Multidimensional data stream processing method, device, equipment and storage medium

Abstract

The embodiment of the present invention discloses a multidimensional data stream processing method, device, equipment and storage medium. The method comprises: obtaining an N-dimensional data stream, wherein N is a positive integer greater than or equal to 2; sequentially writing the signal data in the N-dimensional data stream into an N-dimensional storage space, and in the process of writing the signal data, looping back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loopback cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1]. This enables real-time caching of a large number of multidimensional data streams without consuming a large amount of hardware resources and ensuring signal quality.

Description

Technical Field

[0001] The embodiments of the present invention relate to signal processing technology, and in particular to a multi-dimensional data stream processing method, device, equipment and storage medium. Background Art

[0002] With the advent of the information age and the digital world, real-time signal processing has become an extremely important discipline and technical field. The processing of large amounts of data streams in real-time signals, especially the processing of multi-dimensional data streams in real-time signals, is becoming increasingly important.

[0003] In the system design of real-time signal processing, FIFO memory is widely used for the purpose of increasing data transmission rate, processing large data streams, and matching systems with different transmission rates, thereby improving system performance. FIFO memory is a first-in-first-out double-port buffer, that is, the first data entering it is the first to be removed, one of which is the input port of the memory and the other is the output port of the memory.

[0004] Currently, when storing multidimensional data streams, for example, when storing three-dimensional data streams, the three-dimensional data streams are sliced ​​along the Z axis, and the three-dimensional data streams are reduced in dimension until they are reduced to two dimensions, and then stored in a two-dimensional form.

[0005] In the above technical solution, since the three-dimensional data stream has been reduced to two dimensions, when the data volume is large and the storage capacity of the FIFO memory cannot completely store all the three-dimensional data streams, the function body can only capture part of the data of the three-dimensional data stream, that is, it may only capture the neighborhood data on the same two-dimensional plane, and cannot capture the neighborhood data between different two-dimensional planes. Therefore, the correlation between the data in the third dimension and the first two dimensions cannot be fully utilized, thereby sacrificing a large part of the quality of signal processing. Summary of the invention

[0006] The embodiments of the present invention provide a multi-dimensional data stream processing method, apparatus, device and storage medium to achieve real-time caching of a large number of multi-dimensional data streams without consuming a large amount of hardware resources and ensuring signal quality.

[0007] In a first aspect, an embodiment of the present invention provides a multidimensional data stream processing method, the method comprising:

[0008] Obtain an N-dimensional data stream, where N is a positive integer greater than or equal to 2;

[0009] The signal data in the N-dimensional data stream are sequentially written into the N-dimensional storage space, and during the writing process of the signal data, each M-dimensional data set that has been written into the N-dimensional storage space is looped back based on the corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0010] In a second aspect, an embodiment of the present invention further provides a multi-dimensional data stream processing device, the device comprising:

[0011] An N-dimensional data stream acquisition module, used to acquire an N-dimensional data stream, where N is a positive integer greater than or equal to 2;

[0012] A data writing module is used to write the signal data in the N-dimensional data stream into the N-dimensional storage space in sequence, and during the writing process of the signal data, loop back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0013] In a third aspect, an embodiment of the present invention further provides a device, the device comprising:

[0014] one or more processors;

[0015] A storage device for storing one or more programs;

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement any multi-dimensional data stream processing method described in any one of the embodiments of the present invention.

[0017] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute any of the multi-dimensional data stream processing methods described in the embodiments of the present invention.

[0018] The technical solution of the embodiment of the present invention improves the architecture of the N-dimensional storage space, pre-sets the loopback period, writes the signal data in the acquired N-dimensional data stream into the N-dimensional storage space in sequence, and in the process of writing the signal data, loops back each M-dimensional data set that has been written into the N-dimensional storage space based on the corresponding loopback period until the writing of the N-dimensional data stream is completed, thereby improving the shortcomings of the traditional N-dimensional storage space in processing the N-dimensional data stream, and achieving the effect of real-time caching of a large number of multi-dimensional data streams without consuming a large amount of hardware resources and ensuring signal quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flowchart of a multi-dimensional data stream processing method in Embodiment 1 of the present invention;

[0020] Figure 2 A schematic diagram of the corresponding relationship between an N-dimensional data stream and an M-dimensional data set in the first embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the logic block diagram of writing a one-dimensional data stream into a one-dimensional FIFO in the first embodiment of the present invention in five different clock cycles;

[0022] Figure 4 This is a schematic diagram of two-dimensional slicing of a frame of multi-dimensional data stream along the Z axis in the first embodiment of the present invention;

[0023] Figure 5 This is a flowchart of a multi-dimensional data stream processing method in Embodiment 2 of the present invention;

[0024] Figure 6 This is a schematic diagram of determining the associated N-dimensional data stream block in the second embodiment of the present invention;

[0025] Figure 7 This is a flowchart of a multi-dimensional data stream processing method in Embodiment 3 of the present invention;

[0026] Figure 8 A schematic diagram of position adjustment of a storage space in an embodiment of the present invention in Embodiment 3 of the present invention;

[0027] Fig. 9 A schematic diagram of position adjustment of storage space in the prior art in Embodiment 3 of the present invention;

[0028] Fig.10 is a schematic diagram of a one-dimensional data set in Embodiment 3 of the present invention;

[0029] Fig.11 A schematic diagram of determining that a one-dimensional data set in the third embodiment of the present invention is a preset number of signal data;

[0030] Fig.12 It is a schematic diagram of the solution of the embodiment of the present invention and the prior art solution after storing an N-dimensional data stream in the third embodiment of the present invention;

[0031] Fig.13 is a schematic diagram of the structure of a multi-dimensional data stream processing device in Embodiment 4 of the present invention;

[0032] Fig.14 It is a structural diagram of a device in Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0033] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0034] Embodiment 1

[0035] Figure 1 This is a flow chart of a multidimensional data processing method provided in Embodiment 1 of the present invention. This embodiment is applicable to the case where a large amount of multidimensional data streams are cached in real time without consuming a large amount of hardware resources and ensuring signal quality. The method can be executed by a multidimensional data processing device, which can be implemented by software and / or hardware. The multidimensional data processing device can be configured on a computing device, and specifically includes the following steps:

[0036] S110. Obtain an N-dimensional data stream, where N is a positive integer greater than or equal to 2.

