Terrain recognition result sequence online decision-making method, system, device and medium

Abstract

This application discloses a method, system, device, and medium for online decision-making on a sequence of terrain recognition results. The method comprises: determining the current gait cycle based on the online decision-making trigger time point; filtering out erroneous terrain recognition results within the current gait cycle based on the gait pattern; and performing a weighted decision on the remaining terrain recognition results within the current gait cycle based on a weight matrix to obtain a decision result. This method is simple and easy to implement and can improve the accuracy of terrain recognition results.

Description

Technical Field

[0001] The present application relates to a method, system, device and medium for online decision-making of a sequence of terrain recognition results, and belongs to the field of data decision-making. Background Art

[0002] Currently, in the field of lower limb exoskeletons, in order to achieve better control effects, researchers are trying to apply terrain information to control input.

[0003] However, due to various limitations, existing terrain recognition methods cannot provide a relatively stable and accurate sequence of terrain recognition results. Summary of the Invention

[0004] In view of this, the present application provides a method, system, electronic device and readable storage medium for online decision-making of terrain recognition result sequence.

[0005] In a first aspect, embodiments of the present application disclose an online decision-making method for a terrain recognition result sequence, which is implemented based on a gait pattern sequence. The terrain recognition result sequence and the gait pattern sequence share an absolute time axis. The method includes:

[0006] Determine the current gait cycle based on the online decision triggering time point;

[0007] filtering out erroneous terrain recognition results according to the gait pattern in the current gait cycle;

[0008] A weighted decision is performed on the remaining terrain recognition results in the current gait cycle according to the weight matrix to obtain a decision result.

[0009] As a preferred embodiment, the online decision triggering time point is calculated as follows:

[0010]

[0011]

[0012] in, is the decision time point of the nth gait cycle, is the starting timestamp of the nth gait cycle, ΔT n-1 is the duration of the n-1th gait cycle.

[0013] As a preferred embodiment, the determining of the current gait cycle according to the online decision triggering time point includes:

[0014] Determine the current gait cycle time interval according to the decision time point and the starting timestamp

[0015] As a preferred embodiment, the filtering out erroneous terrain recognition results according to the gait pattern in the current gait cycle includes:

[0016] At the online decision triggering time point, the terrain recognition result sequence of the current gait cycle time interval is counted;

[0017] Use the preset state matrix to filter the sequence of terrain recognition results;

[0018] Based on the elements in the filtered terrain recognition result sequence, according to the gait pattern and terrain results, the state Boolean value s corresponding to each frame is obtained i ;

[0019] When s i When it is the first threshold, the terrain result of the frame is retained;

[0020] When s i When it is the second threshold, the terrain result of the frame is filtered out.

[0021] As a preferred embodiment, performing weighted decision on the remaining terrain recognition results in the current gait cycle according to the weight matrix to obtain a decision result includes:

[0022] For the remaining N frames of terrain result sequence, calculate the normalized position of each frame in the gait cycle;

[0023] Obtain the weight corresponding to the terrain result of each frame according to the normalized position of the gait cycle and the gait mode of each frame;

[0024] The weights are used to perform weighted voting on N terrain results to obtain the decision result.

[0025] As a preferred embodiment, the process of obtaining the weight matrix includes:

[0026] The weights of each gait interval of a single gait pattern are defined as a weight sequence;

[0027] The weight sequences of the five gait patterns including walking, climbing stairs, descending stairs, climbing uphill and descending downhill are combined to define a weight matrix.

[0028] As a preferred embodiment, the terrain recognition result sequence includes terrain recognition results, and the gait pattern sequence includes lower limb gait patterns;

[0029] The terrain recognition results are divided into flat ground, obstacles, ascending stairs, descending stairs, ascending slope and descending slope, and the lower limb gait patterns are divided into walking, ascending stairs, descending stairs, uphill, downhill and turning.

[0030] A second aspect of an embodiment of the present application discloses an online decision system for a terrain recognition result sequence, which is implemented based on a gait pattern sequence. The terrain recognition result sequence and the gait pattern sequence share an absolute time axis. The system includes:

[0031] a determination unit, configured to determine a current gait cycle according to an online decision triggering time point;

[0032] a screening unit, configured to screen out erroneous terrain recognition results according to the gait pattern in the current gait cycle;

[0033] The decision unit is configured to perform weighted decision on the remaining terrain recognition results in the current gait cycle according to a weight matrix to obtain a decision result.

