A human action recognition method and system based on dual-stream network

Abstract

The present invention relates to a human action recognition method and system based on a dual-stream network. Based on skeleton information, an undirected space-time graph and an RGB image are extracted respectively, and a first vector and a second vector are obtained by processing using a GCN network and a CNN network respectively, which represent space-time features and action size and direction features respectively, so as to obtain a more accurate human action category recognition effect. Starting from improving the explicit features of skeleton information data and improving the recognition rate of human actions, a dual-stream network based on a GCN network and a CNN is proposed on the basis of skeleton information, which enhances the representation of skeleton information, reduces the influence of skeleton information noise, and effectively improves the recognition accuracy of human actions.

Description

Technical Field

[0001] The present invention relates to the technical field, and in particular to a human motion recognition method and system based on a dual-stream network. Background Art

[0002] Existing human action recognition methods mainly include two categories: RGB-based human action recognition and skeleton-based human action recognition. Among them, human action recognition based on RGB video has been applied earlier and its implementation method is relatively mature, but it generally has problems such as poor algorithm robustness and low recognition rate. At the same time, human action recognition is affected by dynamic background, ambient light and occlusion in RGB video, and its application scope is greatly limited.

[0003] With the popularization of depth cameras and the introduction of RGB image key point detection algorithms, existing technologies can extract human skeleton information more accurately. Skeleton information has the characteristics of relatively simple features and is not affected by external factors. It provides effective support for human action recognition and effectively improves the action recognition ability. Common human action recognition based on skeleton information can be divided into human action recognition based on manual features, human action recognition based on CNN, human action recognition based on RNN, and human action recognition based on GCN.

[0004] However, the existing human action recognition based on skeleton information is generally affected by the noise of bone information points, making the accuracy of human action recognition less than ideal. Furthermore, the calculation of human skeleton information features extracted based on manual features is cumbersome, the generalization ability is too weak, and it is not suitable for recognizing more complex human actions; action recognition based on RNN can extract temporal information well, but the extraction of spatial information is poor, and the calculation is also more complicated; action recognition based on CNN network is good at extracting spatial information, but is more sensitive to temporal information. Therefore, the existing human action recognition methods based on skeleton information still have many inconveniences and limitations in practical applications. Summary of the invention

[0005] In order to address the shortcomings of the prior art, the present invention proposes a human motion recognition method and system based on a dual-stream network. Starting from improving the explicit features of skeleton information data and improving the recognition rate of human motion, a dual-stream network based on GCN network and CNN is proposed on the basis of skeleton information, which enhances the representation of skeleton information, reduces the influence of skeleton information noise, and effectively improves the recognition accuracy of human motion.

[0006] To achieve the above objectives, the technical solutions adopted by the present invention include:

[0007] A human action recognition method based on a dual-stream network, characterized by comprising:

[0008] S1, obtaining skeleton information corresponding to human body action behavior;

[0009] S2, construct an undirected spatiotemporal graph based on skeleton information;

[0010] S3, encoding the skeleton information to obtain an RGB image containing motion information;

[0011] S4. Processing the undirected spatiotemporal graph using a pre-trained GCN network to obtain a first vector, wherein the first vector includes spatial and temporal features of a natural connection of a human skeleton;

[0012] S5. Process the RGB image using a pre-trained CNN network to obtain a second vector, where the second vector includes size and direction features of a human skeleton motion;

[0013] S6, fusing the features of the first vector and the second vector to obtain a third vector;

[0014] S7. Use the third vector to predict the human action category.

[0015] Furthermore, the undirected spatiotemporal graph based on skeleton information is expressed using Formula 1:

[0016] G=(V,E) Formula 1

[0017] Among them, the undirected space-time graph G is composed of a vertex set V and an edge set E;

[0018] The vertex set V is represented by Formula 2:

[0019] V={Vti|t=1,...,T,i=1,...,N} Equation 2

[0020] Among them, T represents the number of source video frames corresponding to the skeleton information, and N represents all the joint points in the skeleton information;

[0021] The edge set E includes connections between different joint points in the same frame of the corresponding source video, and connections between the same joint points between different frames of the corresponding source video.

[0022] Furthermore, the network definition formula of the GCN network is shown in Formula 3,

[0023]

[0024] Among them, f in Indicates the vertex set V corresponding to the input feature, f out Represents the output feature, Λ represents the diagonal matrix, A represents the single-frame human skeleton connection corresponding to the connection between the same joint points between different frames, the unit matrix I represents the joint self-connection corresponding to the connection between different joint points in the same frame, and the weight W represents the weight parameters of multiple outputs.

