JP2018007214A - Image encoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a circuit scale of quantization and encoding processing after wavelet transformation.SOLUTION: An image encoding device comprises: a transformation unit for inputting image data subjected to encoding and performing wavelet transformation on the inputted image data to generate transformed coefficient data of multiple sub-bands; an encoding unit for quantizing and encoding the coefficient data of the sub-bands obtained by the transformation unit; and a buffer unit for temporarily storing a preset sub-band among the multiple sub-bands obtained by the transformation unit. The transformation unit then supplies one of generated sub-bands HL, LH and HH and coefficient data of a sub-band LL to the encoding unit without interposing the buffer unit and supplies the remaining two sub-bands including the sub-band LL to the buffer unit.SELECTED DRAWING: Figure 1

Description

  The present invention relates to image coding, and more particularly to a technique for performing frequency conversion on an image and coding a generated coefficient.

  In recent years, with the development of digital imaging devices such as digital cameras and digital camcorders, various image data compression encoding methods have been studied. JPEG (Joint Photographic Experts Group) 2000 has been proposed as one of the compression coding methods. One of the features of JPEG 2000 is that wavelet transform (hereinafter referred to as DWT transform) is used instead of discrete cosine transform (DCT) used in JPEG. By using DWT conversion, even an image having a large size can be subjected to frequency conversion without being divided into a plurality of blocks. As a result, JPEG2000 can eliminate image quality degradation that may occur at block boundaries when an image is divided into a plurality of blocks.

  In order to perform DWT conversion, it is necessary to filter all the images in the vertical direction using a DWT filter and then filter in the horizontal direction. Therefore, as in Patent Document 1, a large-scale memory (buffer) is required when performing filtering.

  On the other hand, Patent Document 2 proposes to reduce the memory capacity. However, according to Patent Document 2, there arises a problem that timings at which a plurality of frequency regions (subbands) obtained by performing DWT conversion are output are almost the same. That is, the circuit scale of the subsequent quantization unit and encoding unit is increased.

JP 2003-274185 A JP 2001-285643 A

  The present invention has been made in view of the above problems, and intends to provide a technique for leveling quantization and encoding processing after discrete wavelet transform and reducing the circuit scale.

In order to solve this problem, for example, an image processing apparatus of the present invention has the following configuration. That is,
An image encoding device for encoding image data,
Conversion means for inputting image data to be encoded, wavelet transforming the input image data, and generating conversion coefficient data of a plurality of subbands;
Encoding means for quantizing and encoding the coefficient data of each subband obtained by the conversion means;
Buffer means for temporarily holding the conversion coefficient data obtained by the conversion means,
Among the plurality of subbands generated by the conversion means, the conversion coefficient data of a preset subband is supplied to the encoding means without going through the buffer means,
The transform coefficient data of other subbands among the plurality of subbands is supplied to the buffer means.

  According to the present invention, it is possible to reduce the circuit scale of quantization and encoding processing after wavelet transform.

1 is a block diagram illustrating a configuration example of an image encoding device according to a first embodiment. The figure for demonstrating the timing which implements reversible 5-3DWT conversion. The block diagram which shows the structure of the image coding apparatus for demonstrating a subject. The figure which shows the process timing of the image coding apparatus of FIG. The figure which shows the process timing of the image coding apparatus which concerns on 1st Embodiment. The figure which shows the relationship of each subband which generate | occur | produces by wavelet transformation. The figure which shows the process timing at the time of performing wavelet transformation to the decomposition | disassembly level 3 in the image coding apparatus which concerns on 1st Embodiment. The block diagram which shows the structural example of the image coding apparatus which concerns on 2nd Embodiment. The figure which shows the process timing of the image coding apparatus which concerns on 2nd Embodiment. The figure for demonstrating the relationship of the number of horizontal DWT coefficients of each subband which generate | occur | produces by wavelet transformation when the number of pixels of a horizontal direction is an odd number.

  First, prior to the description of the embodiment, wavelet coefficients (hereinafter referred to as DWT coefficients) in a DWT conversion using a reversible 5-3 tap filter proposed in JPEG 2000 (hereinafter referred to as reversible 5-3 DWT conversion) are output. The timing to be described will be described in detail with reference to FIGS. FIG. 2 is a diagram for explaining the timing at which the DWT coefficient is output when the reversible 5-3DWT conversion is realized using the lifting structure.

In FIG. 2A, symbols a to e indicate input pixel data arranged in the horizontal direction. In FIG. 2 (a), b 'is a DWT coefficient of a high-frequency component of decomposition level 1 generated using pixel data a, b, c, and d' is a decomposition level 1 generated using pixel data c, d, e. It is a DWT coefficient of the high frequency component. Further, c ″ is a DWT coefficient of a low frequency component generated using b ′, d ′, and c. Each DWT coefficient is as shown in the following equations (1) to (3).
b '= b- (a + c) / 2 (1)
d '= d- (c + e) / 2 (2)
c "= c + (b '+ d' + 2) / 4 ... (3)
As shown in the equations (1) and (2), b ′ and d ′ are DWT coefficients of high frequency components, and therefore the types of the equations are the same except for the pixel data used.