[0034] A third aspect of an embodiment of the present application discloses a mobile terminal device, comprising a processor and a memory for storing a program executable by the processor. When the processor executes the program stored in the memory, it implements any one of the terrain recognition result sequence online decision methods disclosed in the embodiment of the present application.

[0035] A fourth aspect of an embodiment of the present application discloses a readable storage medium storing a program. When the program is executed by a processor, any one of the terrain recognition result sequence online decision-making methods disclosed in the embodiment of the present application is implemented.

[0036] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0037] The electronic device determines the current gait cycle based on the online decision trigger time. It then filters out erroneous terrain recognition results within the current gait cycle based on the gait pattern. It then performs a weighted decision on the remaining terrain recognition results within the current gait cycle using a weight matrix to determine a decision result. This computational process is fast, the algorithm is simple, and it can improve the accuracy of terrain recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without any creative work.

[0039] Figure 1 A flowchart of an online decision-making method for a sequence of terrain recognition results provided in an embodiment of the present application.

[0040] Figure 2 A flowchart of an online decision-making method for a sequence of terrain recognition results provided in an embodiment of the present application.

[0041] Figure 3 A data diagram of an online decision-making method for a sequence of terrain recognition results provided in an embodiment of the present application.

[0042] Figure 4 A schematic diagram of a state matrix provided in an embodiment of the present application.

[0043] Figure 5 A schematic diagram of a weight matrix provided in an embodiment of the present application.

[0044] Figure 6 A schematic diagram of the structure of an online decision-making system for terrain recognition result sequence provided in an embodiment of the present application.

[0045] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0047] The inventors discovered that during lower limb movement, there's a logical relationship between lower limb gait patterns and terrain types. For example, on flat terrain, lower limb movements are typically walking or turning. Similarly, within a sequence of lower limb gait patterns, gait patterns that follow a temporal sequence also exhibit a logical relationship. For example, a gait while descending stairs often doesn't connect to a gait while ascending stairs. This logical relationship can be applied to optimize the sequence of terrain recognition results.

[0048] like Figure 1 and Figure 2 As shown, Figure 1 and Figure 2 1 is a flow chart of an online decision-making method for a sequence of terrain recognition results provided in an embodiment of the present application.

[0049] like Figure 3 As shown, the input of the method is a sequence of terrain recognition results and a sequence of gait patterns that are continuously updated in real time over time, and the two sequences use the same absolute time axis.

[0050] The terrain recognition result sequence includes terrain recognition results and timestamps. The terrain recognition results refer to the terrain categories in front of the human lower limb movement, which are divided into six terrain categories: flat ground, obstacles, ascending stairs, descending stairs, ascending slope, and descending slope.

[0051] The obstacles refer to insurmountable objects and terrain structures that prevent the human body from moving forward.

[0052] The gait pattern sequence includes a lower limb gait pattern and a gait cycle start timestamp. The gait pattern refers to the gait pattern of the lower limb movement of the human body, including six gait patterns: walking, climbing stairs, descending stairs, going uphill, going downhill, and turning.

[0053] The terrain recognition result sequence online decision-making method may include the following steps:

[0054] S1. Calculate the online decision triggering time point.

[0055] To achieve online decision making, it is necessary to execute the decision before the next state cycle occurs, such as Figure 3 As shown, it is necessary to estimate the gait length and calculate the decision time point before completing the current gait cycle. The calculation formula for the decision time point of the current cycle is as follows:

[0056]

[0057]

[0058] Among them, during the online process, the current gait cycle is the nth gait cycle, is the decision time point of the nth gait cycle, is the starting timestamp of the nth gait cycle, ΔT n-1 is the length of the n-1th gait cycle.

[0059] S2, filter out the wrong terrain recognition result frame. After determining the decision time point, Execute the subsequent decision-making process. First, count the current gait cycle time interval The terrain recognition result sequence Y = {y1, y2, y3, ..., y m}, use the state matrix S to screen the sequence, the state matrix is ​​composed as follows Figure 4 As shown. In the decision making during the nth gait cycle, for the elements in the terrain sequence Y, according to the gait pattern gait n and the terrain result y i , get the state Boolean value s corresponding to each frame in the sequence i :

[0060] s i =S(gait n ,y i )

[0061] When s i When it is 1, the terrain result of this frame is retained; when s iWhen it is 0, the terrain result of this frame is filtered out.

