KR102142946B1 - Method and apparatus for lossless image data compression using multiple dpcm - Google Patents

Abstract

The present invention relates to an apparatus and method for compressing lossless image data using multiple DPCM techniques, wherein the apparatus receives pixel data, performs multiple DPCM, and outputs a DPCM mode and DPCM result through minimum cost prediction, the A sign conversion unit that outputs a sign value and a magnitude value as a result of performing sign conversion by receiving the DPCM result, and performing VSC encoding in parallel by receiving the sign value and the magnitude value, and A parallel VSC encoder unit that outputs VSC, a K-splitter unit that performs K-split by receiving the magnitude value and outputs a minimum K, a quotient value, and a residual value, and receives the quotient value And a parallel GR encoder unit that performs GR encoding in parallel and outputs unary data, and a data packing unit that packs data according to the DPCM mode and outputs compressed data.

Description

Lossless image data compression apparatus and method using multiple DPCM techniques {METHOD AND APPARATUS FOR LOSSLESS IMAGE DATA COMPRESSION USING MULTIPLE DPCM}

The present invention relates to a lossless image data compression technique, and more particularly, to a lossless image data compression apparatus and method using a multiple DPCM technique capable of selecting a DPCM mode with a minimum cost by performing multiple DPCMs on an image.

As the image resolution of the latest mobile devices and GPUs increases rapidly, the amount of memory bandwidth required to refer to the image stored in the frame buffer increases significantly, and the resulting increase in the amount of memory access has a significant impact on performance degradation and increased power consumption. Can cause. Therefore, there is a need for a lossless image data compression technique supporting high-efficiency compression.

Korean Patent Registration No. 10-1519653 (2015.05.06) relates to an image lossless and lossy compression device. The compression device consists of a simple spatial predictor and a GR encoder to operate in block units, thereby processing large-capacity image data and memory. It is suitable for reducing the bandwidth, and it is easy to implement a chip due to the low complexity of the compression device, and has advantageous advantages in terms of data processing speed and cost.

Korean Patent Registration No. 10-0837410 (2008.06.04) relates to a subjective lossless image data compression method and apparatus, compressing current image data according to each of a plurality of modes for compressing current image data, and compressing current image data. By checking whether the image data can be expressed in bits of a fixed length, selecting any one of the modes confirmed to be expressible in bits of a fixed length, and outputting the current data compressed according to the selected mode. The complexity of the video encoder/decoder system can be drastically reduced, and the constant bit rate (CBR) of picture units required by LCD DCC (Liquid Crystal Display Dynamic Capacitance Compensation) devices can be accurately matched.

Korean Patent Registration No. 10-1519653 (2015.05.06) Korean Patent Registration No. 10-0837410 (2008.06.04)

An embodiment of the present invention is to provide a lossless image data compression apparatus and method using a multiple DPCM technique capable of selecting a DPCM mode with a minimum cost by performing multiple DPCM on an image.

An embodiment of the present invention is lossless image data using a multiplex DPCM technique that performs prediction of selecting a DPCM mode with the least cost by performing multiple horizontal/vertical DPCM and horizontal/vertical DDPCM, and performing entropy coding with an encoding algorithm. It is intended to provide a compression apparatus and method.

An embodiment of the present invention provides a lossless image data compression apparatus and method capable of predicting the minimum cost well for images having horizontal spatial locality or vertical spatial locality by reflecting both horizontal and vertical directions through multiple DPCMs. I want to.

An embodiment of the present invention is to provide a lossless image data compression apparatus and method capable of predicting a minimum cost well for an image in which pixel values rapidly change because DDPCM that performs DPCM once more is used.

Among embodiments, a lossless image data compression apparatus using a multiple DPCM technique receives pixel data, performs multiple DPCM, and outputs a DPCM mode and DPCM result through minimum cost prediction, and receives the DPCM result and encodes it. A sign conversion unit that outputs a sign value and a magnitude value as a result of performing the conversion, and a parallel VSC that performs VSC encoding in parallel by receiving the sign value and the magnitude value, and outputs VSC Encoder unit, a K-splitter unit that performs K-split by receiving the size value and outputs a minimum K, a quotient value, and a residual value, and performs GR encoding in parallel by receiving the quotient value. And a parallel GR encoder unit configured to output unary data and a data packing unit that packs data according to the DPCM mode and outputs compressed data.