  FIG. 2B is a diagram illustrating a state in which pixel data f is newly input in the horizontal direction with respect to FIG. As shown in FIG. 2 (b), there are not enough input pixels to perform a new DWT conversion even when pixel data f is input, so that a new DWT coefficient is output at the timing when the pixel data f is input. I can't.

  On the other hand, when the pixel data g is further input to FIG. 2B as shown in FIG. 2C, DWT conversion can be executed, and the DWT coefficient f ′ of the high frequency component and the DWT coefficient of the low frequency component are executed. e "can be output.

As described above, the timing for executing the DWT conversion in the horizontal direction is once for every two pixel inputs of pixel data. Note that when performing DWT transformation hierarchically to decomposition level 2, it is necessary to use the DWT coefficient of the low frequency component of decomposition level 1. As described above, since the timing at which the DWT coefficient of the low frequency component at the decomposition level 1 is output is once for every two pixel inputs of each pixel data, the timing at which the DWT coefficient at the resolution level 2 is output is once for every four pixel inputs. It becomes. From the above relation, the timing at which the DWT coefficient at the decomposition level N is output is once for each X pixel using X shown in the following equation (4).
X = 2 ^N (4)

  Since the same relationship applies to the vertical direction, the timing at which the DWT coefficient in the vertical direction is output is once per pixel data X-line input according to equation (4).

  FIG. 6 shows the relationship between the subbands when the input image is subjected to vertical / horizontal DWT conversion up to decomposition level 2. In FIG. 6, L indicates a low frequency region, and H indicates a high frequency region. For example, 1HH indicates a subband in the high frequency region in both the horizontal direction and the vertical direction at the decomposition level 1. As shown in FIG. 6, the size of each subband at the decomposition level 1 is half of the input image data in the horizontal and vertical directions. That is, the number of coefficients of each subband is 1/4 of the number of pixels of the original image. Further, each subband at resolution level 2 is half of the horizontal and vertical directions of each subband at resolution level 1.

  Based on the above description, a series of processing timings for DWT conversion and encoding of image data will be described with reference to FIGS.

  FIG. 3 is a block diagram illustrating a configuration example of an image encoding device considered by the present inventors. As shown in FIG. 3, the image coding apparatus includes a wavelet transform unit 300 and a quantization / coding unit 303. The wavelet transform unit 30 includes a Lev1-DWT conversion unit 301 that performs decomposition level 1 DWT conversion and a Lev2-DWT conversion unit 302 that performs decomposition level 2 DWT conversion. Although an example of the encoding method will be described using JPEG2000, the encoding method is not particularly limited.

  4 shows the conversion timing of the DWT coefficients of the subbands of decomposition levels 1 and 2 and the converted DWT when the image data is input to the image encoding apparatus of FIG. 3 in raster order and DWT conversion is performed up to decomposition level 2. It is a figure which shows the timing when a coefficient is quantized and encoded.

  The way of viewing FIG. 4 will be described. In FIG. 4, each interval from t0 to t14 on the horizontal axis represents a period during which one line of pixel data is input. Note that the input of image data is started from the period t0.

  A rectangular bar shown in “Pixel data” indicates that pixel data is input during that period. In FIG. 4, since there are bars in all periods, pixel data for one line is input in all periods.

  On the other hand, the subband bar of each decomposition level of “DWT transform” indicates whether or not the DWT transform was executed in each period and the data of each subband was output to the quantization / encoding unit 303. For example, in the period t0, since there is no bar in any sub-band of the decomposition level, it indicates that DWT conversion is not executed. In addition, since the Lev1-LH, Lev1-HL, and Lev1-HH bars exist in the period t2, the decomposition level 1DWT is executed and the coefficients of the level 1 subbands LH, HH, and LH are output. Show.

  The “quantization / encoding” bar indicates at which timing the DWT coefficient output by DWT conversion in each period is quantized / encoded. For example, in the period t0, since there is no bar in any decomposition level subband, it indicates that the coefficients of all subbands are not quantized and encoded. Further, since the Lev1-LH, Lev1-HL, and Lev1-HH bars exist in the period t2, it indicates that the subbands HL, LH, and HH at the decomposition level 1 are quantized and encoded.

  As shown in the figure, in this image encoding apparatus, the DWT conversion bar and the quantization / encoding bar exist in the same period. This indicates that the DWT coefficient output by the DWT transform is immediately quantized and encoded in the same line input period. The length of the bar indicates the period during which quantization / encoding is executed.

  As shown in FIG. 6, only 0.5 lines of DWT coefficients (Lev1-LH, HL, HH) for each subband of decomposition level 1 are output during a period in which one line of pixel data is input. However, as described with reference to FIG. 2, there is only one coefficient input to the quantization / encoding unit, so that there is only one coefficient per two input pixels. Therefore, as shown in FIG. Encoding.

  Next, the encoding process executed in each section of FIG. 4 will be described with reference to FIG. In the period from t0 to t1, pixel data for two lines is input to the Lev1-DWT conversion unit 301. However, at this point, data for the number of lines sufficient to execute the reversible 5-3DWT conversion has not yet been input. For this reason, the Lev1-DWT conversion unit 301 only holds the two-line input pixel data in the line buffer in the Lev1-DWT conversion unit 301, and does not execute the DWT conversion in this period.