[0037] Exemplarily, the N-dimensional data stream may be, but is not limited to, an N-dimensional image data stream, an N-dimensional video data stream, an N-dimensional audio data stream, or an N-dimensional electric energy data stream.

[0038] Optionally, the acquisition of N-dimensional data stream may specifically be collecting signal data based on a signal collector array arranged in N-1 dimensions; combining the signal data collected by each signal collector based on the arrangement of the signal collector array and the collection timestamp to form an N-dimensional data stream.

[0039] Exemplarily, the acquisition timestamp may be a time mark when the signal acquisition array acquires signal data.

[0040] Taking the N-dimensional data stream as a three-dimensional signal data stream as an example, the signal here can be a solar signal, a wind signal, etc. The two-dimensional solar signal collected by the solar signal collector array located on the solar panel is combined with the collection time of the solar signal according to the arrangement of the solar signal collector array to form a three-dimensional solar signal data stream.

[0041] S120. Write the signal data in the N-dimensional data stream into the N-dimensional storage space in sequence, and in the process of writing the signal data, loop back each M-dimensional data set that has been written into the N-dimensional storage space based on the corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0042] Exemplarily, the N-dimensional storage space may be an N-dimensional memory that can be used to store data, for example, may be an N-dimensional first-in first-out queue (FIFO) space.

[0043] The M-dimensional data set can be the signal data of each dimension that has been written into the N-dimensional storage space during the signal data writing process. Here, each M-dimensional data set includes the M-dimensional data set corresponding to each integer value in the interval [1, N-1]. For example, when N is 1, M is 1; when N is 2, M is 1 and 2; when N is 3, M is 1, 2 and 3.

[0044] refer to Figure 2 A schematic diagram of the correspondence between the N-dimensional data stream and the M-dimensional data set, Figure 2 Take 8*8*8 3D image data stream as an example. Figure 2 A total of 12 transient diagrams of this three-dimensional storage space (such as a three-dimensional FIFO) are listed. The clock cycle corresponding to each state = (cycle factor X) * (cycle factor Y) - 1. According to the above calculation formula, it is not difficult to find that: Figure 2 The 4th and 5th pictures are at the same time, and the 8th and 9th pictures are at the same time, but they represent different meanings. The detailed discussion is as follows:

[0045] (1) Figure 2 The process of the 1st to 4th images in the 8*8*8 three-dimensional image data stream is that the 0th, 2nd, 3rd, and 7th pixels are pushed into the FIFO_00 of the first dimension in the 0th, 2nd, 3rd, and 7th clock cycles respectively. Among them, Figure 2 The 7th pixel in the 4th picture is pushed into FIFO_00 of the 1st dimension in the 7th clock cycle, marking the completion of the first data loading of the 1st dimension.

[0046] (2) Figure 2 The fifth figure in FIG. 1 represents treating the first eight pixels described in (1) as a whole, that is, the first "one-dimensional data set" starts to load data into the second dimension of this three-dimensional FIFO. Figure 2 The process of the 5th to 8th images in the figure is that the 0th, 1st, 2nd, and 7th "one-dimensional data sets" composed of the 0th to 7th, 8th to 15th, 16th to 23rd, and 24th to 63rd pixels are pressed into the FIFO_OX (FIFO_00~07) of the second dimension in the 7th, 15th, 23rd, and 63rd clock cycles respectively. Among them, Figure 2 The 7th "one-dimensional data set" in the 8th figure is pushed into FIFO_0X (FIFO_00~07) in the 63rd cycle, marking the completion of the first data loading of the second dimension.

[0047] (3) Figure 2 The ninth picture in the figure represents treating the first eight "one-dimensional data sets" (64 pixels in total) described in (2) as a whole, that is, the first "two-dimensional data set" starts to load data into the third dimension of this three-dimensional FIFO. Figure 2 The process of the 9th to 12th figures is that the 0th, 1st, 2nd, and 7th "two-dimensional data sets" composed of the 0th to 7th, 8th to 15th, 16th to 23rd, and 24th to 63rd "one-dimensional data sets" are pushed into the FIFO_XX (FIFO_0X~7X) of the third dimension in the 55th, 127th, 199th, and 271st clock cycles respectively. Among them, Figure 2 The 7th "two-dimensional data set" in the 12th figure is pushed into FIFO_XX (FIFO_0X~7X) in the 271st cycle, marking the completion of the first data loading of the third dimension.

[0048] From the above, we can see that Figure 2 Each FIFO_X in is called a "one-dimensional data set", 7 "one-dimensional data sets" are collectively called a two-dimensional data set, and 7 "two-dimensional data sets" are collectively called a "three-dimensional data set". When all 7 "two-dimensional data sets" are written, it means that the writing of the 8*8*8 three-dimensional image data stream is completed.