The multi-DPCM unit is a parallel horizontal DPCM module that outputs a first DPCM error by performing parallel horizontal DPCM (Parallel Horizontal DPCM) based on the pixel data, and a parallel horizontal DDPCM based on the pixel data and the first DPCM error. A parallel horizontal DDPCM module configured to output a first DDPCM error, a parallel vertical DPCM module configured to output a second DPCM error by performing parallel vertical DPCM (parallel vertical DPCM) based on the pixel data, the pixel data and the second A parallel vertical DDPCM module that outputs a second DDPCM error by performing parallel vertical DDPCM based on a DPCM error, and the minimum cost prediction based on the first and second DPCM errors and the first and second DDPCM errors Thus, it may include a DPCM cost calculation module for outputting the DPCM mode and the DPCM result.

The multiple DPCM unit may perform a pair of DPCMs and a pair of DDPCMs in parallel in a process of sequentially performing DPCM and DDPCM.

Each of the DPCM modules and the DDPCM modules includes a plurality of subtractors, and each of the subtractors may perform a subtraction operation on a pair of inputs and output as an operation result.

The DPCM cost calculation module may perform the minimum cost estimation through the following equation based on the first and second DPCM errors and the first and second DDPCM errors.

[Mathematics]

(Here, n is the number of pixels, hd _k is the result of horizontal DPCM, vd _k is the result of vertical DPCM, hdd _k is the result of horizontal DDPCM, and vdd _k is the result of vertical DDPCM.)

The DPCM cost calculation module receives operation results from each of the DPCM modules and the DDPCM modules to perform code conversion, and the DPCM modules and the DDPCM modules are code converted. Accumulators for accumulating calculation results, a comparator for outputting a DPCM mode associated with a minimum cost as a result of comparing each of the accumulated results output by the accumulators, and the calculation result according to the DPCM mode Among them, multiplexers (MUXs) that select and output the DPCM result may be included.

The multiplexers are a first multiplexer that selects and outputs a first selection result from calculation results of the parallel horizontal DPCM module and the parallel horizontal DDPCM module according to the DPCM mode, and the parallel vertical DPCM module and the parallel A second multiplexer that selects and outputs a second selection result from the calculation results of the vertical DDPCM module, and a third multiplexer that selects and outputs the DPCM result from among the first and second selection results according to the DPCM mode. I can.

The first and third multiplexers may output the DPCM mode received as an input together with a selection result.

Among embodiments, the lossless image data compression method using the multiple DPCM technique includes the steps of: (a) receiving pixel data, performing multiple DPCM, and outputting a DPCM mode and DPCM result through minimum cost prediction, (b) the DPCM result. Receiving and outputting a sign value and a magnitude value as a result of performing sign conversion, (c) receiving the sign value and the magnitude value, performing VSC encoding in parallel, and performing VSC Outputting, (d) performing a K-split by receiving the size value, and outputting a minimum K, a quotient value, and a remaining value, (e) receiving the quotient value, Performing GR encoding in parallel and outputting unary data; and (f) packing data according to the DPCM mode and outputting compressed data.

The step (a) includes: (a1) performing a pair of DPCMs in parallel, (a2) performing a pair of DDPCMs in parallel based on the operation result of each of the pair of DPCMs, and ( a3) performing the minimum cost prediction based on the operation result of the pair of DPCMs and the operation result of the pair of DDPCMs, and outputting the DPCM mode and the DPCM result.

The step (a3) may include performing the minimum cost prediction through the following equation based on the calculation result of the pair of DPCMs and the calculation result of the pair of DDPCMs.

[Mathematics]

(Here, n is the number of pixels, hd _k is the result of horizontal DPCM, vd _k is the result of vertical DPCM, hdd _k is the result of horizontal DDPCM, and vdd _k is the result of vertical DDPCM.)

The step (a3) is a step of performing code conversion by receiving the operation result of the pair of DPCMs and the operation result of the pair of DDPCMs, and the result of code conversion for each of DPCM modules and DDPCM modules. Accumulating each, outputting a DPCM mode associated with a minimum cost as a result of comparing the accumulated results of each of the DPCM modules and DDPCM modules, and selecting the DPCM result from among the calculation results according to the DPCM mode It may include the step of outputting.

The step of selecting and outputting the DPCM result may include selecting and outputting a first selection result from among calculation results of a parallel horizontal DPCM module and a parallel horizontal DDPCM module according to the DPCM mode, and a parallel vertical DPCM module according to the DPCM mode. Selecting and outputting a second selection result from among the calculation results of the parallel vertical DDPCM module, and selecting and outputting the DPCM result from among the first and second selection results according to the DPCM mode.

