JP6280266B2 - Decoding device and decoding method - Google Patents

Description

  Embodiments described herein relate generally to a decoding device and a decoding method.

  Known encoding methods for image data include fixed-length encoding with a fixed code length and variable-length encoding with a non-uniform code length. For example, in MPEG-2, which is one of the encoding methods, the quantization parameter can be set in the range of 0 to 31, and it is known to perform 5-bit fixed length encoding. H. In coding systems such as H.264, it is known to employ variable length coding.

  When variable length coding is performed, encoded data having a higher compression rate than that of fixed length coding can be obtained. However, it is known that decoding of encoded data obtained by variable length encoding takes time compared to decoding of encoded data obtained by fixed length encoding. On the other hand, it is known that encoded data obtained by fixed-length encoding has a low compression rate efficiency, although the processing time required for decoding is shorter than that of variable-length encoding.

  Also disclosed is a technique for dividing variable-length encoded data into fixed-length packets with high importance and fixed-length packets with low importance, and retransmitting high-importance packets in the event of a transmission failure. Yes.

JP 2003-101416 A Japanese Patent No. 3662171 Japanese Patent No. 3669277 Japanese Patent No. 3669281

  However, in the past, it has been difficult to decode variable-length encoded data at high speed.

  The problem to be solved by the present invention is to provide a decoding device and a decoding method capable of reducing the decoding processing time of encoded data subjected to variable length encoding.

  The decoding device according to the embodiment includes a cutout unit and a decoding unit. The cutout unit sequentially cuts input data including a plurality of pieces of encoded data subjected to variable length encoding for each predetermined first code length that is equal to or larger than the maximum code length of the plurality of pieces of encoded data. Data is stored in the storage unit as stored data. The decoding unit sequentially reads the stored data stored in the storage unit, and performs variable length decoding on the encoded data included in the read storage data. The decoding unit includes a determination unit that determines whether variable length decoding based on the stored data is possible. When it is determined by the determination unit that variable length decoding is impossible, the extraction unit extracts the extraction data from the input data and stores it as the storage data in the storage unit.

The block diagram of a decoding apparatus. The flowchart which shows the procedure of a decoding process. The schematic diagram which shows encoded data. Explanatory drawing of extraction of extraction data. The schematic diagram of stored data. The schematic diagram of the data structure of a lookup table. The schematic diagram of the data structure of a lookup table. The schematic diagram of stored data. The schematic diagram of the data structure of a lookup table. Explanatory drawing of the process timing of a decoding process. The flowchart which shows the other procedure of a decoding process. The block diagram of a decoding apparatus. The flowchart which shows the procedure of a decoding process. The flowchart which shows the other procedure of a decoding process. The flowchart which shows the other procedure of a decoding process. The flowchart which shows the other procedure of a decoding process. Explanatory drawing of the process timing of a decoding process. The block diagram of an encoding apparatus. The schematic diagram of the data structure of output data. The flowchart which shows the procedure of an encoding process. Explanatory drawing of variable length encoding and packetization. The schematic diagram of the data structure of output data. The schematic diagram of the data structure of output data.

(Embodiment 1)
FIG. 1 is a block diagram illustrating a functional configuration of the decoding device 10 according to the present embodiment. The decoding device 10 is a device that performs image decoding. The decoding device 10 is incorporated in a device or system that mainly handles CG (computer graphics) images, such as a PC (personal computer) screen.

  The decoding device 10 is incorporated in a device implemented by hardware such as an image decoding LSI (Large Scale Integration). By performing decoding at high speed by the decoding process by the decoding device 10, for example, display responsiveness to the display can be improved.

  The decoding device 10 according to the present embodiment receives and decodes input data to be decoded, and outputs the decoded data.

  The input data received by the decoding device 10 includes a plurality of encoded data subjected to variable length encoding. The encoded data indicates data obtained by variable length encoding by an encoding device (described later in detail). Encoded data includes, but is not limited to, quantization parameter differences, motion vectors, and the like.

  The decryption device 10 includes a cutout unit 12, a storage unit 14, and a decryption unit 16.

  The cutout unit 12 cuts out the input data received by the decoding device 10 for each predetermined first code length.

  The first code length is a value that is greater than or equal to the maximum code length among the code lengths of a plurality of pieces of encoded data included in the received input data. The maximum code length indicates the maximum value among the code lengths that can be taken when the input data is variable-length encoded. The maximum code length is determined by the variable length coding method used. Further, the first code length is a value equal to or less than the data length of the input data. Note that the first code length is predetermined for each input data to be received by the decoding device 10 and is stored in advance in a memory or the like (not shown) of the decoding device 10.

  For example, it is assumed that the data length of input data is X bits (X is an integer of 1 or more). Further, it is assumed that the first code length is M bits (M is an integer of 1 to X). In this case, the cutout unit 12 cuts the input data every M bits that are the first code length. In the following description, the M bits that are the first code length may be simply referred to as “first code length M”.

  The storage unit 14 stores the cutting data cut out by the cutting unit 12. The storage unit 14 is a known storage medium. The decoding unit 16 performs variable length decoding on the encoded data included in the cut out data. The decoding unit 16 includes a determination unit 16A. The determination unit 16A analyzes data stored in the storage unit 14 (hereinafter referred to as storage data) and determines whether variable-length decoding based on the stored data is possible.

  FIG. 2 is a flowchart showing the procedure of the decoding process executed by the decoding device 10.

  When receiving the input data, the decoding device 10 executes the decoding process shown in FIG.

  First, the cutting unit 12 cuts the received input data for each first code length (step S100). Then, the cutting unit 12 stores the cut out data in the storage unit 14 (step S102). In the case where data is stored in the storage unit 14, the cutout unit 12 combines the newly cut out data at the end of this data and stores it in the storage unit 14. Although the details will be described later, the stored data indicates data remaining in the storage unit 14 among the cut data cut out by the cutout unit 12 and stored in the storage unit 14.

  Next, the decoding unit 16 decodes the encoded data included in the stored data stored in the storage unit 14 (step S104) (details will be described later). Then, the decoding unit 16 outputs the decoded data. The decoding unit 16 deletes the decoded encoded data from the stored data in the storage unit 14. For this reason, in the storage unit 14, undecoded undecoded data remains as stored data among the cut data stored in the storage unit 14 by the cutout unit 12.

