a state in which an object scene is photographed by an electronic still camera, and FIG. 4 is an area of the object scene image displayed in the finder display section where high bit allocation should be performed during compression processing. 6 is an explanatory view showing a state in which the main subject is designated, FIG. 5 is an explanatory diagram of coordinate data necessary for positioning a frame displayed when a region is designated in the finder display section, FIG. 6 FIG. 3 is an explanatory diagram showing a state in which a shooting screen is divided into a plurality of areas. Description of main part code 20 …… Block activity calculation unit 24 …… Total activity calculation unit 44 …… Bit factor calculation unit 46 …… Bit allocation calculation unit 70 …… Area designation unit 72 …… Area calculation unit 74 …… Area Storage unit 76 …… Image quality weighting factor setting unit 78 …… Area bit distribution calculation unit 80 …… Finder control unit 82 …… Finder display unit

Claims (1)

(57) [Claims] 1. Digital image data constituting one screen is divided into a plurality of blocks on the screen, two-dimensional orthogonal transformation is performed for each block, and the orthogonally transformed image data is given a normalized coefficient. In an image signal compression coding apparatus for normalizing based on the above, and for coding the normalized image data, the apparatus comprises a block activity calculating means for calculating an activity for each block, and the block activity calculating means. A total activity calculating means for calculating the total value of the block activity for one screen output from each, and a specific area on the screen for increasing the bit allocation during the compression coding process of the image data for one screen. First bit allocation calculating means for calculating the number of bits allocated to the area, and the first bit allocation calculation Based on a total value of the number of bits calculated by the means and distributed to each area, the activity of each block belonging to each area calculated by the block activity calculation means, and the activity of each block calculated by the total activity calculation means A second bit allocation calculating means for calculating the number of bits allocated to each block belonging to each area, wherein the first bit allocation calculating means has a bit in the specific area as compared with other areas. An image signal compression encoding apparatus, characterized in that weighting is performed so that the distribution becomes higher.