The disclosed technology can have the following effects. However, since it does not mean that a specific embodiment should include all of the following effects or only the following effects, it should not be understood that the scope of the rights of the disclosed technology is limited thereby.

The apparatus and method for compressing lossless image data according to an embodiment of the present invention perform prediction to select a DPCM mode with the least cost by performing multiple horizontal/vertical DPCM and horizontal/vertical DDPCM, and perform entropy coding with an encoding algorithm. can do.

The apparatus and method for compressing lossless image data according to an embodiment of the present invention can well predict the minimum cost for both images having horizontal spatial locality or vertical spatial locality because both horizontal and vertical directions are reflected through multiple DPCMs. .

The apparatus and method for compressing lossless image data according to an embodiment of the present invention uses DDPCM to perform DPCM once more, so that a minimum cost can be predicted well for an image in which pixel values rapidly change.

1 is a diagram illustrating a hardware structure of a lossless image data compression apparatus according to the present invention.
2 is a diagram illustrating a process of performing a horizontal/vertical DPCM according to the present invention.
3 is a diagram illustrating a process of performing a horizontal/vertical DDPCM according to the present invention.
4 is a diagram illustrating a configuration of a multiple DPCM unit according to the present invention.
5 is a diagram illustrating a hardware structure of a horizontal DPCM/DDPCM according to the present invention.
6 is a diagram illustrating a hardware structure of a vertical DPCM/DDPCM according to the present invention.
7 is a diagram illustrating a hardware structure of a DPCM cost calculation module according to the present invention.
8 is a flowchart illustrating a lossless image data compression process according to the present invention.

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be interpreted as being limited by the embodiments described in the text. That is, since the embodiments can be variously changed and have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing technical ideas. In addition, the purpose or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such an effect, and the scope of the present invention should not be understood as being limited thereby.

Meanwhile, the meaning of terms described in the present application should be understood as follows.

Terms such as "first" and "second" are used to distinguish one component from other components, and the scope of rights is not limited by these terms. For example, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component.

When a component is said to be "connected" to another component, it may be understood that other components may exist in the middle, although they may be directly connected to other components. On the other hand, when a component is said to be "directly connected" to another component, it should be understood that no other component exists in the middle. On the other hand, other expressions describing the relationship between the components, that is, "between" and "immediately between" or "adjacent to" and "directly neighboring to" should be interpreted similarly.

Singular expressions are to be understood as including plural expressions unless the context clearly indicates otherwise, and terms such as “comprises” or “have” are used features, numbers, steps, actions, elements, parts or the like. It is to be understood that a combination is intended to indicate the existence, and does not preclude the existence or addition possibility of one or more other features or numbers, steps, operations, components, parts or combinations thereof.

In each step, the identification code (for example, a, b, c, etc.) is used for convenience of explanation. The identification code does not describe the order of each step, and each step clearly identifies a specific order in context. Unless stated, it may occur in a different order than specified. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be embodied as computer readable code on a computer readable recording medium, and the computer readable recording medium includes all kinds of recording devices in which data readable by a computer system is stored. . Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, HDD, and SSD. In addition, the computer-readable recording medium can be distributed over network coupled computer systems so that the computer-readable code is stored and executed in a distributed fashion.

All terms used herein have the same meaning as generally understood by a person skilled in the art to which the present invention pertains, unless otherwise defined. The terms defined in the commonly used dictionary should be interpreted as being consistent with meanings in the context of the related art, and cannot be interpreted as having ideal or excessively formal meanings unless explicitly defined in the present application.

1 is a diagram illustrating a hardware structure of a lossless image data compression apparatus according to the present invention.

Referring to FIG. 1, the lossless image data compression apparatus 100 includes a multiple DPCM unit (MDPCM) 110 performing multiple DPCMs, and a negative DPCM result with a positive value. ), a parallel VSC encoder unit 130 that performs VSC encoding in parallel, and a K-splitter unit that performs K Spilt. unit) 140, a parallel GR encoder unit 150 for performing GR encoding in parallel, and a data packing unit for generating compressed data by packing data ( 160) may be included.

The multiple DPCM unit (MDPCM) 110 may receive pixel data, perform multiple DPCM, and output a DPCM mode and a DPCM result through minimum cost prediction. More specifically, the multi-DPCM unit 110 may determine the DPCM mode of the lowest cost by applying various DPCM techniques using original pixel data. The multiple DPCM unit 110 can provide two outputs, one of which is a DPCM mode (DPCM Mode) is transferred to the data packing unit 160, and the other DPCM result (DPCM Result) is a code conversion unit ( 120).