  Next, the determination unit 16A analyzes the stored data stored in the storage unit 14 (step S106), and determines whether variable length decoding based on the stored data is possible (step S108) (details will be described later). .

  If an affirmative determination is made in step S108 (step S108: Yes), the process returns to step S104.

  On the other hand, if a negative determination is made in step S108 (step S108: No), the process proceeds to step S110. In step S110, the clipping unit 12 determines whether there is undecoded undecoded data in the input data to be decoded (step S110). For example, the cutout unit 12 determines whether or not all data of the input data has been cut out by the cutout unit 12 to perform the determination in step S110.

  If an affirmative determination is made in step S110 (step S110: Yes), the process returns to step S100. On the other hand, if a negative determination is made in step S110 (step S110: No), this routine ends.

  Next, details of the decoding process executed by the decoding device 10 will be described with a specific example.

  FIG. 3 is a schematic diagram illustrating an example of a plurality of encoded data included in input data. FIG. 3 shows a case where the encoded data is encoded data obtained by encoding a difference between quantization parameters (refer to delta_qp in FIG. 3 and may be simply referred to as delta_qp hereinafter). ing. The encoded data shown in FIG. 3 is variable-length encoded data corresponding to the difference between the quantization parameters. In FIG. 3, the code length indicates the code length of the corresponding encoded data.

  For example, as shown in FIG. 3, it is assumed that the value of delta_qp is −4 to +4, and the encoded data corresponding to each delta_qp is the value shown in FIG. Also assume that the code length of each encoded data is the value shown in FIG.

  Specifically, the encoded data corresponding to delta_qp “−2” is “0011” in binary number and the code length is “4”. In the following, when a binary number is expressed, “b” is added. That is, the binary number “0011” is expressed as “0011b”.

  The maximum code length in the plurality of pieces of encoded data illustrated in FIG. 3 is “6”. For this reason, when the input data is data including a plurality of pieces of encoded data shown in FIG. 2, the decoding apparatus 10 performs extraction with the maximum code length “6”.

  FIG. 4 is an explanatory diagram showing the extraction of the extraction data by the extraction unit 12. As shown in FIG. 4, the cutout unit 12 cuts out input data including a plurality of pieces of encoded data shown in FIG. 3 for each predetermined first code length M that is equal to or larger than the maximum code length of these encoded data. I do. In the present embodiment, the case where the first code length M is “6” is assumed. Further, description will be made assuming that the data length of the input data is X bits (X is an integer larger than M).

  In the present embodiment, the first code length M is described as a constant value for one input data. The first code length M is preferably a constant value for one input data from the viewpoint of performing decoding processing at high speed, but the value of the first code length M for one input data is You may change according to the arrangement | sequence of the encoding data contained in data. Specifically, when the code lengths of the encoded data are arranged in the order in which they are regularly changed, the first code length M may be changed according to the change rules.

  The cutout unit 12 cuts out the input data for each first code length M in order from the top of the input data. Then, the cutout unit 12 sequentially stores the cut out M-bit cutout data in the storage unit 14. As shown in FIG. 4, when there is residual data (for example, α bits) in the storage unit 14, the cutout unit 12 combines the cutout data with the end of the residual data. For this reason, for example, the storage unit 14 stores storage data having a data length of M + α bits by cutting (see step S100 in FIG. 2) and storing (see step S102 in FIG. 2) by the cutting unit 12. It becomes a state.

  Although details will be described later, the decoding unit 16 reads and decodes the encoded data from the stored data. At this time, the decoding unit 16 cuts out the encoded data to be decoded from the storage data stored in the storage unit 14 and decodes the extracted encoded data. Therefore, the storage unit 14 is in a state in which undecoded data is stored as stored data by the decoding unit 16 among the cut data cut out by the cutting unit 12.

  As shown in FIG. 4, the data length X of the input data may not be an integral multiple of the first code length M, which is a unit cut out by the cutout unit 12. In this case, the cutting unit 12 cuts the remaining data having a data length less than the first code length M (remainder value of X / M) at the time of last cutting when the input data is cut out in order from the top. Just put it out.

  The storage capacity (hereinafter referred to as memory size) of the storage unit 14 that stores the cut-out data is preferably minimal from the viewpoint of device miniaturization and the like. In the present embodiment, the memory size of the storage unit 14 is preferably greater than or equal to the memory size represented by the following formula (1).

  Memory size of storage unit 14 = M × 2-L (1)

  In formula (1), M represents the first code length. L indicates the minimum code length of the encoded data included in the input data.

  When the cutting unit 12 stores the cut data in the storage unit 14 in a state where no data is stored in the storage unit 14, the maximum value that the data length of the stored data stored in the storage unit 14 can take is the first value. The code length is M. When the decoding unit 16 extracts and decodes the encoded data to be decoded from the stored data, the data length of the stored data in the storage unit 14 is the data of the encoded data decoded from the data length of the stored data in the state before decoding. Subtracted by subtracting the length. For this reason, the maximum value that the data length of the storage data stored in the storage unit 14 in the state after decoding by the decoding unit 16 can take from the first code length M to the minimum of the encoded data included in the input data A subtraction value (a value of ML) obtained by subtracting the code length L is obtained.

  It is assumed that the next extraction (step S100) and storage (step S102) are further performed by the extraction unit 12 in a state where the storage data having the data length of “ML” is stored in the storage unit 14. The maximum value of the data length of the stored data in this state is “(ML) + M”.

  Therefore, the storage unit 14 needs to have at least a memory size of “(M−L) + M”, that is, “M × 2−L” represented by the above formula (1). For this reason, it is preferable that the memory size of the storage part 14 is more than the memory size shown by said Formula (1).

  In the present embodiment, a case will be described in which the memory size of the storage unit 14 is the memory size represented by the above equation (1).

  Next, decoding of encoded data by the decoding unit 16 will be described. In the present embodiment, the decoding unit 16 reads out the data from the beginning of the stored data stored in the storage unit 14 to extract and decode the encoded data (for example, N bytes).

  For this reason, when the decoding unit 16 performs decoding while the data length of the stored data stored in the storage unit 14 is “M × 2-L”, the data length of the stored data is “M × 2-L−”. N ".

  In the present embodiment, a case where the decoding unit 16 performs decoding using a lookup table will be described.