[0049] The traditional way to write signal data in an N-dimensional data stream into an N-dimensional storage space (N is FIFO) is as follows:

[0050] Let’s take a one-dimensional data stream as an example. Figure 3 Schematic diagram of the logic block diagram for writing a one-dimensional data stream into a one-dimensional FIFO in five different clock cycles. Figure 3 In the first figure, the data stream that is about to be input into the FIFO buffer enters the FIFO through the data_in port. After the data stream enters the FIFO and is processed by the pipeline quasi-parallel signal, it will be output from data_out in the same code stream format. This FIFO can cache 3 lines of data in a frame at the same time. This FIFO has four tap outputs. Each tap is cached for 4 beats, which are AX0, AX1, AX2, and AX3 (where: X = 0, 1, 2, 3). The read enable start time of each tap is 0, 3839, 7679, and 11519 respectively. After the read enable is turned on, the data stream in the FIFO continues to be pushed to the right, and on the other hand, it is pushed into the 4 registers below the tap in sequence.

[0051] The conventional writing method of writing signal data in a one-dimensional data stream into a one-dimensional storage space (N is FIFO) is consistent with the writing method in the embodiment of the present invention, and will not be described in detail here.

[0052] However, when a multidimensional data stream is written into a multidimensional storage space, the conventional writing method is different from the solution of the embodiment of the present invention.

[0053] refer to Figure 4 The schematic diagram of two-dimensional slicing of a frame of multidimensional data stream along the Z axis is as follows. Here, a frame of 1920*1920*1920 three-dimensional data stream is taken as an example. The 1920*1920*1920 three-dimensional data stream is reduced in dimension in the form of two-dimensional slicing. Each fragment is 0-3686399, 3686400-7372799, 7372800-11059199, ..., 7066828800-7070515199, 7070515200-7074201599, 7074201600-7077007999. Here, each fragment can be regarded as a two-dimensional subframe of one frame of three-dimensional code stream, and is processed by the above-mentioned "writing signal data in one-dimensional data stream into one-dimensional storage space". However, since the three-dimensional code stream has been reduced to two dimensions, the function body can only capture the neighborhood data on the same two-dimensional plane, and cannot capture the neighborhood data between different two-dimensional planes. Therefore, the correlation between the data in the third dimension and the first two dimensions cannot be fully utilized, thereby sacrificing a large part of the signal processing quality.

[0054] In the scheme of the embodiment of the present invention, each M-dimensional data set that has been written into the N-dimensional storage space is looped back based on the corresponding loopback cycle. Here, each dimensional data set has its corresponding pre-set loopback cycle. By adjusting the data processing architecture of the FIFO and pre-setting the loopback cycle, each M-dimensional data set that has been written into the N-dimensional storage space can be looped back based on the corresponding loopback cycle during the process of writing signal data. In this way, real-time caching of a large number of multi-dimensional data streams can be achieved without consuming a large amount of hardware resources and ensuring signal quality, thereby improving the deficiencies of the traditional N-dimensional storage space when processing N-dimensional data streams. The specific loopback process is described in detail in the following embodiments.

[0055] It should be noted that, by writing the N-dimensional data stream into the N-dimensional storage space using the traditional method and the solution of the embodiment of the present invention respectively, the effect comparison results shown in Table 1 can be obtained. It can be seen from Table 1 that by using the method of the embodiment of the present invention, the N-dimensional data stream is written into the N-dimensional storage space, which improves the shortcomings of the traditional N-dimensional storage space in processing the N-dimensional data stream, and achieves the effect of real-time caching of a large number of multi-dimensional data streams without consuming a large amount of hardware resources and ensuring signal quality.

[0056]

[0057]

[0058] The technical solution of the embodiment of the present invention improves the architecture of the N-dimensional storage space, pre-sets the loopback period, writes the signal data in the acquired N-dimensional data stream into the N-dimensional storage space in sequence, and in the process of writing the signal data, loops back each M-dimensional data set that has been written into the N-dimensional storage space based on the corresponding loopback period until the writing of the N-dimensional data stream is completed, thereby improving the shortcomings of the traditional N-dimensional storage space in processing the N-dimensional data stream, and achieving the effect of real-time caching of a large number of multi-dimensional data streams without consuming a large amount of hardware resources and ensuring signal quality.

[0059] Embodiment 2

[0060] Figure 5 A flowchart of a multi-dimensional data stream processing method provided for Embodiment 2 of the present invention, the embodiment of the present invention can be combined with the various optional schemes in the above embodiments. In the embodiment of the present invention, optionally, before sequentially writing the signal data in the N-dimensional data stream into the N-dimensional storage space, the method further includes: processing the N-dimensional data stream in blocks to obtain multiple N-dimensional data stream blocks. Correspondingly, sequentially writing the signal data in the N-dimensional data stream into the N-dimensional storage space includes: for any N-dimensional data stream block, sequentially writing the signal data in the current N-dimensional data stream block into the N-dimensional storage space.

[0061] like Figure 5 As shown, the method of the embodiment of the present invention specifically includes the following steps:

[0062] S210. Obtain an N-dimensional data stream, where N is a positive integer greater than or equal to 2.

[0063] S220 , dividing the N-dimensional data stream into blocks to obtain multiple N-dimensional data stream blocks.

[0064] Exemplarily, the N-dimensional data stream block can be obtained by dividing the N-dimensional data stream into blocks. Because for the obtained N-dimensional data stream, when the amount of data in the N-dimensional data stream is large, and the storage space of the N-dimensional storage space is limited, it is impossible to write all the information data in the N-dimensional data stream, so it is necessary to divide the N-dimensional data stream into blocks to obtain multiple N-dimensional data stream blocks. In this way, when writing the N-dimensional data stream into the N-dimensional storage space, the N-dimensional data stream blocks can be written into the N-dimensional storage space in sequence according to the association of the multiple N-dimensional data stream blocks. In this way, it is ensured that the associated N-dimensional data streams can be stored in the N-dimensional storage space at the same time.