The sign converter unit 120 may receive the DPCM result and output a sign value and a magnitude value as a result of performing the sign conversion. More specifically, when the input value is a negative value, the sign conversion unit 120 may convert it into a two's complement positive value (2's compliment positive), and the sign value of each value ( sign value) can be transferred to the parallel VSC encoder unit 130. In addition, the code conversion unit 120 may transmit a magnitude value of each value to the parallel VSC encoder unit 130 and the K-splitter unit 140.

The parallel VSC encoder unit 130 may receive a sign value and a magnitude value, respectively, perform VSC encoding in parallel, and output VSC as a result. That is, the parallel VSC encoder unit 130 may perform VSC encoding in parallel to compress sign data by omitting a sign value having a magnitude value of 0.

The K-splitter unit 140 may receive a sign value, perform K-split, and output a minimum K, a quotient value, and a remaining value. The K-splitter unit 140 may perform a K split with K=0, 1, 2, or 3 on input magnitude values, and may select K of the minimum cost. The K-splitter unit 140 may transmit the selected minimum K and remaining values to the data packing unit 160, and transmit quotient values to the parallel GR encoder unit 150. I can.

The parallel GR encoder unit 150 may receive a quotient value, perform GR encoding in parallel, and output unary data. The parallel GR encoder unit 150 may deliver unary data to the data packing unit 160 as a result of GR encoding.

The data packing unit 160 may pack data according to the DPCM mode and output compressed data. That is, the data packing unit 160 includes a seed value of the original pixel data, a DPCM mode received from the multiple DPCM unit 110, and a VSC, K-splitter received from the parallel VSC encoder unit 130. Compressed data is generated as a result by performing data packing based on the minimum K received from the unit 140 and the remaining value (Remainder), and the unary data received from the parallel GR encoder unit 150 can do.

2 is a diagram illustrating a process of performing a horizontal/vertical DPCM according to the present invention.

Referring to FIG. 2, horizontal/vertical DPCM may be performed by multiple DPCM units 110 of the lossless image data compression apparatus 100.

Methods for applying DPCM to a 2D image include horizontal DPCM (horizontal DPCM) and vertical DPCM (vertical DPCM). In the horizontal DPCM, the DPCM algorithm is applied in a horizontal direction, whereas in the vertical DPCM, the DPCM algorithm can be applied in a vertical way.

In addition, the equation used in each DPCM is as follows. That is, Equation 1 is an equation used in horizontal DPCM, and Equation 2 is an equation used in vertical DPCM.

[Equation 1]

[Equation 2]

In particular, if there is no previous pixel value to be referenced by the DPCM algorithm, such as the leftmost pixel or uppermost pixel, the leftmost pixel is the upper pixel. A difference value can be calculated with reference, and an uppermost pixel can be calculated with reference to a left pixel.

Fig. 2 (a) shows a process of performing horizontal DPCM and vertical DPCM for an input block having a size of 4*4. In the case of vertical DPCM, a result value may be calculated by calculating a difference between a current pixel and a left pixel. However, since the leftmost pixels with j=0 do not have a left pixel to be referred to, a difference from the upper pixel can be calculated. A first pixel with i=0 and j=0 may be marked as a seed value for performing inverse DPCM (DPCM).

Meanwhile, the vertical DPCM is in contrast to the horizontal DPCM, and the result value can be calculated by calculating a difference between a current pixel and an upper pixel. However, since the uppermost pixels with i=0 do not have an upper pixel to be referred to, a difference from the left pixel can be calculated.

Figure (b) of FIG. 2 shows an example of performing a horizontal DPCM and a vertical DPCM for an example block having a size of 4*4. Pixels at positions i=0, j≠0 or i≠0, j=0 may have the same result of horizontal DPCM and vertical DPCM. However, the results of each DPCM for pixels located other than that may be different. For example, the result of each DPCM of the pixel at the position i=1 and j=1 is 5 * (Err _h (1,1) = P(1,1)-P(0,1)) for horizontal DPCM And, in the case of vertical DPCM, -1 * (Err _v (1,1) = P(1,1)-P(1,0)).

3 is a diagram illustrating a process of performing a horizontal/vertical DDPCM according to the present invention.

Referring to FIG. 3, the horizontal/vertical DDPCM may be performed in multiple DPCM units 110 of the lossless image data compression apparatus 100.