[0062] S3, use the weight matrix W to make weighted decisions on the remaining results and output them. The weight matrix is ​​as follows Figure 5 As shown, it is a 4×9 matrix. For the N frames of terrain result sequence remaining after the screening in step S2, for example, for the k-th frame terrain result, its corresponding timestamp is t k , calculate the normalized position j in the gait cycle k :

[0063]

[0064]

[0065] in, is the starting timestamp of this gait cycle, such as Figure 3 As shown in ; ΔT n-1 is the duration of the n-1th gait cycle.

[0066] According to the gait mode n and the normalized position j of the gait cycle of the k-th frame terrain k , get the weight w corresponding to the terrain result of this frame.

[0067] Finally, the weights are used to perform a weighted voting decision on the terrain results of N frames in the gait cycle. That is, the weighted sum of the posterior probabilities of each terrain category is calculated, and the classification result with the largest probability sum is output:

[0068]

[0069] y∈{flat ground, obstacle, ascending stairs, descending stairs, ascending slope, descending slope}

[0070] The weight matrix W is a parameter matrix used to assign weights to terrain recognition results in different intervals within the gait cycle under different gait patterns. The weight matrix W is calculated as follows:

[0071] Specifically, the weights of each gait interval of a single gait pattern are defined as a weight sequence. The weight sequences of the five gait patterns that require decision making, including walking, ascending stairs, descending stairs, ascending slope, and descending slope, are combined to form the weight matrix W. The steps for calculating the weight sequence of a single gait pattern are as follows:

[0072] a) Collect terrain recognition result data and gait cycle data samples of the subject in a single gait mode, and manually annotate the correct terrain recognition results;

[0073] b) normalizing the gait period to [0, 1) and dividing it into N intervals. In this embodiment, N is 9;

[0074] c) Assume that the sample data contains M gait cycles and T frames of terrain recognition results. The manually labeled terrain recognition true value sequence is R = {r1, r2, ..., r T};

[0075] d) For the sample x at position i in the mth normalized gait cycle t , the true value of the classification result is r t , the corresponding posterior probability is:

[0076]

[0077] Then the nth interval The weight calculation formula is:

[0078]

[0079] e) Calculate the weight of each interval and obtain the discrete weight sequence w={w1,w2,...,w N}, where N = 9; the weight sequences of the five gait patterns are combined into a matrix, such as Figure 5 shown.

[0080] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above embodiments may be completed by instructing related hardware through a program, and the corresponding program may be stored in a computer-readable storage medium.

[0081] It should be noted that although the method operations of the above embodiments are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all of the illustrated operations must be performed to achieve the desired results. Rather, the depicted steps may be performed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into a single step, and / or a single step may be broken down into multiple steps.

[0082] like Figure 6 As shown, Figure 6 This is a schematic diagram of the structure of an online decision-making system for terrain recognition result sequences provided by an embodiment of the present application. This online decision-making system for terrain recognition result sequences is implemented based on a gait pattern sequence, and the terrain recognition result sequence and the gait pattern sequence share an absolute time axis. This online decision-making system for terrain recognition result sequences may include:

[0083] A determination unit 601 is configured to determine a current gait cycle according to an online decision triggering time point;

[0084] A screening unit 602 is configured to screen out erroneous terrain recognition results according to the gait pattern in the current gait cycle;

[0085] The decision unit 603 is configured to perform weighted decision on the remaining terrain recognition results in the current gait cycle according to a weight matrix to obtain a decision result.

[0086] like Figure 7 As shown, Figure 7 7 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include a processor 702, a memory, an input device 703, an output device 704, and a network interface 705 connected via a system bus 701. The processor 702 is used to provide computing and control capabilities, and the memory includes a non-volatile storage medium 706 and an internal memory 707. The non-volatile storage medium 706 stores an operating system, a computer program, and a database. The internal memory 707 provides an environment for the operation of the operating system and computer program in the non-volatile storage medium 706. When the computer program is executed by the processor 702, any one of the online decision-making methods for terrain recognition result sequences disclosed in the embodiments of the present application is implemented.

[0087] The present application discloses a storage medium that is computer-readable and stores a computer program that, when executed by a processor, implements any one of the terrain recognition result sequence online decision-making methods disclosed in the present application.