[0065] It should be noted that when the N-dimensional data stream is divided into blocks, the blocks can be divided according to the storage space of the N-dimensional storage space. For example, if the N-dimensional storage space can only accommodate 300 signal data, a frame of 64*64*64 3-dimensional data stream cannot be completely stored in the N-dimensional storage space. In this case, the 64*64*64 3-dimensional data stream can be divided into 8*8*8 8*8*8 3-dimensional data stream blocks.

[0066] S230. For any N-dimensional data stream block, write the signal data in the current N-dimensional data stream block into the N-dimensional storage space in sequence.

[0067] Exemplarily, the current N-dimensional data stream block may be an N-dimensional data stream block that is currently writing signal data.

[0068] When the signal data in multiple N-dimensional data stream blocks are sequentially written into the N-dimensional storage space, the data should be stored according to the association of each N-dimensional data stream block, so that the associated N-dimensional data stream can be stored in the N-dimensional storage space.

[0069] Optionally, for any N-dimensional data flow block, the signal data in the current N-dimensional data flow block is written into the N-dimensional storage space in sequence. Specifically, it can be: determining an associated N-dimensional data flow block that is relevant to the current N-dimensional data flow block; and synchronously writing the current N-dimensional data flow block and the associated N-dimensional data flow block into the N-dimensional storage space.

[0070] Exemplarily, the associated N-dimensional data stream block may be an N-dimensional data stream block that is correlated with the current N-dimensional data stream block.

[0071] If a certain N-dimensional data stream block is the current N-dimensional data stream block, it is necessary to first determine the associated N-dimensional data stream block with which it is correlated. The specific determination method can be: extract any original signal data in the current N-dimensional data stream block, and determine the coordinate information of the associated signal data based on the coordinate information of the original signal data and the number of dimensions N of the data stream block, wherein the coordinate information of the associated signal data is the coordinate information of at least one coordinate value in the coordinate information of the original signal data and 2 N Perform addition / subtraction operations to obtain; determine the N-dimensional data stream block where the associated signal data is located as the associated N-dimensional data stream block.

[0072] For example, taking the N-dimensional data flow block as a three-dimensional data flow block, refer to Figure 6 A schematic diagram of determining the associated N-dimensional data flow block. Figure 6 The structural block in represents a three-dimensional data stream, wherein A is the current three-dimensional data stream block, and any original signal data in A is selected, and at least one coordinate value in the coordinate information of the signal data in A, for example, the coordinate value of the signal data at the most boundary in A is combined with 2 3For example, if the coordinate value of the signal data at the upper right corner of A is (1,1,1), then add this coordinate value to 2 3 Perform addition / subtraction operations to obtain the coordinate value (8,8,8). The three-dimensional data flow block where the signal data corresponding to the coordinate value is located is the associated three-dimensional data flow block of A, that is, Figure 6 B, C and D in the figure are all associated N-dimensional data flow blocks that are related to A.

[0073] After determining the associated N-dimensional data stream block that is correlated with the current N-dimensional data stream block, the current N-dimensional data stream block and the associated N-dimensional data stream block can be synchronously written into the N-dimensional storage space. This ensures that the data stream written into the N-dimensional storage space is a correlated data stream, and achieves the effect of real-time caching of a large number of multi-dimensional data streams without consuming a large amount of hardware resources and ensuring signal quality.

[0074] S240. During the writing process of the signal data, each M-dimensional data set that has been written into the N-dimensional storage space is looped back based on the corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0075] The technical solution of the embodiment of the present invention is to obtain multiple N-dimensional data stream blocks by dividing the N-dimensional data stream into blocks, and for any N-dimensional data stream block, sequentially write the signal data in the current N-dimensional data stream block into the N-dimensional storage space, so that when writing the N-dimensional data stream into the N-dimensional storage space, the N-dimensional data stream blocks can be sequentially written into the N-dimensional storage space according to the association of the multiple N-dimensional data stream blocks. This ensures that the associated N-dimensional data streams can be stored in the N-dimensional storage space at the same time.

[0076] Embodiment 3

[0077] Figure 7The flowchart of the multi-dimensional data stream processing method provided for the third embodiment of the present invention, the embodiment of the present invention can be combined with each optional scheme in the above embodiment. In the embodiment of the present invention, optionally, the sequentially writing the signal data in the current N-dimensional data stream block into the N-dimensional storage space includes: in the process of writing into the N-dimensional storage space, sequentially writing into the current M-dimensional storage space according to the writing order of the signal data in the data stream, when the current M-dimensional storage space completes writing, adjusting the position of each M-dimensional storage space, and writing into the next looped M-dimensional storage space, wherein the current M-dimensional storage space belongs to the N-dimensional storage space. The looping back of each M-dimensional data set that has been written into the N-dimensional storage space based on the corresponding looping cycle includes: according to the looping cycle of the M-dimensional data set, controlling the storage position of the M-1-dimensional data set to change in a ring, wherein the one-dimensional data set includes a first preset number of signal data, and the second preset number of M-1-dimensional data sets constitute the M-dimensional data set. After the writing of the N-dimensional data stream is completed, the method also includes: compressing, encoding, and feature extraction of the signal data written into the N-dimensional storage space.

[0078] like Figure 7 As shown, the method of the embodiment of the present invention specifically includes the following steps:

[0079] S310. Obtain an N-dimensional data stream, where N is a positive integer greater than or equal to 2.

[0080] S320, dividing the N-dimensional data stream into blocks to obtain multiple N-dimensional data stream blocks.

[0081] S330. For any N-dimensional data stream block, during the writing process to the N-dimensional storage space, the signal data in the data stream are written into the current M-dimensional storage space in sequence according to the writing order. When the current M-dimensional storage space is completed, the position of each M-dimensional storage space is adjusted, and the next looped M-dimensional storage space is written, wherein the current M-dimensional storage space belongs to the N-dimensional storage space.