The multiple DPCM unit 110 may use a horizontal DDPCM that performs DDPCM based on the horizontal DPCM and a vertical DDPCM that performs DDPCM based on the vertical DPCM. In FIG. 3, a process of performing horizontal DDPCM and vertical DDPCM for an input block having a size of 4*4 can be confirmed. In this case, d _h a denotes a difference between a and a left value, and d _v a denotes a difference between a and an upper value.

For example, in the first step of the horizontal DDPCM, d _h a ₀₁ is a ₀₁ -a ₀₀ , the difference value between a ₀₁ and a ₀₀ , the left value, and d _v a ₁₀ is a _10, and the upper value (upper value) and the value of the difference between a ₀₀ a ₁₀ -a _00. The same rule, the second stage of the horizontal _{DDPCM (2nd step), d v} d h a 11 is the d _h a ₁₁ d _h a ₁₁ and upper value (upper vaule) of d _h a ₀₁ difference value (difference value) between the - d _h a _01, and reorganizing this expression is (a ₁₁ -a ₁₀ )-(a ₀₁ -a ₀₀ ).

The process of performing horizontal DDPCM and vertical DDPCM is as follows. First, the corresponding DPCM is performed in the first step of DDPCM. When DPCM is performed, for a result value of i> 1, the difference value for the left DPCM result value is calculated for horizontal DDPCM, and the difference value for the upper DPCM result value is calculated for horizontal DDPCM.

For the DPCM result value of i=1, the difference value for the upper DPCM result value is calculated for the horizontal DDPCM, and the difference value for the left DPCM result value is calculated for the horizontal DDPCM. If there is no previous DPCM result value to be referenced by the DDPCM algorithm, such as the leftmost DPCM result and the uppermost DPCM result, the leftmost DPCM result is different by referring to the upper DPCM result. A value is calculated, and a difference value is calculated by referring to the left DPCM result for the uppermost DPCM result.

4 is a diagram illustrating a configuration of a multiple DPCM unit according to the present invention.

4, the multiple DPCM unit 110 includes a parallel horizontal DPCM module 410, a parallel horizontal DDPCM module 430, a DPCM cost calculation module 450, a parallel vertical DPCM module 470, and a parallel vertical DDPCM module ( 490).

The multiple DPCM unit 110 may receive original pixel data as an input and perform multiple DPCM, and as a result, may provide a DPCM result and a DPCM mode as an output.

More specifically, the parallel horizontal DPCM module 410 may output a first DPCM error by performing parallel horizontal DPCM (Parallel Horizontal DPCM) based on pixel data. The parallel horizontal DDPCM module 430 may output a first DDPCM error by performing parallel horizontal DDPCM (Parallel Horizontal DDPCM) based on the pixel data and the first DPCM error. The DPCM cost calculation module 450 may pack data according to the DPCM mode and output compressed data. The parallel vertical DPCM module 470 may output a second DPCM error by performing parallel vertical DPCM (parallel vertical DPCM) based on the pixel data. The parallel vertical DDPCM module 490 may output a second DDPCM error by performing parallel vertical DDPCM (parallel vertical DDPCM) based on the pixel data and the second DPCM error.

In one embodiment, the multiple DPCM unit 110 may perform a pair of DPCMs and a pair of DDPCMs in parallel in a process of sequentially performing DPCM and DDPCM. More specifically, the multi-DPCM unit 110 may perform horizontal DPCM in parallel in the parallel horizontal DPCM module 410 for input pixel data, and perform vertical DPCM in parallel in the parallel vertical DPCM module 470. I can. Each calculated DPCM error may be transmitted to the DPCM cost calculation module 450 for cost calculation and may be transmitted to the corresponding DDPCM module. The parallel horizontal/vertical DDPCM modules 430 and 490 may perform DDPCM in parallel based on the received DPCM error, and then transmit the result to the DPCM cost calculation module 450.

That is, each DPCM module can perform horizontal DPCM and vertical DPCM for input pixels in parallel, and can deliver the result to the corresponding DDPCM module. In addition, each DDPCM module may perform each DDPCM operation on the result in parallel when the DPCM operation is completed. All DPCM results may be delivered to the DPCM cost calculation module 450.

The DPCM cost calculation module 450 may calculate a cost for each DPCM error and may determine a DPCM mode of a minimum cost. When the DPCM mode is determined, the DPCM mode may be transmitted to the data packing unit 160 and the corresponding DPCM error may be transmitted to the code conversion unit 120.

In an embodiment, the DPCM cost calculation module 450 may perform the minimum cost prediction through Equation 3 below based on the first and second DPCM errors and the first and second DDPCM errors.