  Specifically, it is assumed that a plurality of pieces of encoded data included in the input data to be decoded are the encoded data shown in FIG. Further, it is assumed that storage data “01011000b” composed of remaining data “01b” and cut-out data “011000b” is stored in the storage unit 14.

  First, the decoding unit 16 generates a loop-up table for acquiring the data length of the encoded data to be decoded.

  FIG. 6 is a schematic diagram illustrating an example of the data structure of a lookup table that is being generated that is generated during decoding by the decoding unit 16. In FIG. 6, “X” represents an arbitrary value (0 or 1). The read data is data to be read by the decryption unit 16 from the stored data. The decoding unit 16 expands all combinations of “X” shown in FIG. 6 and creates a lookup table for acquiring encoded data and the cut-out length of the encoded data from the read data.

  FIG. 7 is a schematic diagram illustrating an example of a data structure of a lookup table generated by the decoding unit 16 at the time of decoding. As shown in FIG. 7, when delta_qp is “0”, the value of the first bit when represented by a binary number is “1”. For this reason, with respect to “6” which is the value of the first code length M, the remaining 5 bits may be any value. Therefore, as shown in FIG. 7, when the value of the read data read from the stored data is in the range of “32” to “63”, the decryption unit 16 has a value of delta_qp of “0”, and the extraction length Is analyzed to be “1”.

  FIG. 5 is a schematic diagram illustrating an example of stored data. Assume that the decoding unit 16 performs decoding from the stored data illustrated in FIG. 5 based on the lookup table illustrated in FIG. In this case, the data for 6 bits (first code length M) from the beginning in the stored data is “010110b”. For this reason, this stored data corresponds to “010XXX” in the lookup table shown in FIG. 6, so that it is understood that delta_qp is “−1” and the cut-out length is “3”.

  Therefore, the decoding unit 16 outputs the quantization parameter difference (delta_qp) = “− 1” as the decoded data from the stored data “01011000b”, and stores the data “010b” for 3 bits used for the decoding as the stored data. Cut out from. FIG. 8 is a schematic diagram illustrating an example of the remaining stored data stored in the storage unit 14. As shown in FIG. 8, in this case, the storage data stored in the storage unit 14 after decoding is 5-bit data “11000b”.

  Next, the determination unit 16A of the decoding unit 16 analyzes the stored data remaining in the storage unit 14 and determines whether variable length decoding is possible (see step S106 and step S108 in FIG. 2).

  Specifically, the determination unit 16A determines whether decoding is possible by reading and analyzing the storage data stored in the storage unit 14. As one of the analysis and determination methods, there is a method of actually decoding and determining whether or not encoded data can be obtained.

  In this case, when the decoding unit 16 uses the lookup tables of FIGS. 6 and 7 for the stored data stored in the storage unit 14, it is necessary to devise.

  The lookup tables shown in FIG. 6 and FIG. 7 are created based on the maximum code length of the encoded data that is variable-length encoded and included in the input data, as described above. For this reason, in the present embodiment in which the memory size of the storage unit 14 is minimized to a value satisfying the above equation (1), the data length of the stored data is a value less than the first code length M.

  For this reason, in this case, the decoding unit 16 needs to create a loop-up table in which the lower bits of the lookup table shown in FIGS. 6 and 7 are deleted.

  Specifically, it is assumed that the storage data shown in FIG. The data length of the stored data is 5 bits (“11000b”). For this reason, the decoding unit 16 deletes “6-5 = 1” bits from the lower bits of the lookup table shown in FIG. 7 and creates the lookup table shown in FIG.

  FIG. 9 is a schematic diagram illustrating an example of a lookup table from which lower bits are deleted.

  That is, when the stored data is “00001b”, the cut-out length is “6”. In this case, it is found that the data length of the encoded data is less than 1 bit. Therefore, in this case, the determination unit 16A determines that variable length decoding based on the stored data is impossible (step S108: No in FIG. 2).

  In the example shown in FIG. 8, since the stored data remaining in the storage unit 14 is “1XXXb”, it is found that delta_qp as decoded data is “0” and the cut-out length is “1”. . Therefore, in this case, the determination unit 16A determines that variable-length decoding based on the stored data is possible (step S108: Yes in FIG. 2).

  In the case where the determination of whether or not variable length decoding in step S108 in FIG. 2 is possible by performing decoding, if the determination in step S108 is affirmative (step S108: Yes), the processing in step S104 is omitted. Then, the process of step S106 may be executed. As a result, the decoding processing efficiency can be further improved.

  In the above description, the case has been described in which the decoding unit 16 performs the determination in step S108 by decoding the stored data. However, the decoding unit 16 may determine whether variable length decoding based on stored data is possible without performing decoding by using the following method.

  For example, the decoding unit 16 determines whether the upper 4 bits of the logical sum operation result of the stored data stored in the storage unit 14 is “0” or “1”, and thus based on the stored data. It may be determined whether variable length decoding is possible.

  For example, it is assumed that the storage data stored in the storage unit 14 is the storage data “11000b” illustrated in FIG. 8.

  In this case, the decoding unit 16 determines that variable-length decoding is not possible if the result of a logical sum operation (sometimes referred to as an OR operation) for the upper 4 bits of the stored data is “0”. This is because the stored data in which the logical sum operation result of the upper 4 bits of the stored data is “0” is “0000XXb”, that is, “000010b” or “000011b”. That is, in this case, 6 bits are required for decoding, but the data length of the stored data is 5 bits, and there is not enough data.

  On the other hand, the decoding unit 16 determines that variable length decoding is possible when the logical sum operation result of the upper 4 bits of the stored data is not “0”, that is, “1”.

  As described above, the decoding unit 16 determines whether or not variable length decoding is possible from the stored data by determining whether or not the logical sum operation result of the upper 4 bits of the stored data is “0”. Good. In this case, the processing can be simplified and the processing time can be shortened compared to the method of determining whether variable length decoding is possible while performing decoding.

  Next, a determination process (see step S110 in FIG. 2) in which the decoding unit 16 determines whether there is undecoded undecoded data in the input data to be decoded will be described.

  In step S110, the decoding unit 16 reads the data length of the undecoded data in the input data to be decoded. Note that this input data is the remaining data cut out by the cutout unit 12. For this reason, the decoding part 16 should just read the data length of undecoded data by reading the remaining data length of the input data of decoding object.