[0082] Exemplarily, when any N-dimensional data stream block is written into the N-dimensional storage space, in the process of writing into the N-dimensional storage space, the signal data in the data stream are sequentially written into the current M-dimensional storage space according to the writing order, and when the current M-dimensional storage space is finished writing, the position of each M-dimensional storage space is adjusted according to the preset loopback rule, and the writing is performed into the next looped M-dimensional storage space. In this way, the effect of writing the associated N-dimensional data stream into the N-dimensional storage space can be achieved.

[0083] Optionally, the position of each M-dimensional storage space is adjusted and the M-dimensional storage space of the next loop is written, including: using the M-dimensional storage space that has not been written at the current tail position in the N-th dimensional storage space as the M-dimensional storage space of the next loop; adjusting the M-dimensional storage space of the next loop to the head position in the N-th dimensional storage space, and sequentially adjusting the positions of other M-dimensional storage spaces, wherein each M-dimensional storage space includes a set of M-dimensional storage spaces corresponding to each integer value in the interval [1, N-1].

[0084] For example, taking the current data stream block as 8*8*8 signal data and the N-dimensional storage space as N-dimensional FIFO as an example, when writing to the second storage space, the signal data of the second storage space is written into the first storage space, as shown in Figure 2 The correspondence between the N-dimensional data stream and the M-dimensional data set in the N-dimensional storage space and the M-dimensional storage space Figure 2 The corresponding relationship is consistent, and a preset number of M-dimensional storage spaces are combined together to form an N-dimensional storage space.

[0085] refer to Figure 8 The schematic diagram of the position adjustment of the storage space in the embodiment of the present invention takes the current data stream block as 8*8*8 signal data and the N-dimensional storage space as an N-dimensional FIFO as an example. In the figure, FIFO-0-FIFO-7 respectively represent different storage spaces in the M-dimensional storage space.

[0086] like Figure 8 As shown in Figure a, before the signal data is written into the FIFO, the arrangement order of FIFO-0-FIFO-7 is FIFO-7-FIFO-0. When the first signal data "0" in the current 8*8*8 three-dimensional data stream block is written into the first storage space FIFO-0 in the three-dimensional FIFO, FIFO-0 moves from the end to the beginning, and FIFO-7-FIFO-1 moves down in sequence, as shown in Figure a. Figure 8 As shown in Figure b. After the first signal data "0" in the 8*8*8 three-dimensional data flow block is written into the first storage space FIFO-0 in the three-dimensional FIFO, when the data "1" is input, the "0" is pushed forward and the "1" enters the upper right corner, as shown in Figure b. Figure 8 As shown in Figure c, when data is written, it is written from the upper right corner. During the writing process of the first storage space FIFO-0, the storage space FIFO-1 that has not been written at the tail position in the current sequence is used as the storage space for the next loop. When the storage space in the storage space FIFO-0 is filled, the storage space FIFO-1 of the next loop is adjusted to the head position of the storage space FIFO-0, and the other FIFO-7-FIFO-2 are moved down in sequence, as shown in Figure 2. Figure 8As shown in Figure d in the figure. And so on, until the current 8*8*8 three-dimensional data stream block is completely written into the three-dimensional FIFO.

[0087] In addition, during the loopback process between the storage spaces FIFO-0-FIFO-7, loopback is also performed inside each storage space in FIFO-0-FIFO-7, such as Figure 8 As shown in Figure d, when the 9th signal data "8" begins to be written into the storage space of FIFO-1, the 8 signal data "0-7" in the storage space of FIFO-0 are also looped back. When the signal data "8" is written into the storage space of FIFO-1, the signal data "0" in the storage space of FIFO-0 is looped back to the tail position of the storage space of FIFO-0, and the signal data "1-7" moves forward in sequence. When the signal data "9" is written into the storage space of FIFO-1, the signal data "1" in the storage space of FIFO-0 is looped back to the tail position of the storage space of FIFO-0, and the signal data "2-7" moves forward in sequence. And so on, until the storage space of FIFO-1 is filled.

[0088] In the prior art, reference Fig. 9 A schematic diagram of the position adjustment of the storage space in the prior art, such as Fig. 9 As shown in Figure a, before the signal data is written into the FIFO, the arrangement order of FIFO-0-FIFO-7 is FIFO-7-FIFO-0. When the first signal data "0" in the current 8*8*8 three-dimensional data stream block is written into the first storage space FIFO-0 in the three-dimensional FIFO, the arrangement order of FIFO-0-FIFO-7 remains unchanged, and the 8 signal data "0-7" are written into the storage space FIFO-0 in sequence, as shown in Figure a. Fig. 9 As shown in Figure b, when data is written here, Figure 8 Same, both are written from the upper right corner. When the storage space in FIFO-0 is filled, continue to write the 8 signal data "8-15". During this process, it is still written in FIFO-0. Every time a data is written, the first data in FIFO-0 will be moved to the next storage space first. That is, when "8" is written, "0" is pushed into the last position in FIFO-1, and when "9" is written, "1" is pushed into the last position in FIFO-1, and "0" moves forward, and so on. However, the arrangement order of FIFO-0-FIFO-7 remains unchanged, such as Fig. 9 As shown in Figure c in the figure. And so on, until the 8*8*8 three-dimensional data stream is written into the three-dimensional FIFO.

[0089] This loopback method in the embodiment of the present invention can write the associated N-dimensional data stream into the N-dimensional storage space.

[0090] S340. During the writing process of the signal data, the storage position of the M-1 dimensional data set is controlled to change in a circular manner according to the loop period of the M dimensional data set, wherein the one dimensional data set includes a first preset number of signal data and a second preset number of M-1 dimensional data sets constitute the M dimensional data set.

[0091] Exemplarily, the first preset number may be the number of preset 1-dimensional data sets.