[Equation 3]

Here, n is the number of pixels, hd _k is a result of horizontal DPCM, vd _k is a result of vertical DPCM, hdd _k is a result of horizontal DDPCM, and vdd _k may correspond to a result of vertical DDPCM.

That is, when the horizontal DPCM, vertical DPCM, horizontal DDPCM, and vertical DDPCM are all performed, the DPCM cost calculation module 450 selects an optimal prediction function for entropy coding for the result of each prediction function. I can. In this case, the entropy coding cost may be calculated through the sum of the absolute values of each DPCM result value.

The lossless image data compression apparatus 100 may perform golumb-rice (GR) encoding as entropy coding, and the bit length of the GR encoding may increase as the value increases. Therefore, if a DPCM having the smallest sum of DPCM result values is selected, entropy coding cost may be lowest.

5 is a diagram illustrating a hardware structure of a horizontal DPCM/DDPCM according to the present invention.

5, Figure (a) is a hardware structure of a parallel horizontal DPCM module 410 of multiple DPCM units 110, Figure (b) is a parallel horizontal DDPCM module 430 of multiple DPCM units 110 It is a hardware structure.

In one embodiment, the parallel horizontal DPCM module 410 and the parallel horizontal DDPCM module 430 include a plurality of subtractors 510, and each subtractor 510 performs a subtraction operation on a pair of inputs. Then, it can be output as an operation result. For example, in Figure (a), the first subtractor 511 included in the parallel horizontal DPCM module 410 performs a subtraction operation based on a pair of inputs a ₀₀ and a ₁₀ , and d _v a You can print ₁₀ . In Figure (b), the second subtractor 513 included in the parallel horizontal DDPCM module 430 performs a subtraction operation based on a pair of inputs d _v a ₁₀ and d _h a ₁₁ , and d _v d _h a ₁₁ can be printed.

6 is a diagram illustrating a hardware structure of a vertical DPCM/DDPCM according to the present invention.

Referring to Figure 6, Figure (c) is a hardware structure of the parallel vertical DPCM module 470 of the multiple DPCM units 110, Figure (d) is the parallel vertical DDPCM module 490 of the multiple DPCM units 110 It is a hardware structure.

In one embodiment, the parallel vertical DPCM module 470 and the parallel vertical DDPCM module 490 include a plurality of subtractors 510, and each subtractor 510 performs a subtraction operation on a pair of inputs. Then, it can be output as an operation result. For example, in Figure (c), the third subtractor 515 included in the parallel vertical DPCM module 470 performs a subtraction operation based on a pair of inputs a ₀₀ and a ₁₀ , and d _h a ₀₁ can be printed. In Figure (d), the fourth subtractor 517 included in the parallel vertical DDPCM module 490 performs a subtraction operation based on a pair of inputs d _h a ₀₁ and d _v a ₁₁ , and d _h d _v a ₁₁ can be printed.

7 is a diagram illustrating a hardware structure of a DPCM cost calculation module according to the present invention.

Referring to FIG. 7, the lossless image data compression apparatus 100 may include multiple DPCM units 110, and the multiple DPCM units 110 may include a DPCM cost calculation module 450.

In one embodiment, the DPCM cost calculation module 450 is a sign converter 710, an accumulator 730, a comparator 750, and a multiplexer (MUX) 771, 773, 775. It may include. More specifically, the sign converter 710 may perform code conversion by receiving operation results from each of the DPCM modules and the DDPCM modules. The accumulator 730 may accumulate a code-converted operation result for each of the DPCM modules and the DDPCM modules. The comparator 750 may output a DPCM mode associated with the minimum cost as a result of comparing the accumulated results output by the accumulators 730 with each other. The multiplexer (MUX) 771, 773, and 775 may select and output a DPCM result from among operation results according to the DPCM mode.

In one embodiment, the multiplexer (MUX) is a first multiplexer 771 that selects and outputs a first selection result among the calculation results of the parallel horizontal DPCM module 410 and the parallel horizontal DDPCM module 430 according to the DPCM mode, A second multiplexer 773 that selects and outputs a second selection result among the calculation results of the parallel vertical DPCM module 470 and the parallel vertical DDPCM module 490 according to the DPCM mode, and the first and second selection according to the DPCM mode A third multiplexer 775 that selects and outputs a DPCM result among the results may be included.