  In the t2 period, the Lev1-DWT conversion unit 301 performs DWT conversion using the pixel data for two lines held in the line buffer in the period t0 and t1 and the input pixel data for the third line. As a result, coefficient data of subbands LL, LH, HH, and LH of decomposition level 1 are generated. Pixel data that is insufficient to execute the 5-tap filter process is generated by mirroring. Among the generated DWT coefficients, the DWT coefficients of the decomposition level 1 subband LL are transmitted to the Lev2-DWT conversion unit 302 because they are used for the decomposition level 2 DWT conversion. The coefficient data of the other subbands, that is, the subbands LH, HH, and LH of the decomposition level 1 are output to the subsequent quantization / encoding unit 303.

  Note that only the DWT coefficients of the subband LL at the decomposition level 1 are output for one line in the period t2, and the DWT conversion at the decomposition level 2 cannot be executed. For this reason, the Lev2-DWT conversion unit 302 does not execute the DWT conversion, but only holds the coefficient data in the internal line buffer. In addition, the number of coefficients in each subband of decomposition level 1 is half of the original pixel data. Accordingly, by combining the three subbands LH, HH, and LH at the decomposition level 1, DWT coefficients for 1.5 lines (0.5 × 3) are output based on the original pixel data reference. The quantization / encoding unit 303 quantizes the DWT coefficients for a total of 1.5 lines sent from the wavelet transform unit 300 using the quantization parameter. In the encoding, entropy encoding such as EBCOT (Embedded Block Coding with Optimized Truncation) is performed for each subband and output as encoded data.

  In the period t3, as in the period t0 to t1, data corresponding to the number of lines sufficient to perform the reversible 5-3DWT conversion is not input. Therefore, the input pixel data is held in the line buffer in the Lev1-DWT conversion unit 301, and no DWT coefficient data is output during this period.

  In the t4 period, DWT conversion is performed using the pixel data for two lines input in the period t2 and t3 and held in the line buffer, and the pixel data for the third line, and the subbands LL, LH, and HH at the decomposition level 1 , LH coefficient data is calculated. However, since the DWT coefficient data of the subband LL at the decomposition level 1 is output only for two lines including the amount held in the line buffer in the period t2, the decomposition level 2 DWT conversion cannot be executed. Thereafter, the decomposition level 1 DWT conversion is executed once every two lines (that is, in a period of t6, t8, t10, t12, t14,. The quantization / encoding unit 303 quantizes the DWT coefficients for a total of 1.5 lines sent from the wavelet transform unit 300 using quantization parameters, performs entropy encoding such as EBCOT for each subband, Output as digitized data.

  In the period t5, the DWT conversion of the decomposition levels 1 and 2 is not executed, and the pixel data is held in the line buffer in the Lev1-DWT conversion unit 301 as input data.

  In the period t6, the Lev1-DWT converter 301 performs decomposition level 1 DWT conversion. Also, the DWT coefficient data of the decomposition level 1 subband LL output at that time, and two lines of the decomposition level 1 subband LL held in the line buffer in the Lev2-DWT conversion unit 302 in the period t2 and t4 The Lev2-DWT conversion unit 302 calculates the decomposition level level 2 subband LL, LH, HH, and LH coefficient data using the minute DWT coefficient data.

  As shown in FIG. 4, since the number of decomposition level 2 coefficients is half of the decomposition level 1 coefficients, the subbands LL, LH, HH, and LH of decomposition level 2 are sub-bands based on the original pixel data standard In total, the DWT coefficients for one line (0.25 × 4) are output to the quantization / encoding unit 303 at the subsequent stage. Combine this with 1.5 lines for the 3 subbands of resolution level 1 subbands LH, HH, and LH, and a DWT coefficient of 2.5 lines (1 + 1.5) based on the original pixel data reference is output during this period. The

  The decomposition level 2 DWT conversion is executed once every four lines (that is, t10, t14,...) As shown in Equation 4 above. The quantization / encoding unit 303 quantizes the DWT coefficients for a total of 2.5 lines sent from the wavelet transform unit 300 using quantization parameters, and performs entropy encoding such as EBCOT for each subband, Output as encoded data. Thereafter, the same processing is repeated until the last line of the image data.

  As can be seen from the above description, when DWT conversion is performed up to decomposition level 2, DWT coefficients for a maximum of 2.5 lines are output during a period in which one line of input pixels is input (t6, t10 in FIG. 4). , t14). Therefore, when performing quantization and encoding immediately at the timing of executing the DWT conversion, it is necessary to have a performance capable of processing DWT coefficients for 2.5 lines in a period of inputting one line of pixel data. It can be understood that the circuit scale of the encoding unit becomes large. Hereinafter, based on such points, embodiments according to the present invention will be described.

[First Embodiment]
FIG. 1 is a block diagram illustrating a configuration example of an image encoding device according to an embodiment of the present invention in the first embodiment. In order to simplify the description, the present embodiment will be described on the assumption that the discrete wavelet transform is performed twice.