  Then, the decoding unit 16 determines whether there is undecoded data by determining whether the data length of the undecoded data is not “0”. Specifically, when the data length of the undecoded data is “0”, the decoding unit 16 makes a negative determination (step S110: No) in step S110. Then, the decoding process ends. On the other hand, when the data length of the undecoded data is not “0”, the decoding unit 16 makes an affirmative determination (step S110: Yes) in step S110, and returns to the extraction by the extraction unit 12 (step S100).

  As described above, in the decoding device 10 according to the present embodiment, input data including a plurality of pieces of variable length encoded data is cut out for each predetermined first code length M equal to or greater than the maximum code length, The encoded data included in the cut out data is subjected to variable length decoding.

  Here, conventionally, the input data is read from the head, the encoded data located at the head position (0th) is decoded, and the code length (data length) of the encoded data is analyzed, and according to the code length. The variable length decoding was performed on the encoded data. As described above, conventionally, so-called sequential processing is performed in which search is performed in order from the top of input data.

  On the other hand, as described above, decoding apparatus 10 according to the present embodiment cuts out input data including a plurality of pieces of variable-length encoded data for each predetermined first code length M that is greater than or equal to the maximum code length. Then, the encoded data included in the cut out data is variable-length decoded.

  FIG. 10 is an explanatory diagram of the processing timing of the decoding process for the input data to be decoded by the decoding device 10 according to the present embodiment.

  As shown in FIG. 10, in the decoding device 10 according to the present embodiment, the cutout unit 12 reads input data to be decoded sequentially from the top (see the arrow Y direction in FIG. 10), and a predetermined first code Data extraction is performed for each length M (see data extraction # 0, data extraction # 1, and data extraction # 2 in FIG. 10). For this reason, the decoding unit 16 performs the 0th data extraction (see # 0 in FIG. 10) in the extraction unit 12 without waiting for all the data to be extracted in the extraction unit 12. Decryption is started for the extracted data. That is, the decoding unit 16 performs decoding on the already cut out data while the cutting unit 12 performs the cutting. For this reason, in the decoding apparatus 10 of this Embodiment, at least one part of the data extraction by the extraction part 12 and the decoding by the decoding part 16 can be performed in parallel.

  As described above, in the decoding device 10 of the present embodiment, data can be extracted from the input data to be decoded at high speed and the decoding process can be executed.

  Therefore, the decoding apparatus 10 of the present embodiment can decode encoded data that has been subjected to variable length encoding at high speed.

  Further, as described above, in the decoding device 10 according to the present embodiment, the value of the first code length M, which is a unit in which the cutout unit 12 cuts out the cutout data from the input data, is encoded into a plurality of pieces of input data. It is set to a value greater than the maximum data code length. For this reason, in the decoding apparatus 10 according to the present embodiment, data longer than the data length necessary for decoding by the decoding unit 16 can be stored in the storage unit 14, so that the decoding process can be shortened.

  Moreover, in the decoding apparatus 10 of this Embodiment, after storing the cut-out data cut out by the cut-out unit 12 in the storage unit 14, the decoding unit 16 sequentially reads and decodes the data. For this reason, before the code length of the encoded data is determined by decoding, the extracted data can be extracted from the input data. Therefore, a plurality of encoded data included in the cut data cut out at different timings can be decoded in parallel or by high-speed decoding by pipeline processing.

  In the present embodiment, the decoding unit 16 (determination unit 16A) determines whether variable-length decoding is possible based on the stored data stored in the storage unit 14. When it is determined that decoding using the stored data is impossible, the cutout unit 12 cuts out the first code length cutout data from the input data and stores it in the storage unit 14. For this reason, the storage unit 14 can store only data having a minimum necessary data length used for decoding in the decoding unit 16. Thus, in the present embodiment, cut-out data having the first code length is cut out from input data only when data used for decoding is insufficient. For this reason, the memory size of the storage unit 14 can be minimized.

  In the present embodiment, the case has been described in which the encoded data included in the input data is encoded data of delta_qp (quantization parameter difference). However, the encoded data included in the input data is not limited to the encoded data of the quantization parameter difference. For example, the input data may include a plurality of types of encoding parameters having different maximum code lengths, such as quantization parameter differences (delta_qp) and motion vectors (mv_x, mv_y). When the input data includes a plurality of types of encoded data, the decoding apparatus 10 may perform the decoding process a number of times corresponding to the type of encoded data.

  Further, in the decoding device 10 of the present embodiment, the case has been described where analysis of stored data and determination as to whether variable length decoding is possible for the stored data are performed. However, the decoding apparatus 10 may omit the analysis of stored data and the determination of whether or not variable length decoding is possible in the decoding process shown in FIG. 2 (see step S106 and step S108 in FIG. 2).

  In this case, the decoding device 10 may perform the decoding process shown in FIG. FIG. 11 is a flowchart showing another procedure of the decoding process.

  First, the cutting unit 12 cuts input data for each first code length M (step S200) and stores it in the storage unit 14 (step S202). Next, the decryption unit 16 decrypts the stored data stored in the storage unit 14 (step S204). Next, the decoding unit 16 determines whether there is undecoded data (step S206). If an affirmative determination is made in step S206 (step S206: Yes), the process returns to step S200. On the other hand, if a negative determination is made in step S206 (step S206: No), the process proceeds to step S208. In step S208, the decryption unit 16 decrypts all stored data stored in the storage unit 14 (step S208). Then, this routine ends.

  Note that the processing of step S200, step S202, step S204, and step S206 is the same as that of step S100, step S102, step S104, and step S110 in FIG.

  As described above, the decoding device 10 executes the decoding process shown in FIG. 11 instead of the decoding process shown in FIG. 2, thereby eliminating the need to analyze the stored data and further simplifying the decoding process. Can do. However, when the decoding process shown in FIG. 11 is executed, the stored data remaining in the storage unit 14 may be larger than when the decoding process shown in FIG. 2 is executed.

  Specifically, when the decoding process shown in FIG. 11 is executed, the data amount of the stored data stored in the storage unit 14 is much larger than “M × 2-L” described above. For this reason, when the decoding process shown in FIG. 11 is executed, it is necessary to prepare the storage unit 14 having a larger memory size as the data size of the input data is larger.