[0092] The second preset number may be a preset number of M-1 dimensional data sets.

[0093] When M is greater than or equal to 2, taking M as 3 as an example, if the data stream block is an 8*8*8 data stream block, after the 8 one-dimensional data sets of signal data "0-63" in the 8*8*8 three-dimensional data set are written in the manner in S330 above, the following is formed: Fig.10 The schematic diagram of the one-dimensional data set is that the eight one-dimensional data sets are regarded as a combination to form the first two-dimensional data set, that is, a preset number of M-1 dimensional data sets constitute an M dimensional data set.

[0094] When the eight one-dimensional data sets are regarded as a combination to form the first two-dimensional data set, the signal data "64-127" are written in sequence. When the two-dimensional data set "64-127" is written into the N-dimensional FIFO in the manner described in S330, the eight one-dimensional data sets "0-63" are still changing in a circular manner at their respective storage locations and have not stopped looping back. In other words, the two-dimensional data set currently looping back also controls the one-dimensional data set that has been written previously to change in a circular manner at its storage location.

[0095] It should be noted that the reference Fig.11 The one-dimensional data set is a schematic diagram for determining a preset number of signal data. Taking the N-dimensional data stream as a one-dimensional data stream composed of 8 signal data "0-7" as an example, the first preset number here is 8, that is, the 8 signal data "0-7". At this time, when M is 1, there is only a one-dimensional data set, and the one-dimensional data set is the preset number of signal data, such as Fig.11 As shown, the one-dimensional data set in FIFO-0 is 8 signal data "0-7".

[0096] It should be noted that the loop period of the M-dimensional data set here is the writing period of the M-1-dimensional data set, and the loop period of the 1-dimensional data set is the writing period of one signal data. For example, taking M as 3, if the data stream block is an 8*8*8 data stream block, when the writing of the 8 one-dimensional data sets of the signal data "0-63" in the 8*8*8 three-dimensional data set is completed, these 8 one-dimensional data sets are regarded as a combination to form the first two-dimensional data set. The loop period of this two-dimensional data set is its writing period during the writing process, and when it is written, it is a one-dimensional data set. 8 one-dimensional data sets constitute this two-dimensional data set. Therefore, the loop period of this two-dimensional data set is the writing period of the one-dimensional data set. The loop period of the 1-dimensional data set is the writing period of one signal data.

[0097] In this way, the associated N-dimensional data streams are written into the N-dimensional storage space and are placed together in the N-dimensional storage space, so that the associated N-dimensional data streams can be subsequently calculated.

[0098] S350, performing at least one of compression, encoding, and feature extraction on the signal data written in the N-dimensional storage space.

[0099] Exemplarily, after the N-dimensional data stream is written into the N-dimensional storage space, at least one of compression, encoding and feature extraction may be performed on the signal data written into the N-dimensional storage space according to user requirements.

[0100] refer to Fig.12 The schematic diagram of the solution of the embodiment of the present invention and the prior art solution after storing N-dimensional data stream, Fig.12 In the example, N=2, a single frame signal data is a 8*8 two-dimensional data stream, and a block size of 4*4 two-dimensional data stream blocks are used. Fig.12 It can be seen that signal data 5, 6, 7, 20, ..., 33, 34, 35, 48 are correlated, and similarly, 20-31, 48-51 are correlated. After the storage method of the above embodiment of the present invention, the correlated signal data can be stored together, for example, all stored in a rectangle, such as the rectangle surrounded by 5, 6, 7, 20, ..., 33, 34, 35, 48, the rectangle surrounded by 20-31, 48-51, etc. In this way, during feature extraction, the correlated signal data can be easily extracted for subsequent calculations. If the existing technical solution is used, it can be seen that the correlated data are not together, and are even far apart, so the correlated signal data cannot be extracted well. This makes subsequent calculations difficult.

[0101] By writing the N-dimensional data stream into the N-dimensional storage space in the above manner, the extracted features are signal data with correlation, so that the signal data with correlation can be subsequently calculated and processed.

[0102] It should be noted that at least one of compression, encoding and feature extraction of the signal data written into the N-dimensional storage space may also be performed during the process of writing the N-dimensional data stream into the N-dimensional storage space, which is not limited here.

[0103] The technical solution of the embodiment of the present invention is that for any N-dimensional data stream block, in the process of writing into the N-dimensional storage space, the signal data in the data stream are sequentially written into the current M-dimensional storage space according to the writing order, and when the current M-dimensional storage space is completed, the position of each M-dimensional storage space is adjusted, and the next looped M-dimensional storage space is written, so that the effect of writing the associated N-dimensional data stream into the N-dimensional storage space can be achieved. In the process of writing the signal data, according to the loop cycle of the M-dimensional data set, the storage position of the M-1-dimensional data set is controlled to change in a circular manner, and the signal data written in the N-dimensional storage space is compressed, encoded, and feature extracted. At least one of the processes is performed, so that the associated N-dimensional data stream is written into the N-dimensional storage space and is together in the N-dimensional storage space, so that the associated N-dimensional data stream can be subsequently calculated.

[0104] Embodiment 4

[0105] Fig.13 A schematic diagram of the structure of a multi-dimensional data stream processing device provided in Embodiment 4 of the present invention is shown in FIG. Fig.13 As shown, the device includes: an N-dimensional data stream acquisition module 31 and a data writing module 32.

[0106] Wherein, the N-dimensional data stream acquisition module 31 is used to acquire the N-dimensional data stream, where N is a positive integer greater than or equal to 2;

[0107] The data writing module 32 is used to write the signal data in the N-dimensional data stream into the N-dimensional storage space in sequence, and during the writing process of the signal data, loop back each M-dimensional data set that has been written into the N-dimensional storage space based on the corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0108] Based on the technical solution of the above embodiment, the device further includes:

[0109] The block module is used to process the N-dimensional data stream in blocks to obtain multiple N-dimensional data stream blocks.