  The image coding apparatus according to the first embodiment includes a wavelet transform unit 300, a subband line buffer 103, and a quantization / coding unit 104, as illustrated. It is assumed that pixel data of image data to be encoded is supplied to the wavelet transform unit 300 in the raster scan order. The wavelet transform unit 300 includes a Lev1-DWT transform unit 301 that performs a decomposition level 1 discrete wavelet transform (hereinafter also referred to as a DWT transform in the embodiment), and a Lev2-DWT transform unit 302 that performs a decomposition level 2 DWT transform. The In the present embodiment, JPEG2000 is described as an example of an encoding method, but the type of encoding method is not particularly limited. The wavelet transform unit 300, the Lev1-DWT transform unit 301, and the Lev2-DWT transform unit 302 use the same configuration as that of the encoding apparatus of FIG. 2 described above, but output coefficient data obtained by DWT transform. The destination is different. Further, it is assumed that the quantization / encoding unit 104 has a performance capable of quantizing / encoding a DWT coefficient for one line in a period in which one line of pixel data is input.

  FIG. 5 shows the timing at which the image data is converted into DWT coefficient data for each subband of level 1 and level 2 when the image data is DWT converted to resolution level 2 in raster order in FIG. 1, and the converted DWT coefficient data is quantized. -It is a timing diagram for demonstrating the timing encoded. The way of viewing FIG. 5 is the same as that of FIG.

  Hereinafter, the encoding process performed in each section of FIG. 5 will be described with reference to FIG.

  Pixel data for two lines is input to the Lev1-DWT conversion unit 301 in the period from t0 to t1. However, at this time, data for the number of lines sufficient to perform the reversible 5-3DWT conversion has not been input yet. For this reason, the Lev1-DWT conversion unit 301 only holds two lines of input pixel data in the line buffer in the Lev1-DWT conversion unit 301, and does not perform DWT conversion in this period.

  In the t2 period, the Lev1-DWT conversion unit 301 performs DWT conversion using the pixel data for two lines held in the line buffer in the period t0 and t1 and the input pixel data for the third line, and generates a decomposition level 1 sub-pixel. Coefficient data for bands LL, LH, HH, and LH are output. At this time, the DWT coefficient data of the subband LL of decomposition level 1 is supplied to the Lev2-DWT conversion unit 302 because it is used for level 2 DWT conversion. Further, the DWT coefficient data of the subbands LH and HL of the decomposition level 1 are directly output to the quantization / encoding unit 104 without the interposition of the subband line buffer 103, but the DWT of the subband HH of the level 1 The coefficient data is transmitted to the subband line buffer 103 and temporarily held.

  Note that only one line of the DWT coefficient data of the subband LL at the decomposition level 1 is still output during the period t2, and the DWT conversion at the decomposition level 2 cannot be executed. For this reason, the Lev2-DWT conversion unit 302 does not perform DWT conversion, but only holds the coefficient data in the internal line buffer. The quantization / encoding unit 104 quantizes the DWT coefficient data for a total of 1.0 lines of the decomposition level 1 subbands LH and HL sent from the wavelet transform unit 300, alternately using quantization parameters for each cycle. To do. In encoding, for each subband, entropy encoding such as EBCOT (Embedded Block Coding with Optimized Truncation) is performed and output as encoded data.

  In the period t3, as in the period t0 to t1, data corresponding to the number of lines sufficient to perform the reversible 5-3DWT conversion is not input. Therefore, the input pixel data is held in the line buffer in the Lev1-DWT conversion unit 301, and no DWT coefficient data is output during this period. However, the quantization / encoding unit 104 reads the DWT coefficient data of the decomposition level 1 subband HH held in the subband line buffer 103 in the period t2, quantizes it using the quantization parameter, and performs entropy encoding. And output as encoded data. Since only DWT coefficient data of subband HH of decomposition level 1 is encoded during this period, only DWT coefficient data for 0.5 lines is quantized and encoded on the basis of pixel data. Therefore, quantization / encoding can be executed in a half of the period during which one line of pixel data is input.

  In the t4 period, the Lev1-DWT conversion unit 301 performs DWT conversion using the pixel data input in the t2 and t3 periods and held in the line buffer, and the pixel data input this time, and the subband LL of decomposition level 1 , LH, HH, LH coefficient data is generated. At this time, since the DWT coefficient data of the subband LL at the decomposition level 1 is used for the DWT conversion at the decomposition level 2, it is transmitted to the Lev2-DWT conversion unit 302. However, even at this time, the DWT coefficients of the level 1 subband LL are output for only two lines including the amount held in the line buffer in the period t2, so that the level 2 DWT conversion cannot be executed. Similarly to the period t2, the DWT coefficient data of subbands LH and HL of decomposition level 1 are output to the quantization / encoding unit 104, while the DWT coefficient data of subband HH of decomposition level 1 is output to the subband line buffer 103. And temporarily held (delayed by one cycle). The quantization / encoding unit 104 quantizes the DWT coefficient data for 1.0 lines of the decomposition level 1 subbands LH and HL sent from the wavelet transform unit 300 alternately using quantization parameters every cycle. Entropy encoding is performed and the encoded data is output.