  For this reason, when the decoding device 10 executes the decoding process shown in FIG. 11, the device capable of mounting a memory having a larger memory size as the storage unit 14 than the case of executing the decoding process shown in FIG. It is preferable to apply the decoding device 10.

  If the decoding process shown in FIG. 11 is executed instead of the decoding process shown in FIG. 2, the process from step S200 to step S206 is executed, and if a negative determination is made in step S206 (step S206: No), a step is performed. In S208, it is necessary to decrypt all stored data stored in the storage unit 14. Since the decoding in step S208 is the same as the decoding process in step S204, the processing amount of the entire decoding process does not increase compared to the decoding process shown in FIG.

(Embodiment 2)
In this embodiment, in addition to the decoding process based on the storage data storing the cut data cut out from the input data in the first embodiment, a form for performing the decoding process on the encoded data included in the input data will be described. .

  FIG. 12 is a block diagram illustrating a functional configuration of the decoding device 10A according to the present embodiment. The decoding apparatus 10A accepts input data including a plurality of encoded data that has been subjected to variable length encoding.

  In decoding apparatus 10A of the present embodiment, a case will be described in which input data includes a plurality of types of encoded data having different maximum code lengths.

  The decoding device 10A includes a cutout unit 12, a storage unit 14, a first decoding unit 17A, and a second decoding unit 17B. The first decoding unit 17A includes a determination unit 170A. The cutout unit 12 and the storage unit 14 are the same as those in the first embodiment. Further, the first decoding unit 17A and the determination unit 170A are the same as the decoding unit 16 and the determination unit 16A of the first embodiment, respectively.

  The second decoding unit 17B sequentially extracts and decodes types of encoded data having a maximum code length equal to or greater than a predetermined value among a plurality of types of encoded data included in the input data (hereinafter referred to as second decoding). Sometimes). On the other hand, the first decoding unit 17A sequentially extracts and decodes encoded data of a type other than the types extracted by the second decoding unit 17B from among a plurality of types of encoded data included in the input data (hereinafter referred to as the first data). May be referred to as 1 decoding).

  This predetermined value may be set in advance according to the encoding method of the encoded data included in the input data handled by the decoding apparatus 10A.

  In the decoding apparatus 10A of the present embodiment, a case will be described in which input data includes two types of encoded data having different maximum code lengths. For this reason, the first decoding unit 17A cuts out and decodes the encoded data of the type having the smallest maximum code length (hereinafter sometimes referred to as the first type) out of the two types of encoded data. On the other hand, the second decoding unit 17B cuts out and decodes the encoded data of the type having the largest maximum code length (hereinafter sometimes referred to as the second type) out of the two types of encoded data.

  The first type of encoded data having a small maximum code length is, for example, encoded data of a difference between quantization parameters. The second type of encoded data having a large maximum code length is, for example, encoded data of orthogonal transform coefficients used in MPEG.

  Also, in the input data used in the present embodiment, a plurality of first type encoded data having a small maximum code length and a second type encoded data having a maximum maximum code length are in this order in the head of the input data. A description will be given assuming the case of arrangement from

  FIG. 13 is a flowchart showing the procedure of the decoding process executed by the decoding device 10A.

  First, the cutout unit 12 cuts out a group of first type encoded data having a small maximum code length in the received input data for each first code length (step S300). Then, the cutting unit 12 stores the cut out data in the storage unit 14 (step S302). The processes in step S300 and step S302 are the same as the processes in step S100 and step S102 of the first embodiment. The definition of the first code length is also the same as the first code length (first code length M) in the first embodiment.

  Next, the second decoding unit 17B directly reads the second type of encoded data having a large maximum code length from the input data, and performs second decoding for decoding the encoded data (step S304).

  Next, the first decoding unit 17A performs step S306 on the first type of encoded data in the stored data stored in the storage unit 14 in the same manner as in steps S104 to S110 performed by the decoding unit 16 of the first embodiment. Step S312 is executed.

  In this embodiment, when an affirmative determination is made in step S310 corresponding to the determination in step S108 in the first embodiment (step S310: Yes), the process returns to step S304. On the other hand, if a negative determination is made in step S310 (step S310: No), the process proceeds to step S312. If an affirmative determination is made in step S312 (step S312: Yes), the process returns to step S300. On the other hand, if a negative determination is made in step S312 (step S312: No), this routine ends.

  As described above, in the present embodiment, the first type of encoded data having a small maximum code length is first extracted after being cut out for each first code length M by the cutout unit 12 and stored in the storage unit 14. The decoding unit 17A sequentially performs decoding (first decoding). On the other hand, for the second type of encoded data having a large maximum code length, the second decoding unit 17B directly extracts and decodes (second decoding) from the input data without storing in the storage unit.

  Therefore, in the decoding device 10A of the present embodiment, in addition to the effects of the first embodiment, the decoding processing speed can be further improved.

  For example, when the encoded data included in the input data to be decoded is the second type of encoded data having a large maximum code length, such as encoded data of orthogonal transform coefficients used in MPEG, the storage unit 14 Therefore, it is necessary to prepare the storage unit 14 having a larger memory size than that of the first embodiment.

  On the other hand, in the decoding apparatus 10A according to the present embodiment, the second type of encoded data having a large maximum code length is not stored in the storage unit 14, but is directly extracted from the input data by the second decoding unit 17B and decoded ( Second decoding). For this reason, in addition to the effects of the first embodiment, the memory size of the storage unit 14 can be further reduced.

  Note that FIG. 13 illustrates a case where the decoding apparatus 10A executes the first decoding by the first decoding unit 17A after the second decoding by the second decoding unit 17B. However, the second decoding by the second decoding unit 17B may be executed after the first decoding by the first decoding unit 17A.

  In this case, the decoding device 10A may execute the decoding process shown in FIG.

  FIG. 14 is a flowchart showing another procedure of the decoding process executed by the decoding device 10A.

  First, similarly to FIG. 13, the cutout unit 12 cuts out a group of first type encoded data having a small maximum code length in the received input data for each first code length (step S300). Then, the cutting unit 12 stores the cut out data in the storage unit 14 (step S302).

  Next, the first decoding unit 17A performs the first decoding in the same manner as in step S306 (step S3040). Next, as in step S304, the second decoding unit 17B directly reads the second type of encoded data having a large maximum code length from the input data and performs second decoding for decoding the encoded data ( Step S3060).