[0088] It should be noted that the computer-readable storage medium of the present embodiment may be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0089] In this embodiment, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. Furthermore, in this embodiment, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable program. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable storage medium other than a computer-readable storage medium that can transmit, propagate, or transfer a program for use by or in conjunction with an instruction execution system, apparatus, or device. The computer program contained on a computer-readable storage medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0090] The computer readable storage medium can be written in one or more programming languages ​​or a combination thereof to execute the computer program for performing the present embodiment, including object-oriented programming languages ​​such as Java, Python, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect via the Internet).

[0091] In summary, the present invention makes decisions on the terrain recognition result sequence based on the existing gait pattern sequence, and outputs a terrain sequence corresponding to the lower limb movement in the next gait cycle. Its input is the terrain recognition result sequence and the gait pattern sequence that are continuously updated in real time over time on the same absolute time axis. The method is characterized by calculating the online decision triggering time point. According to the gait pattern, the terrain recognition results are preliminarily screened using the state matrix. The weight matrix is ​​used to make weighted decisions on the terrain recognition results preliminarily screened within the gait cycle, and the decision results are finally output. This method is simple and easy to implement and can improve the terrain recognition results.

[0092] The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes based on the technical solution and inventive concept of the present invention within the scope disclosed by the present invention, which falls within the scope of protection of the present invention.

Claims

1. A method for online decision-making of terrain recognition result sequence, characterized in that: Based on the gait pattern sequence, the terrain recognition result sequence and the gait pattern sequence share an absolute time axis, and the method includes: The current gait cycle is determined according to the online decision triggering time point, and the calculation of the online decision triggering time point is as follows: in, is the decision time point of the nth gait cycle, is the starting timestamp of the nth gait cycle, ΔT n-1 is the duration of the n-1th gait cycle; filtering out erroneous terrain recognition results according to the gait pattern in the current gait cycle; A weighted decision is performed on the remaining terrain recognition results in the current gait cycle according to the weight matrix to obtain a decision result, specifically: For the remaining N frames of terrain result sequence, calculate the normalized position of each frame in the gait cycle; Obtain the weight corresponding to the terrain result of each frame according to the normalized position of the gait cycle and the gait mode of each frame; Use the weights to perform weighted voting on N terrain results to obtain the decision result; The process of obtaining the weight matrix includes: The weights of each gait interval of a single gait pattern are defined as a weight sequence; the weight sequences of five gait patterns are combined to define a weight matrix, and the five gait patterns include walking, climbing stairs, descending stairs, going uphill, and going downhill.

2. The method according to claim 1, characterized in that Determining the current gait cycle according to the online decision triggering time point includes: Determine the current gait cycle time interval according to the decision time point and the starting timestamp 3. The method according to any one of claims 1 to 2, characterized in that: The filtering out erroneous terrain recognition results according to the gait pattern in the current gait cycle includes: At the online decision triggering time point, the terrain recognition result sequence of the current gait cycle time interval is counted; Use the preset state matrix to filter the sequence of terrain recognition results; Based on the elements in the filtered terrain recognition result sequence, according to the gait pattern and terrain results, the state Boolean value s corresponding to each frame is obtained i ; When s i When it is the first threshold, the terrain result of the frame is retained; When s i When it is the second threshold, the terrain result of the frame is filtered out.

4. The method according to claim 1, characterized in that The terrain recognition result sequence includes terrain recognition results, and the gait pattern sequence includes lower limb gait patterns; The terrain recognition results are divided into flat ground, obstacles, ascending stairs, descending stairs, ascending slope and descending slope, and the lower limb gait patterns are divided into walking, ascending stairs, descending stairs, uphill, downhill and turning.

5. An online decision system for terrain recognition result sequence based on gait pattern sequence, characterized in that: The method according to any one of claims 1 to 4 is implemented based on a gait pattern sequence, wherein the terrain recognition result sequence and the gait pattern sequence share an absolute time axis, and the system comprises: a determination unit, configured to determine a current gait cycle according to an online decision triggering time point; a screening unit, configured to screen out erroneous terrain recognition results according to the gait pattern in the current gait cycle; The decision unit is configured to perform weighted decision on the remaining terrain recognition results in the current gait cycle according to a weight matrix to obtain a decision result.

6. An electronic device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, the method according to any one of claims 1 to 4 is implemented.

7. A readable storage medium storing a program, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images