[0110] Based on the technical solution of the above embodiment, the data writing module 32 includes:

[0111] The first data writing unit is used to write the signal data in the current N-dimensional data stream block into the N-dimensional storage space in sequence for any N-dimensional data stream block, and during the writing process of the signal data, loop back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loop cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1].

[0112] Based on the technical solution of the above embodiment, the data writing module 32 further includes:

[0113] An associated N-dimensional data stream block determining unit, used to determine an associated N-dimensional data stream block having correlation with the current N-dimensional data stream block;

[0114] The second data writing unit is used to synchronously write the current N-dimensional data stream block and the associated N-dimensional data stream block into the N-dimensional storage space.

[0115] Based on the technical solution of the above embodiment, the associated N-dimensional data stream block determination unit is specifically used for:

[0116] Extract any original signal data in the current N-dimensional data stream block, and determine the coordinate information of the associated signal data based on the coordinate information of the original signal data and the number of dimensions N of the data stream block, wherein the coordinate information of the associated signal data is the sum of at least one coordinate value in the coordinate information of the original signal data and 2 N Perform addition / subtraction operations to obtain;

[0117] The N-dimensional data stream block where the associated signal data is located is determined as the associated N-dimensional data stream block.

[0118] Based on the technical solution of the above embodiment, the first data writing unit includes:

[0119] The storage space adjustment subunit is used for, for any N-dimensional data stream block, writing the signal data in the data stream into the current M-dimensional storage space in sequence according to the writing order of the signal data in the data stream during the writing process of the N-dimensional storage space; when the current M-dimensional storage space is completed, adjusting the position of each M-dimensional storage space, and writing to the next looped M-dimensional storage space, wherein the current M-dimensional storage space belongs to the N-dimensional storage space.

[0120] Based on the technical solution of the above embodiment, the storage space adjustment subunit is specifically used for:

[0121] Using the unwritten M-dimensional storage space at the current tail position in the N-dimensional storage space as the M-dimensional storage space of the next loop;

[0122] The M-dimensional storage space of the next loop is adjusted to the head position in the N-dimensional storage space, and the positions of other M-dimensional storage spaces are adjusted sequentially, wherein each M-dimensional storage space includes a set of M-dimensional storage spaces corresponding to each integer value in the interval [1, N-1].

[0123] Based on the technical solution of the above embodiment, the first data writing unit further includes:

[0124] The loop control subunit is used to control the storage position of the M-1 dimensional data set to change in a loop according to the loop period of the M dimensional data set, wherein the one dimensional data set includes a first preset number of signal data and a second preset number of M-1 dimensional data sets constitute the M dimensional data set.

[0125] Optionally, the loop period of the M-dimensional data set is the writing period of the M-1-dimensional data set, wherein the loop period of the one-dimensional data set is the writing period of one signal data.

[0126] Based on the technical solution of the above embodiment, the N-dimensional data stream acquisition module 31 is specifically used for:

[0127] The signal collector array collects signal data based on an N-1 dimensional arrangement;

[0128] The signal data collected by each signal collector are combined based on the arrangement of the signal collector array and the collection timestamp to form an N-dimensional data stream.

[0129] Optionally, the N-dimensional data stream includes: an N-dimensional image data stream, an N-dimensional video data stream, an N-dimensional audio data stream, and an N-dimensional signal data stream.

[0130] Optionally, the N-dimensional storage space is an N-dimensional first-in-first-out queue space.

[0131] Based on the technical solution of the above embodiment, the device further includes:

[0132] The data processing module is used to perform at least one of compression, encoding, and feature extraction on the signal data written in the N-dimensional storage space.

[0133] The multidimensional data stream processing device provided in the embodiment of the present invention can execute the multidimensional data stream processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0134] Embodiment 5

[0135] Fig.14 A schematic diagram of a device provided in Embodiment 5 of the present invention is shown in FIG. Fig.14 As shown, the device includes a processor 70, a memory 71, an input device 72 and an output device 73; the number of processors 70 in the device can be one or more. Fig.14 A processor 70 is taken as an example; the processor 70, the memory 71, the input device 72 and the output device 73 in the device can be connected by a bus or other means. Fig.14 The example of connecting through bus is taken in the following.

[0136] The memory 71, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the multi-dimensional data stream processing method in the embodiment of the present invention (for example, the N-dimensional data stream acquisition module 31 and the data writing module 32). The processor 70 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 71, that is, implements the above-mentioned multi-dimensional data stream processing method.

[0137] The memory 71 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 71 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 71 may further include a memory remotely arranged relative to the processor 70, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0138] The input device 72 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 73 may include a display device such as a display screen.

[0139] Embodiment 6

[0140] Embodiment 6 of the present invention further provides a storage medium containing computer executable instructions, wherein the computer executable instructions are used to execute a multi-dimensional data stream processing method when executed by a computer processor.

[0141] Of course, the computer executable instructions of a storage medium containing computer executable instructions provided by an embodiment of the present invention are not limited to the method operations described above, and can also execute related operations in the multidimensional data stream processing method provided by any embodiment of the present invention.