  Next, processing similar to that in steps S308 to S312 is performed, and this routine is terminated.

  Thus, the order of the first decoding by the first decoding unit 17A and the second decoding by the second decoding unit 17B may be reversed.

  Note that, in the present embodiment, a mode has been described in which decoding is performed by one of the first decoding unit 17A and the second decoding unit 17B, and then decoding by the other is performed. However, the first decoding by the first decoding unit 17A and the second decoding by the second decoding unit 17B may be executed in parallel.

  In this case, the decoding device 10A may execute the processes shown in FIGS.

  FIG.15 and FIG.16 is a flowchart which shows the other procedure of the decoding process performed with 10 A of decoding apparatuses of this Embodiment.

  First, the cutout unit 12 cuts out a group of first type encoded data having a small maximum code length in the received input data for each first code length (step S400). Then, the cutting unit 12 stores the cut out data in the storage unit 14 (step S402). Next, the second decoding unit 17B directly reads the second type of encoded data having a large maximum code length from the input data, and performs second decoding for decoding the encoded data (step S404).

  The processes in steps S400 to S404 are the same as the processes in steps S300 to S304.

  Next, in the same way as steps S106 to S110 by the decoding unit 16 of the first embodiment, the first decoding unit 17A analyzes the stored data, determines whether variable length decoding is possible, and whether there is undecoded data. It is determined whether or not (steps S406 to S410). If an affirmative determination is made in step S410 (step S410: Yes), the process returns to step S400. On the other hand, if a negative determination is made in step S410 (step S410: No), this routine ends.

  In the decoding device 10A, the first decoding unit 17A executes the parallel processing shown in FIG. 16 in parallel while executing the decoding processing of FIG.

  FIG. 16 is a flowchart showing a procedure of parallel processing executed by the first decoding unit 17A.

  First, the first decryption unit 17A determines whether the storage data has been stored in the storage unit 14 (step S500). If a negative determination is made in step S500 (step S500: No), this routine ends. On the other hand, if a positive determination is made in step S500 (step S500: Yes), the process proceeds to step S502. In step S502, the first decoding unit 17A determines whether variable-length decoding is possible in the same manner as in step S408 (step S502).

  If a negative determination is made in step S502 (step S502: No), the process returns to step S500. On the other hand, if an affirmative determination is made in step S502 (step S502: Yes), the process proceeds to step S504.

  In step S504, the first decoding unit 17A performs the first decoding similarly to step S306 (step S504). Then, this routine ends.

  When the decoding device 10A performs the parallel processing shown in FIG. 16 during the decoding processing shown in FIG. 15, in the decoding device 10A, the first decoding by the first decoding unit 17A and the second decoding by the second decoding unit 17B are performed. Decoding and can be performed in parallel.

  For this reason, in addition to the effects of the first embodiment, the processing time can be further shortened.

  As described above, according to decoding apparatus 10A of the present embodiment, even when input data includes a plurality of types of encoded data having different maximum code lengths, the input data is decoded at high speed. Can be executed.

  Here, the input data includes a plurality of types of encoded data having greatly different maximum code lengths, and in particular, the first type of encoded data having a small maximum code length (eg, #a) of the encoded data. And a second type of encoded data having a large maximum code length (eg, #b) of the encoded data. In such a case, the decoding processing time may become long.

  On the other hand, in the present embodiment, even when a plurality of types of encoded data having different maximum code lengths are mixed, occurrence of waiting time for decoding processing can be suppressed.

  FIG. 17 is an explanatory diagram of processing timing of decoding processing on input data to be decoded by the decoding device 10A according to the present embodiment.

  As shown in FIG. 17, in the decoding device 10A of the present embodiment, the cutout unit 12 reads the received input data in order from the top (see the arrow Y direction in FIG. 17), and the first with the smallest maximum code length. A group of types of encoded data is extracted for each first code length M (data extraction # a0 in FIG. 17). The cut out data is stored in the storage unit 14.

  Then, when the first decoding unit 17A executes the first decoding for the extracted data (see decoding process # a0), the second decoding unit 17B reads directly from the input data in parallel. The second type of encoded data having a large maximum code length is second-decoded (see decoding process # b0).

  For this reason, according to the decoding apparatus 10A of the present embodiment, a plurality of decoding processes can be executed in parallel on the same input data.

  Therefore, in addition to the effects of the first embodiment, the decoding process can be executed at a higher speed.

(Embodiment 3)
In this embodiment, an encoding apparatus that creates input data to be decoded in Embodiments 1 and 2 will be described.

  FIG. 18 is a block diagram showing a functional configuration of encoding apparatus 50 according to the present embodiment.

  The encoding device 50 according to the present embodiment generates encoded data by encoding information to be decoded using a variable-length encoding scheme. The encoding device 50 outputs output data including the generated encoded data. In the decoding device 10 and the decoding device 10A described above, the output data including the encoded data created by the encoding device 50 is received as input data to be decoded.

  The encoding device 50 is incorporated in a device implemented by hardware such as an image encoding LSI (Large Scale Integration). By using the encoded data obtained by encoding by the encoding device 50, for example, display responsiveness to a display can be improved.

  The encoding device 50 includes a first encoding unit 52, a second encoding unit 54, and a combining unit 56. The first encoding unit 52 and the second encoding unit 54 each encode the encoding target information using different variable length encoding methods (details will be described later). The combining unit 56 combines the encoded data generated by the first encoding unit 52, packetizes it with a predetermined data length, and sequentially outputs it (details will be described later).

  FIG. 19 is a schematic diagram illustrating an example of a data structure of output data output from the encoding device 50.

  As illustrated in FIG. 19A, the output data includes encoded data encoded by the first encoding unit 52 and the second encoding unit 54. The encoding device 50 packetizes these encoded data into a predetermined data length (details will be described later, for example, every M bytes) (see FIG. 19B), and sequentially outputs them (packet # in FIG. 19). 0 to packet # 1, etc.). The packet corresponds to combined data.

  FIG. 20 is a flowchart showing the procedure of the encoding process executed by the encoding device 50.

  It is assumed that the encoding target data to be encoded by the encoding device 50 includes a plurality of types of encoding target information having different maximum code lengths by encoding. In the present embodiment, description will be made assuming that two types of encoding target information are included. In the present embodiment, these two types of encoding target information will be referred to as first data and second data.