As a result, the multiplexer MUX may select and output DPCM result values according to the DPCM mode through a combination of the first to third multiplexers 771, 773, and 775. For example, in FIG. 7, the multiplexer (MUX) may output e ₀₁ , ..., e ₇₆ , and e ₇₇ as final outputs through a combination of the first to third multiplexers 771, 773, and 775. .

In an embodiment, the first multiplexer 771 and the third multiplexer 775 may output DPCM modes received as inputs together with a selection result. In FIG. 7, the first multiplexer 771 may transmit the DPCM mode together with the selection result to the third multiplexer 775, and the third multiplexer 775 may output the DPCM mode together with the selection result. On the other hand, since the third multiplexer 775 receives the DPCM mode from the first multiplexer 771, the second multiplexer 773, unlike the first and third multiplexers 771 and 775, only the selection result according to the DPCM mode. Can be printed.

8 is a flowchart illustrating a lossless image data compression process according to the present invention.

Referring to FIG. 8, the lossless image data compression apparatus 100 may receive pixel data through multiple DPCM units 110 to perform multiple DPCMs and output a DPCM mode and DPCM result through minimum cost prediction (step S810). The lossless image data compression apparatus 100 may receive the DPCM result through the code conversion unit 120 and output a sign value and a magnitude value as a result of performing the code conversion (step S820). ).

In addition, the lossless image data compression apparatus 100 may receive a sign value and a size value through the parallel VSC encoder unit 130 to perform VSC encoding in parallel and output the VSC (step S830). The lossless image data compression apparatus 100 may receive a size value through the K-splitter unit 140, perform K-split, and output a minimum K, a quotient value, and a remaining value. (Step S840).

In addition, the lossless image data compression apparatus 100 may receive a quotient value through the parallel GR encoder unit 110, perform GR encoding in parallel, and output unary data (step S850). The lossless image data compression apparatus 100 may pack data according to the DPCM mode through the data packing unit 160 and output compressed data (step S860).

Although described above with reference to preferred embodiments of the present invention, those skilled in the art variously modify and change the present invention without departing from the spirit and scope of the present invention as set forth in the claims below. You can understand that you can.

100: lossless image data compression device
110: multiple DPCM unit 120: code conversion unit
130: parallel VSC encoder unit 140: K-splitter unit
150: parallel GR encoder unit 160: data packing unit
410: parallel horizontal DPCM module 430: parallel horizontal DDPCM module
450: DPCM costing module 470: Parallel vertical DPCM module
490: parallel vertical DDPCM module
510: subtractor 511: first subtractor
513: second subtractor 515: third subtractor
517: fourth subtractor
710: code converter 730: accumulator
750: comparator 771: first multiplexer
773: second multiplexer 775: third multiplexer

Claims (13)

A multi-DPCM unit for receiving pixel data, performing multiple DPCMs, and outputting a DPCM mode and DPCM results associated with the minimum cost through prediction of minimum cost based on respective errors occurring in the corresponding process;
A sign conversion unit for receiving the DPCM result and outputting a sign value and a magnitude value as a result of performing sign conversion;
A parallel VSC encoder unit receiving the sign value and the magnitude value, performing VSC encoding in parallel, and outputting VSC;
A K-splitter unit receiving the size value, performing K-split, and outputting a minimum K, a quotient value, and a remaining value;
A parallel GR encoder unit receiving the quotient value, performing GR encoding in parallel and outputting unary data; And
Lossless image data compression apparatus using multiple DPCM techniques including a data packing unit that packs data according to the DPCM mode and outputs compressed data.
The method of claim 1, wherein the multiple DPCM units
A parallel horizontal DPCM module for outputting a first DPCM error by performing parallel horizontal DPCM (Parallel Horizontal DPCM) based on the pixel data;
A parallel horizontal DDPCM module configured to output a first DDPCM error by performing parallel horizontal DDPCM based on the pixel data and the first DPCM error;
A parallel vertical DPCM module for outputting a second DPCM error by performing parallel vertical DPCM (parallel vertical DPCM) based on the pixel data;
A parallel vertical DDPCM module configured to output a second DDPCM error by performing parallel vertical DDPCM based on the pixel data and the second DPCM error; And
And a DPCM cost calculation module configured to output the DPCM mode and the DPCM result by performing the minimum cost prediction based on the first and second DPCM errors and the first and second DDPCM errors. Lossless image data compression device using DPCM technique.
The method of claim 2, wherein the multiple DPCM units
A lossless image data compression apparatus using a multi-DPCM technique, characterized in that in a process of sequentially performing DPCM and DDPCM, a pair of DPCMs and a pair of DDPCMs are respectively performed in parallel.
The method of claim 2, wherein each of the DPCM modules and the DDPCM modules
A lossless image data compression apparatus using a multiple DPCM technique, comprising a plurality of subtractors, each subtractor performing a subtraction operation on a pair of inputs and outputting the result as an operation result.
The method of claim 1, wherein the DPCM cost calculation module
The lossless image data compression apparatus using a multi-DPCM technique, characterized in that the minimum cost prediction is performed through the following equation based on the first and second DPCM errors and the first and second DDPCM errors.
[Mathematics]