  In the period t5, the decomposition level 1 and 2 DWT conversion is not executed, and the input data is held in the line buffer in the Lev1-DWT conversion unit 301 as pixel data. However, as in the t3 period, the quantization / encoding unit 104 reads the DWT coefficient data (stored in the t4 period) of the decomposition level 1 subband HH held in the subband line buffer 103, and stores the quantization parameter. Is quantized using, entropy-coded, and output as encoded data.

  In the period t6, the Lev1-DWT converter 301 performs decomposition level 1 DWT conversion. Further, at this time, the DWT coefficient data of the decomposition level 1 subband LL and the DWT coefficient of the level 1 subband LL held in the line buffer in the Lev2-DWT conversion unit 302 in the periods t2 and t4 are used. Then, the Lev2-DWT conversion unit 302 executes the DWT conversion to calculate and output the coefficient data of the subbands LL, LH, HH, and LH at the decomposition level 2.

  As shown in FIG. 4, since the number of coefficient data at decomposition level 2 is half of the coefficient data at decomposition level 1, 4 sub-bands LL, LH, HH, and LH at decomposition level 2 on the basis of the original pixel data. Together, the DWT coefficient data for one line (0.25 × 4) is transmitted to the subband line buffer 103 and temporarily held. In addition, if 1.5 lines including three subbands LH, HH, and LH output by the DWT conversion at decomposition level 1 are included, the DWT coefficient data for 2.5 lines (1 + 1.5) is based on the original pixel data standard. Output during this period.

  The quantization / encoding unit 104 quantizes DWT coefficient data for a total of 1.0 lines of decomposition level 1 subbands LH and HL sent from the wavelet transform unit 300 alternately using quantization parameters for each cycle. Entropy encoding is performed and the encoded data is output.

  In the t7 period, the DWT conversion at the decomposition levels 1 and 2 is not executed, and the pixel data is held in the line buffer in the Lev1-DWT conversion unit 301 as input data. However, here, the quantization / encoding unit 104 holds the DWT coefficient data of the decomposition level 1 subband HH held in the subband line buffer 103 in the period t6 and the subband line buffer 103 in the period t6. Further, the DWT coefficient data of the subbands LH and HL at the decomposition level 2 are quantized and encoded. The quantization / encoding unit 104 first quantizes and encodes the decomposition level 1 subband HH DWT coefficient data corresponding to 0.5 lines on a pixel data basis. Subsequently, the quantization / encoding unit 104 quantizes and encodes the DWT coefficient data of the decomposition level 2 subband LH corresponding to 0.5 lines on the pixel data basis. Finally, the quantization / encoding unit 104 quantizes and encodes the DWT coefficients of the subband HL of the decomposition level 2 corresponding to 0.5 lines on the pixel data basis. As a result, the DWT coefficients for 1.0 line are quantized and encoded using the entire period during which one line of pixel data is input. In addition, if the decomposition level 1 subband HH is quantized / encoded during this period, the processing order of the DWT coefficients to be quantized / encoded is not particularly limited. For example, instead of processing in order of subband HH at decomposition level 1, subband LH at decomposition level 2, subband HL at decomposition level 1, subband LH at decomposition level 2, subband HL at decomposition level 2, and decomposition level 1 The processing order may be changed, for example, processing is performed in the order of the subbands HH. Further, instead of the decomposition level 2 subband LH and the decomposition level 2 subband HL, the decomposition level 2 subband HH and the decomposition level 2 subband LL may be quantized and encoded during this period. In this case, the decomposition level 2 subband LH and the decomposition level 2 subband HL that are not quantized / encoded in this period may be quantized / encoded in the period t9 described later.

  In the period t8, the DWT coefficients of the subband LL of decomposition level 1 necessary for executing the next decomposition level 2 DWT conversion are insufficient. Therefore, only decomposition level 1 1DWT transformation is performed. Similarly to the periods t2, t4, and t6, the DWT coefficient data of the decomposition level 1 subbands LH and HL are output to the quantization / encoding unit 104, while the DWT coefficient data of the decomposition level 1 subband HH is the sub It is transmitted to the band line buffer 103 and once held. The quantization / encoding unit 104 quantizes the DWT coefficient data for 1.0 lines of the decomposition level 1 subbands LH and HL sent from the wavelet transform unit 300 alternately using quantization parameters every cycle. Entropy encoding is performed and the encoded data is output.

  In the period t9, the DWT conversion at the decomposition levels 1 and 2 is not executed, and the input data is held in the line buffer in the Lev1-DWT conversion unit 301 as pixel data. However, in this t9 period, the quantization / encoding unit 104 performs the decomposition level 1 subband HH DWT coefficient data held in the subband line buffer 103 in the t8 period, and the subband line buffer in the t6 period. The decomposition level 2 subbands HH and LL DWT coefficient data held in 103 are quantized and encoded. Specifically, the quantization / encoding unit 104 first quantizes and encodes the DWT coefficient data of the decomposition level 1 subband HH corresponding to 0.5 lines on a pixel data basis. Subsequently, the quantization / encoding unit 104 quantizes and encodes the decomposition level 2 subband HH DWT coefficient data corresponding to 0.5 lines on a pixel data basis. Finally, the quantization / encoding unit 104 quantizes and encodes the DWT coefficient data of the sub-band LL of decomposition level 2 corresponding to 0.5 lines on the pixel data basis. As a result, the DWT coefficients for 1.0 line are quantized and encoded using the entire period during which one line of pixel data is input.