  The first data is data having a relatively small maximum code length when variable length coding is performed. The first data is, for example, a quantization parameter difference (delta_qp) in MPEG. Thus, the first data is data having a relatively small dynamic range (for example, −51 to +51 in H.264).

  The second data is data having a relatively large maximum code length when variable length coding is performed. The second data has a feature that a range of values that the second data can take is large or a plurality of pieces of information are included. The second data is, for example, an orthogonal transform coefficient in MPEG. Thus, the second data is data having a relatively large dynamic range of values (for example, -32768 to 32767 in H.264). In addition, when encoding the second data, when encoding processing is performed in units of 16 × 16 pixel blocks, 256 orthogonal transform coefficients are encoded.

  The encoded data obtained by encoding the first data corresponds to the above-described first type encoded data having a small maximum code length. The encoded data obtained by encoding the second data corresponds to the above-described second type of encoded data having a large maximum code length.

  First, the first encoding unit 52 performs variable length encoding on the first data included in the encoding target data (step S600). The first encoding unit 52 outputs the encoded data to the combining unit 56.

  The combining unit 56 combines the encoded data encoded by the first encoding unit 52 (step S602).

  Next, the combining unit 56 determines whether or not the encoded data for one packet has been combined (step S604). The combining unit 56 determines whether or not encoded data having a predetermined data length (details will be described later) is combined, thereby determining step S604.

  If a negative determination is made in step S604 (step S604: No), the process returns to step S600. On the other hand, if a positive determination is made in step S604 (step S604: Yes), the process proceeds to step S606.

  In step S606, the second encoding unit 54 performs variable length encoding on the second data (step S606).

  Next, the first encoding unit 52 determines whether or not unencoded unencoded data remains in the encoding target data (step S608). If an affirmative determination is made in step S608 (step S608: Yes), the process returns to step S600. On the other hand, if a negative determination is made in step S608 (step S608: No), this routine ends.

  Next, the encoding process executed by the encoding device 50 will be described in detail.

  First, a process in which the first encoding unit 52 performs variable length encoding on the first data included in the encoding target data will be described. This process corresponds to step S600.

  The first encoding unit 52 performs an encoding process using a variable length encoding method. The variable length coding method may be any method. For example, when the first data is a quantization parameter difference (delta_qp), a coding scheme using the Golomb code shown in FIG. 3 may be used.

  FIG. 21 is an explanatory diagram of variable length encoding by the first encoding unit 52 and combination (packetization) by the combining unit 56.

  Specifically, the first encoding unit 52 performs variable length encoding on the quantization parameter difference (delta_qp) illustrated in FIG. 3 in the order of “−1”, “0”, “+3”, and “+1”. Suppose that Then, the encoded data corresponding to the difference between the quantization parameters is “010b”, “1b”, “00011b”, and “011b”, respectively, as shown in FIG.

  The first encoding unit 52 sequentially outputs the encoded data obtained by encoding to the combining unit 56. The combining unit 56 combines the encoded data received from the first encoding unit 52 and generates a packet having a predetermined data length (corresponding to the processing in steps S602 to S604 in FIG. 20).

  The predetermined data length used at the time of packet generation is set to a value equal to or greater than the maximum code length when the first data and the second data are variable-length encoded. This is because at least one first data can be obtained when one packet is decoded. That is, this data length is set to the same value as the first code length M used in the first and second embodiments.

  As shown in FIG. 3, in the present embodiment, the maximum code length of delta_qp, which is a difference between quantization parameters as the first data, is 6. For this reason, in this embodiment, the combining unit 56 generates a packet having a data length of 6 bits.

  In the step of determining whether or not the encoded data for one packet is combined in step S604, if the combining unit 56 makes an affirmative determination (step S604: Yes), the combining unit 56 combines the packets (combined data). Is output.

  On the other hand, when the encoded data for one packet is not combined, the combining unit 56 makes a negative determination in step S604 (step S604: No), and returns to step S600.

  For example, in a stage where the first encoding unit 52 encodes the quantization parameter difference “delta_qp” = “− 1” among the plurality of first data included in the encoded data to be encoded, the combined data The amount of data is 3 bits. Therefore, in this case, a negative determination is made in step S604. Then, when 1-bit encoded data is created by the encoding of the quantization parameter difference “delta_qp” = “0” by the first encoding unit 52, the combined data becomes 4 bits. Even at this stage, since it is less than 6 bits, a negative determination is made in step S604.

  When 5-bit encoded data is generated by encoding the quantization parameter difference “delta_qp” = “+ 3” by the first encoding unit 52, the combined data becomes 9 bits. At this stage, the coupling unit 56 makes an affirmative determination in the determination in step S604 (step S604: Yes).

  FIG. 22 is a schematic diagram illustrating an example of output data output from the encoding device 50.

  By outputting encoded data of one packet (6 bits in this embodiment) from the combining unit 56, the quantization parameter difference “delta_qp” is included in the first packet (see packet # 0 in FIG. 22). ”=“ − 1 ”(encoded data A # 0 in FIG. 22) and the quantization parameter difference“ delta_qp ”=“ 0 ”(encoded data A # in FIG. 22). 1) and a part of the encoded data of the quantization parameter difference “delta_qp” = “+ 3” (see encoded data A # 2a in FIG. 22) are output. The

  Thereafter, combining unit 56 sequentially generates and outputs packets (see packet # 1 in FIG. 22).

  Returning to FIG. 20, next, the second encoding unit 54 performs variable length encoding on the second data (see step S606 in FIG. 20).

  The second encoding unit 54 performs variable length encoding on the second data using a variable length encoding method. Any method may be used for the first encoding unit 52 to perform variable length encoding.

  For example, it is assumed that the second data is an orthogonal transform coefficient. In this case, the second encoding unit 54 performs variable length encoding using, for example, a Huffman code used in MPEG-2.

  The second encoding unit 54 outputs a packet (combined data) of encoded data obtained by variable-length encoding the second data. In the encoding device 50, the packet unit output from the first encoding unit 52 is output. The encoded data and the encoded data output from the second encoding unit 54 are alternately output. For this reason, the output data output from the encoding device 50 has the data structure shown in FIG.