[0142] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0143] It is worth noting that in the embodiment of the above-mentioned multi-dimensional data stream processing device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0144] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A multidimensional data stream processing method, It is characterized in that The method comprises: Obtain an N-dimensional data stream, where N is a positive integer greater than or equal to 2; Processing the N-dimensional data stream in blocks to obtain multiple N-dimensional data stream blocks; Sequentially write the signal data in the N-dimensional data stream into the N-dimensional storage space, and in the process of writing the signal data, loop back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loopback cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes a set of M-dimensional data corresponding to each integer value in the interval [1, N-1]; Wherein, the step of sequentially writing the signal data in the N-dimensional data stream into the N-dimensional storage space comprises: for any N-dimensional data stream block, sequentially writing the signal data in the current N-dimensional data stream block into the N-dimensional storage space; wherein, for any N-dimensional data stream block, sequentially writing the signal data in the current N-dimensional data stream block into the N-dimensional storage space comprises: determining an associated N-dimensional data stream block having a correlation with the current N-dimensional data stream block, and synchronously writing the current N-dimensional data stream block and the associated N-dimensional data stream block into the N-dimensional storage space; Wherein, in the process of writing to the N-dimensional storage space, the signal data in the data stream are written into the current M-dimensional storage space in sequence according to the writing order of the signal data in the data stream. When the current M-dimensional storage space is completed, the position of each M-dimensional storage space is adjusted, and the next looped M-dimensional storage space is written, and the current M-dimensional storage space belongs to the N-dimensional storage space; wherein, the adjusting the position of each M-dimensional storage space and writing to the next looped M-dimensional storage space include: taking the M-dimensional storage space that has not been written at the current tail position in the N-dimensional storage space as the next looped M-dimensional storage space, adjusting the next looped M-dimensional storage space to the head position in the N-dimensional storage space, and sequentially adjusting the positions of other M-dimensional storage spaces, and each M-dimensional storage space includes a set of M-dimensional storage spaces corresponding to each integer value in the interval [1, N-1].

2. The method according to claim 1, It is characterized in that The determining of an associated N-dimensional data stream block having relevance to the current N-dimensional data stream block comprises: Extract any original signal data in the current N-dimensional data stream block, and determine the coordinate information of the associated signal data based on the coordinate information of the original signal data and the number of dimensions N of the data stream block, wherein the coordinate information of the associated signal data is obtained by performing an addition / subtraction operation on at least one coordinate value in the coordinate information of the original signal data and 2N; The N-dimensional data stream block where the associated signal data is located is determined as the associated N-dimensional data stream block.

3. The method according to claim 1, It is characterized in that The looping back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding looping cycle includes: According to the loop cycle of the M-dimensional data set, the storage position of the M-1-dimensional data set is controlled to change in a circular manner, wherein the one-dimensional data set includes a first preset number of signal data and a second preset number of M-1-dimensional data sets constitute the M-dimensional data set.

4. The method according to claim 3, It is characterized in that The loop period of the M-dimensional data set is the writing period of the M-1-dimensional data set, wherein the loop period of the one-dimensional data set is the writing period of one signal data.

5. The method according to claim 1, It is characterized in that The obtaining of the N-dimensional data stream comprises: The signal collector array collects signal data based on an N-1 dimensional arrangement; The signal data collected by each signal collector are combined based on the arrangement of the signal collector array and the collection timestamp to form an N-dimensional data stream.

6. The method according to claim 1, It is characterized in that The N-dimensional data stream includes: an N-dimensional image data stream, an N-dimensional video data stream, an N-dimensional audio data stream, and an N-dimensional electric energy data stream.

7. The method according to claim 1, It is characterized in that The N-dimensional storage space is an N-dimensional first-in-first-out queue space.

8. The method according to claim 1, It is characterized in that The method further comprises: The signal data written in the N-dimensional storage space is subjected to at least one of compression, encoding, and feature extraction.

9. A multidimensional data stream processing device, It is characterized in that include: An N-dimensional data stream acquisition module, used to acquire an N-dimensional data stream, where N is a positive integer greater than or equal to 2; A block division module, used for dividing the N-dimensional data stream into blocks to obtain multiple N-dimensional data stream blocks; A data writing module, used for sequentially writing the signal data in the N-dimensional data stream into the N-dimensional storage space, and in the process of writing the signal data, looping back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loopback cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1]; Wherein, the data writing module comprises: A first data writing unit is used for writing the signal data in the current N-dimensional data stream block into the N-dimensional storage space in sequence for any N-dimensional data stream block, and in the process of writing the signal data, looping back each M-dimensional data set that has been written into the N-dimensional storage space based on a corresponding loopback cycle until the writing of the N-dimensional data stream is completed, wherein each M-dimensional data set includes an M-dimensional data set corresponding to each integer value in the interval [1, N-1]; An associated N-dimensional data stream block determining unit, used to determine an associated N-dimensional data stream block having correlation with the current N-dimensional data stream block; A second data writing unit, used for synchronously writing the current N-dimensional data stream block and the associated N-dimensional data stream block into the N-dimensional storage space; Wherein, the first data writing unit includes a storage space adjustment subunit, which is used for, for any N-dimensional data stream block, in the process of writing to the N-dimensional storage space, writing the signal data in the data stream into the current M-dimensional storage space in sequence according to the writing order of the signal data in the data stream; when the current M-dimensional storage space is completed, adjusting the position of each M-dimensional storage space, and writing to the next looped M-dimensional storage space, the current M-dimensional storage space belongs to the N-dimensional storage space; wherein, the storage space adjustment subunit is specifically used for: using the M-dimensional storage space that has not been written at the current tail position in the N-dimensional storage space as the next looped M-dimensional storage space; adjusting the next looped M-dimensional storage space to the head position in the N-dimensional storage space, and sequentially adjusting the positions of other M-dimensional storage spaces, wherein each M-dimensional storage space includes a set of M-dimensional storage spaces corresponding to each integer value in the interval [1, N-1].

10. A device, It is characterized in that The device comprises: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the multi-dimensional data stream processing method as described in any one of claims 1 to 8.

11. A storage medium containing computer executable instructions, It is characterized in that When the computer executable instructions are executed by a computer processor, they are used to perform the multi-dimensional data stream processing method as described in any one of claims 1 to 8.

Images