  Thereafter, the above process is repeated until the last line of the image data.

  By performing quantization / encoding with the configuration and processing timing as described above, the quantization / encoding unit 104 has the same number during the period of inputting pixel data for one line from the image data to be encoded. It is only necessary to have the ability to process the DWT coefficient data. That is, the total number of transform coefficient data input by the quantization / encoding unit 104 can be set to the number of pixels in the horizontal direction of the image data to be encoded at the maximum. In the case of the configuration shown in FIG. 3, the quantization / encoding unit needs the ability to process DWT coefficient data for 2.5 lines in a period for inputting pixel data for one line of image data to be encoded. It can be understood that the increase in the circuit scale can be suppressed as compared with the conventional case.

  In this embodiment, a reversible 5-3 tap filter is used. However, any filter can be used as long as the filter satisfies the relationship of Equation 4 such as a nonreciprocal 9-7 tap filter proposed in JPEG2000. It is possible to execute at the processing timing.

  In the present embodiment, the case where the wavelet coefficients are executed up to the decomposition level 2 has been described as an example. However, when the wavelet coefficients are executed up to a higher level after the decomposition level 3, the same idea can be executed. It is.

  For example, in the case of decomposition level 3, a buffer for temporarily storing decomposition level 3 subbands is newly provided in the subband line buffer 103, and the stored DWT coefficients are set to low level subbands as shown in the timing diagram of FIG. This can be realized by preferentially reading from.

  Here, for example, when generalizing the case of decomposition into subbands up to decomposition level N and encoding, the i-th (i ≧ 2) wavelet transform unit is the decomposition level generated by the i−1-th wavelet transform unit. The conversion coefficient data of the subband LL of i-1 is input, and the conversion coefficient data of the subbands LL, HL, LH, and HH of the decomposition level i is output. Then, the first wavelet transform unit positioned first supplies two of the generated decomposition level 1 subbands HL, LH, and HH to the quantization / encoding unit without intervention of the buffer memory. At the same time, the remaining one is supplied to the buffer memory. The other wavelet transform units other than the first wavelet transform unit supply the generated subbands HL, LH, and HH of the respective decomposition levels to the buffer memory. Then, the quantization / encoding unit gives the same number of transform coefficients as the number of pixels in the horizontal direction of the image data to be encoded while giving priority to the lower decomposition level from the first wavelet transform unit or the buffer memory. After data is input, quantization and encoding may be performed.

[Second Embodiment]
Subsequently, a second embodiment of the present invention will be described. FIG. 8 is a block diagram of an image encoding apparatus according to the second embodiment, and FIG. 9 is a timing chart relating to the encoding process. The apparatus configuration of the second embodiment is almost the same as that of the first embodiment, but the object to be held in the subband line buffer 103 is the DWT transform coefficient data of the subband LH of decomposition level 1 Is different. Hereinafter, the processing of the image coding apparatus according to the second embodiment shown in FIG. 8 will be described with reference to the timing chart of FIG.

  In the period from t0 to t1, pixel data for two lines is supplied to the Lev1-DWT conversion unit 301 from the image data to be encoded. However. At this time, data for the number of lines sufficient to perform reversible 5-3DWT conversion has not been input yet. For this reason, the Lev1-DWT conversion unit 301 only holds the input pixel data of two lines in the line buffer in the Lev1-DWT conversion unit 301, and does not execute the DWT conversion in this period.

  In the t2 period, the Lev1-DWT conversion unit 301 executes DWT conversion using the pixel data for two lines of the encoding target image held in the line buffer in the t0 and t1 periods and the pixel data of the third line, and decomposes them. Coefficient data of subbands LL, LH, HH, and LH of level 1 are output. At this time, the DWT coefficient data of the decomposition level 1 subband LL is supplied to the Lev2-DWT conversion unit 302 because it is used to generate decomposition level 2 DWT conversion data. Further, the DWT coefficient data of subbands HH and HL of decomposition level 1 are output to the quantization / encoding unit 104, but the DWT coefficient data of subband LH of decomposition level 1 is supplied to the subband line buffer 103 and temporarily. , Retained (delayed).

  In the period t2, only one line of the DWT coefficient data of the subband LL at the decomposition level 1 is output, and the DWT conversion at the decomposition level 2 cannot be executed. For this reason, the Lev2-DWT conversion unit 302 does not execute DWT conversion for the subband LL of decomposition level 1 and keeps the coefficients in the internal line buffer. The quantization / encoding unit 104 quantizes the DWT coefficient data for 1.0 line total of HH and HL subbands sent from the wavelet transform unit 300 for each cycle alternately using quantization parameters. Is done. In encoding, for each subband, entropy encoding such as EBCOT (Embedded Block Coding with Optimized Truncation) is performed and output as encoded data.