  That is, as shown in FIG. 22, the output data output from the encoding device 50 is a combination of encoded data obtained by variable-length encoding the first data for one packet and variable-length encoding of the second data. The encoded data is alternately arranged.

  Returning to FIG. 20, when an affirmative determination is made in the processing of step S608 in FIG. 20 (step S608: Yes), the processing returns to step S600. On the other hand, if a negative determination is made in step S608 (step S608: No), this routine ends.

  As described above, the encoding apparatus 50 according to the present embodiment packetizes the variable length encoded encoded data for each fixed length for the first data having a small maximum code length when the variable length encoding is performed. To do.

  Therefore, the encoding apparatus 50 according to the present embodiment can generate output data that can be decoded by cutting out the encoded data at high speed during the decoding process.

  In the present embodiment, as described above, for the first data having a small maximum code length when variable length coding is performed, packetization is performed to perform coding for the data length M. Therefore, the first data is repeatedly encoded before executing the encoding of the second data having a large maximum code length when the variable length encoding is performed.

  Therefore, if the number of first data and the number of second data included in the encoding target data to be encoded by the encoding device 50 are the same number, the first data is encoded first. In some cases, the second data may remain unencoded. For this reason, in the encoding process shown in FIG. 20, when the encoding process of the first data is completed, the encoding process of the first data (that is, steps S600 to S1 by the first encoding unit 52 and the combining unit 56). It is preferable to omit the process of S604.

  Further, in determining whether there is unencoded data in step S608, it may be further determined whether each of the first data and the second data remains in the unencoded data. Then, when it is determined that both the first data and the second data remain in the unencoded data, the process may proceed to step S606 by the second encoding unit 54. If only the first data remains in the unencoded data, the process returns to step S600.

  Alternatively, the second encoding unit 54 may output the encoded data obtained by variable-length encoding the second data by the amount corresponding to the variable-length encoded first data.

  FIG. 23 is a schematic diagram illustrating another example of output data output from the encoding device 50.

  That is, the first packet (packet # 0 in FIG. 23) includes the 0th encoded data of the first data (encoded data A # 0 in FIG. 23) and the first encoded of the first data. Data (encoded data A # 1 in FIG. 23). For this reason, the second encoding unit 54 encodes the second data up to the first.

  On the decoding apparatus (10, 10A) side that receives and decodes the output data having the data structure shown in FIG. 22 as input data, the encoded data of the first first data (encoded data A # 1 in FIG. 22) and When decoding the encoded data of the first second data (encoded data B # 1 in FIG. 22), it is necessary to hold up to the third data for the first data. For this reason, it is necessary to prepare a storage unit 14 having a larger memory size.

  On the other hand, if the output data has the data structure shown in FIG. 23, on the decoding device (10, 10A) side that receives and decodes the output data as input data, The data up to the middle of the data (see encoded data A # 2a in FIG. 23) may be held. For this reason, compared with the case where the output data has the data structure shown in FIG. 22, the output data has the data structure shown in FIG. 23, so that the memory in the decoding device (10, 10A) side (storage unit 14) memory. The size can be reduced.

  Note that a program for executing each of the decoding process and the encoding process executed by the decoding device 10, the decoding device 10A, and the encoding device 50 in the above-described embodiment is provided by being incorporated in advance in a ROM or the like. .

  The program for executing each of the decoding process and the encoding process executed by the decoding device 10, the decoding device 10A, and the encoding device 50 in the above embodiment is a file in an installable format or an executable format. You may comprise so that it may record and provide on computer-readable recording media, such as CD-ROM, flexible disk (FD), CD-R, and DVD (Digital Versatile Disk).

  Furthermore, a program for executing each of the decoding process and the encoding process executed by the decoding apparatus 10, the decoding apparatus 10A, and the encoding apparatus 50 in the above embodiment is executed on a computer connected to a network such as the Internet. And may be provided by being downloaded via a network. In addition, a program for executing each of the decoding process and the encoding process executed by the decoding device 10, the decoding device 10A, and the encoding device 50 in the above embodiment is provided or distributed via a network such as the Internet. You may comprise as follows.

  The program for executing each of the decoding process and the encoding process executed by the decoding device 10, the decoding device 10A, and the encoding device 50 in the above embodiment has a module configuration including the above-described units, As actual hardware, a CPU (processor) reads out a program from the ROM and executes the program, so that each unit is loaded on the main storage device, and each unit is generated on the main storage device. In addition, you may comprise each part of the decoding apparatus 10, decoding apparatus 10A, and the encoding apparatus 50 in the said embodiment with hardware, such as a circuit.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

10, 10A Decoding device 12 Extraction unit 14 Storage unit 16 Decoding unit 16A Determination unit 17A First decoding unit 17B Second decoding unit

Claims (4)

Input data including a plurality of variable length encoded data is sequentially cut out for each predetermined first code length that is equal to or larger than the maximum code length of the plurality of encoded data, and the cut out data is used as storage data. A cut-out section to be stored in the storage section;
A decoding unit that sequentially reads the stored data stored in the storage unit, and performs variable-length decoding on the encoded data included in the read storage data;
With
The decoding unit includes a determination unit that determines whether variable length decoding based on the stored data is possible,
The cutout unit cuts out the cutout data from the input data and stores it as the storage data in the storage unit when the determination unit determines that variable length decoding is impossible.
Decoding device. The storage unit
A storage capacity equal to or greater than a subtraction value obtained by subtracting a minimum code length of the plurality of encoded data from a double value of the first code length;
The decoding device according to claim 1. The input data includes a plurality of types of encoded data having different maximum code lengths,
The decoding unit executes in parallel at least one type of variable length decoding of the plurality of types of encoded data and another type of variable length decoding other than the type,
The decoding device according to claim 1. Input data including a plurality of variable length encoded data is sequentially cut out for each predetermined first code length that is equal to or larger than the maximum code length of the plurality of encoded data, and the cut out data is used as storage data. A cutting step to store in the storage unit;
A decoding step of variable-length decoding the encoded data included in the stored data stored in the storage unit;
With
The decoding step includes a determination step of determining whether variable length decoding based on the stored data is possible,
In the cutting step, when it is determined by the determining step that variable length decoding is impossible, the cutting data is cut out from the input data and stored as the storage data in the storage unit.
Decryption method.

Images