(Here, n is the number of pixels, hd _k is the result of horizontal DPCM, vd _k is the result of vertical DPCM, hdd _k is the result of horizontal DDPCM, and vdd _k is the result of vertical DDPCM.)
The method of claim 2, wherein the DPCM cost calculation module
Sign converters for performing code conversion by receiving operation results from each of the DPCM modules and the DDPCM modules;
Accumulators for accumulating the result of the code-converted operation for each of the DPCM modules and the DDPCM modules;
A comparator for outputting a DPCM mode associated with a minimum cost as a result of comparing the accumulated results outputted by the accumulators with each other; And
And multiplexers (MUXs) for selecting and outputting the DPCM result from among the calculation results according to the DPCM mode.
The method of claim 6, wherein the multiplexers are
A first multiplexer for selecting and outputting a first selection result from among calculation results of the parallel horizontal DPCM module and the parallel horizontal DDPCM module according to the DPCM mode;
A second multiplexer for selecting and outputting a second selection result from among operation results of the parallel vertical DPCM module and the parallel vertical DDPCM module according to the DPCM mode; And
And a third multiplexer that selects and outputs the DPCM result from among the first and second selection results according to the DPCM mode.
The method of claim 7, wherein the first and third multiplexers are
A lossless image data compression apparatus using a multi-DPCM technique, characterized in that outputting the DPCM mode received as an input together with a selection result.
In the method performed in a lossless image data compression apparatus,
(a) receiving pixel data, performing multiple DPCM, and outputting a DPCM mode and DPCM results associated with the minimum cost, respectively, through prediction of a minimum cost based on respective errors occurring in the corresponding process;
(b) receiving the DPCM result and outputting a sign value and a magnitude value as a result of performing sign conversion;
(c) receiving the sign value and the magnitude value, performing VSC encoding in parallel, and outputting VSC;
(d) receiving the size value, performing K-split, and outputting a minimum K, a quotient value, and a remaining value;
(e) receiving the quotient value, performing GR encoding in parallel, and outputting unary data; And
(f) A method of compressing lossless image data using multiple DPCM techniques, including the step of packing data according to the DPCM mode and outputting compressed data.
The method of claim 9, wherein step (a)
(a1) performing a pair of DPCMs in parallel;
(a2) performing a pair of DDPCMs in parallel based on an operation result of each of the pair of DPCMs; And
(a3) performing the minimum cost prediction based on the operation result of the pair of DPCMs and the operation result of the pair of DDPCMs, and outputting the DPCM mode and the DPCM result. Lossless image data compression method using the technique.
The method of claim 10, wherein step (a3)
Compressing lossless image data using a multi-DPCM method, comprising the step of performing the minimum cost prediction through the following equation based on the operation result of the pair of DPCMs and the operation result of the pair of DDPCMs. Way.
[Mathematics]

(Here, n is the number of pixels, hd _k is the result of horizontal DPCM, vd _k is the result of vertical DPCM, hdd _k is the result of horizontal DDPCM, and vdd _k is the result of vertical DDPCM.)
The method of claim 10, wherein step (a3)
Performing code conversion by receiving an operation result of the pair of DPCMs and an operation result of the pair of DDPCMs;
Accumulating the code-converted calculation results for each of the DPCM modules and the DDPCM modules, respectively;
Outputting a DPCM mode associated with a minimum cost as a result of comparing the accumulated results of each of the DPCM modules and the DDPCM modules with each other; And
And selecting and outputting the DPCM result from among the calculation results according to the DPCM mode.
The method of claim 12, wherein the step of selecting and outputting the DPCM result
Selecting and outputting a first selection result from calculation results of a parallel horizontal DPCM module and a parallel horizontal DDPCM module according to the DPCM mode;
Selecting and outputting a second selection result from calculation results of a parallel vertical DPCM module and a parallel vertical DDPCM module according to the DPCM mode; And
And selecting and outputting the DPCM result from among the first and second selection results according to the DPCM mode.

Images