  In the t3 period, as in the period from t0 to t1, data corresponding to the number of lines sufficient to perform the reversible 5-3DWT conversion is not input. Therefore, the input pixel data is held in the line buffer in the Lev1-DWT conversion unit 301, and no DWT coefficient is output during this period. However, the quantization / encoding unit 104 reads the DWT coefficient data of the decomposition level 1 subband LH held in the subband line buffer 103 in the period t2, quantizes it using the quantization parameter, and performs entropy encoding. And output as encoded data. During this period, only the DWT coefficient data of the subband LH at the decomposition level 1 is encoded, so that only DWT coefficient data for 0.5 lines is quantized and encoded on the pixel data basis. Therefore, quantization / encoding can be executed in a half of the period during which one line of pixel data is input.

  In the t4 period, the Lev1-DWT conversion unit 301 performs DWT conversion using the pixel data input in the t2 and t3 periods and held in the line buffer, and the newly input pixel data, and the decomposition level 1 sub Coefficient data of the bands LL, LH, HH, and LH are generated. At this time, the DWT coefficient data of the decomposition level 1 subband LL is supplied to the Lev2-DWT conversion unit 302 because it is used to generate decomposition level 2 DWT conversion data. However, at this time, even if the DWT coefficient data of the subband LL at the decomposition level 1 is combined with the amount held in the line buffer in the period t2, it does not correspond to two lines. Therefore, the Lev2-DWT conversion unit 302 does not execute DWT conversion. Similarly to the period t2, the DWT coefficient data of the decomposition level 1 subbands HH and HL are output to the quantization / encoding unit 104, but the decomposition level 1 subband LH DWT coefficient data is output from the subband line buffer. Sent to 103 and once held. The quantization / encoding unit 104 quantizes the DWT coefficient data for 1.0 line of the subbands HH and HL of the decomposition level 1 sent from the wavelet transform unit 300 alternately by using the quantization parameter every cycle. Entropy encoding is performed and the encoded data is output.

  In the period t5, the decomposition level 1 and 2 DWT conversion is not executed, and the input data is stored in the line buffer in the Lev1-DWT conversion unit 301 as pixel data. However, as in the t3 period, the quantization / encoding unit 104 reads the decomposition level 1 subband LH DWT coefficient data held in the subband line buffer 103 in the t4 period, and performs quantization using the quantization parameter. And entropy-coded, and output as encoded data.

  In a period t6, the Lev1-DWT conversion unit 301 performs DWT conversion, generates coefficient data of decomposition level 1 subbands LL, LH, HH, and LH and outputs them. At this time, the DWT coefficient data of the decomposition level 1 subband LL and the DWT of the decomposition level 1 subband LL held in the line buffer in the Lev2-DWT conversion unit 302 in the period t2 and t4. Using the coefficient data, the Lev2-DWT conversion unit 302 performs DWT conversion. As a result, coefficient data of subbands LL, LH, HH, and LH of decomposition level 2 are generated.

  As shown in FIG. 4, since the number of coefficient data at the decomposition level 2 is half of the coefficient data at the decomposition level 1, 4 sub-bands LL, LH, HH, and LH at the decomposition level 2 based on the original pixel data standard. Together, the DWT coefficient data for one line (0.25 × 4) is transmitted to the subband line buffer 103 and temporarily held. Since it is the 1.5-line coefficient data of the decomposition level 1 subbands LH, HH, and LH, the number of coefficient data generated in this t6 period is 2.5 lines (1 + 1.5) based on the original pixel data standard. Become.

  The quantization / encoding unit 104 quantizes the DWT coefficient data for a total of 1.0 lines of the decomposition level 1 subbands HL and HH sent from the wavelet transform unit 300 alternately using a quantization parameter every cycle. Entropy encoding is performed and the encoded data is output.

  In the period t7, the decomposition level 1 and 2 DWT conversion is not executed, and the pixel data is held in the line buffer in the Lev1-DWT conversion unit 301 for the input data. However, in this t7 period, the quantization / encoding unit 104 performs the decomposition level 1 subband LH DWT coefficient data held in the subband line buffer 103 in the t6 period, and the subband line buffer in the t6 period. The decomposition level 2 subbands LH and HL DWT coefficient data held in 103 are quantized and encoded. The quantization / encoding unit 104 first quantizes and encodes LH DWT coefficient data of decomposition level 1 corresponding to 0.5 lines on a pixel data basis. Subsequently, the quantization / encoding unit 104 quantizes and encodes the DWT coefficient data of the decomposition level 2 subband LH corresponding to 0.25 lines on the pixel data basis. Finally, the quantization / encoding unit 104 quantizes and encodes the decomposition level 2 subband HL DWT coefficient data corresponding to 0.25 lines on a pixel data basis. As a result, 1.0 line of DWT coefficient data is quantized and encoded using the entire period during which one line of pixel data is input.

  In the period t8, the DWT coefficient data of the subband LL of decomposition level 1 necessary for executing the next decomposition level 2 DWT conversion is insufficient. Thus, only decomposition level 1 DWT conversion is performed. Similarly to the periods t2, t4, and t6, the DWT coefficient data of the decomposition level 1 subbands HH and HL are output to the quantization / encoding unit 104, but the decomposition level 1 subband LH DWT coefficient data is It is transmitted to the subband line buffer 103 and once held. The quantization / encoding unit 104 quantizes the DWT coefficient data for 1.0 line of the subbands HH and HL of the decomposition level 1 sent from the wavelet transform unit 300 alternately by using the quantization parameter every cycle. Entropy encoding is performed and the encoded data is output.

  In the period t9, the decomposition level 1 and 2 DWT conversion is not executed, and the input data is held in the line buffer in the Lev1-DWT conversion unit 301 as pixel data. However, in this t9 period, the quantization / encoding unit 104 performs the decomposition level 1 subband LH DWT coefficient data held in the subband line buffer 103 in the t8 period, and the subband line buffer in the t6 period. The decomposition level 2 subbands HH and LL DWT coefficient data held in 103 are quantized and encoded. Specifically, the quantization / encoding unit 104 first quantizes and encodes the DWT coefficient data of the decomposition level 1 subband LH corresponding to 0.5 lines on a pixel data basis. Subsequently, the quantization / encoding unit 104 quantizes and encodes the DWT coefficient data of the decomposition level 2 subband HH corresponding to 0.25 lines on a pixel data basis. Finally, the quantization / encoding unit 104 quantizes / encodes the DWT coefficient data of the subband LL of decomposition level 2 corresponding to 0.25 lines on the pixel data basis. As a result, the DWT coefficient data for 1.0 line is quantized / encoded using the entire period during which one line of pixel data is input.

  Thereafter, the same processing is repeated until the last line of the image data.

  The reason why the coefficients held in the subband line buffer 103 are the coefficient data of the decomposition level 1 subband LH will be described with reference to FIG.

  FIG. 10 is a diagram showing the number of horizontal DWT coefficients of each subband when an image having a horizontal resolution (number of pixels) of 9 is subjected to DWT conversion up to decomposition level 2. FIG.

  As shown in FIG. 10, when the horizontal resolution is an odd number, the number of horizontal coefficients of each subband at decomposition level 1 belongs to the subband of the low frequency component (LH = 5 at decomposition level 1) or high frequency. It depends on whether it belongs to components (HL, HH = 4 at decomposition level 1). Therefore, when the level 1 DWT coefficient data input to the quantization / encoding unit 104 in the same period is subband LH, HH or LH, HL, the number of coefficients to be quantized / encoded in one line period is Different timing adjustment is required. Therefore, the number of coefficients to be quantized / encoded in one line period can be made equal by setting the decomposition level 1 DWT coefficients input to the quantizer / encoder 104 in the same period as subbands HL and HH. Because.

DESCRIPTION OF SYMBOLS 103 ... Subband line buffer, 104 ... Quantization / encoding part, 300 ... Wavelet transformation part, 301 ... Lev1-DWT transformation part, 302 ... Lev2-DWT transformation part

Claims (7)

An image encoding device for encoding image data,
Conversion means for inputting image data to be encoded, wavelet transforming the input image data, and generating conversion coefficient data of a plurality of subbands;
Encoding means for quantizing and encoding the coefficient data of each subband obtained by the conversion means;
Buffer means for temporarily holding the conversion coefficient data obtained by the conversion means,
Among the plurality of subbands generated by the conversion means, the conversion coefficient data of a preset subband is supplied to the encoding means without going through the buffer means,
The image coding apparatus, wherein transform coefficient data of another subband of the plurality of subbands is supplied to the buffer means.   Of the plurality of subbands generated by the converting means, subband transform coefficient data having a minimum decomposition level and including a high-frequency component is supplied to the encoding means without passing through the buffer means. The image encoding device according to claim 1.   Of the subbands LL, HL, LH, and HH generated by the conversion means, the coefficient data of the subband HH is supplied to the encoding means without going through the buffer means, and the coefficient data of the subband LL is 3. The image coding apparatus according to claim 1, wherein the image coding apparatus supplies the buffer means.   The coefficient data of subbands HL and HH are supplied to the encoding means without going through the buffer means, and the coefficient data of subbands LH and LL are supplied to the buffer means. The image encoding device described.   The encoding means reads coefficient data, performs quantization and encoding so that the total number of transform coefficient data read from the buffer means is equal to the number of pixels in the horizontal direction of the image to be encoded. The image coding apparatus according to claim 1, wherein the image coding apparatus is an image coding apparatus. 6. The encoding unit according to any one of claims 1 to 5, wherein the encoding unit preferentially reads, quantizes, and encodes subband coefficient data having a low wavelet transform decomposition level from the buffer unit. The image encoding device according to item. The conversion means has a plurality of wavelet conversion means,
Here, the i-th (i ≧ 2) wavelet transform unit inputs the transform coefficient data of the decomposition level i−1 subband LL generated by the i−1th wavelet transform unit, and the decomposition level i subband. Output conversion coefficient data of LL, HL, LH, HH.
The first wavelet transforming means, which is the first of the plurality of wavelet transforming means, is configured to encode two of the generated decomposition level 1 subbands HL, LH, and HH without interposing the buffer means. And supplying the remaining one to the buffer means,
The other wavelet transform means excluding the first wavelet transform means supplies the generated subbands HL, LH and HH of the respective decomposition levels to the buffer means,
The encoding means includes
Quantization and coding are performed by inputting the same number of transform coefficient data as the number of pixels in the horizontal direction of the image data to be encoded while giving priority to the lower decomposition level from the conversion means or the buffer means. The image encoding device according to any one of claims 1 to 6, wherein:

Images