[0025] Furthermore, the step S3 comprises the following sub-steps:

[0026] S31, applying depth-first tree traversal to the skeleton information to generate a chain order sequence j;

[0027] S32, connecting the chain sequence j row by row according to the time frame t to form a matrix M, wherein the dimension of the matrix M is J×T×3, wherein J is the number of joint points in the skeleton information after traversal in the depth-first tree, T is the total number of frames of the source video, and 3 represents the three-dimensional coordinates x, y, z of each joint point in the skeleton information;

[0028] S33, calculate the motion structure using equation 4 according to the matrix M,

[0029] N M,t =M J,t+d -M J,t Formula 4

[0030] Among them, the matrix N M,t The difference is calculated by two matrices M with a difference of d frames, and its dimension is J×(Td)×3;

[0031] S34, according to the motion structure, use formula 5 to calculate the magnitude of the motion, and use formulas 6-1 to 6-4 to calculate the direction of the motion,

[0032]

[0033]

[0034]

[0035]

[0036]

[0037] Among them, A M,t Indicates the magnitude of the movement, D M,t Represents the stacking of three directions of motion, represents the direction of movement of coordinates x and y, represents the direction of movement of coordinates y and z, Indicates the direction of motion of coordinates z and x; the dimension of motion magnitude A is J×(Td)×1; the dimension of motion direction D is J×(Td)×3;

[0038] S35, combining the size of the motion and the direction of movement to obtain an RGB image containing motion information.

[0039] Furthermore, the step S3 further includes:

[0040] According to formula 7, the motion size A is filtered.

[0041]

[0042] Where m is the threshold of motion size, D′ M,t is the filtered motion direction.

[0043] Further, the step S7 comprises:

[0044] The third vector is input into several binary SVM classifiers respectively and the classification task is performed to obtain the human action category prediction result.

[0045] The present invention also relates to a human action recognition system based on a dual-stream network, characterized in that it comprises:

[0046] The spatiotemporal graph processing module is used to construct an undirected spatiotemporal graph based on skeleton information;

[0047] An RGB image processing module is used to encode the skeleton information to obtain an RGB image containing motion information;

[0048] A GCN network processing module, used for processing the undirected spatiotemporal graph using a pre-trained GCN network to obtain a first vector;

[0049] A CNN network processing module, used for processing the RGB image using a pre-trained CNN network to obtain a second vector;

[0050] A vector fusion module, used for fusing the features of the first vector and the second vector to obtain a third vector;

[0051] The action recognition module is used to predict the human action category using the third vector.

[0052] The present invention also relates to a computer-readable storage medium, characterized in that a computer program is stored on the storage medium, and the computer program implements the above method when executed by a processor.

[0053] The present invention also relates to an electronic device, characterized in that it comprises a processor and a memory;

[0054] The memory is used to store skeleton information, undirected spatiotemporal graph and RGB image;

[0055] The processor is used to execute the above method by calling skeleton information, undirected spatiotemporal graph and RGB image.

[0056] The present invention also relates to a computer program product, comprising a computer program and / or instructions, characterized in that the computer program and / or instructions implement the steps of the above method when executed by a processor.

[0057] The beneficial effects of the present invention are:

[0058] The human motion recognition method and system based on the dual-stream network described in the present invention are adopted to improve the explicit characteristics of skeleton information data and improve the recognition rate of human motion. On the basis of skeleton information, a dual-stream network based on GCN network and CNN is proposed, which enhances the representation of skeleton information, reduces the influence of skeleton information noise, and effectively improves the recognition accuracy of human motion. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 The figure is a flow chart of the method for human action recognition based on a dual-stream network of the present invention.

[0060] Figure 2 It is a schematic diagram of the structure of the human action recognition system based on the dual-stream network of the present invention.

[0061] Figure 3 It is a schematic diagram of an embodiment of the undirected space-time graph of the present invention. DETAILED DESCRIPTION

[0062] In order to more clearly understand the content of the present invention, it will be described in detail with reference to the drawings and embodiments.

[0063] The first aspect of the present invention relates to a process of steps as follows Figure 1 The human action recognition method based on the two-stream network shown includes:

[0064] S1. Obtain skeleton information corresponding to human body action behavior. Any suitable method, such as depth camera acquisition or RGB image key point detection algorithm calculation, can be used to process the target source video to obtain skeleton information.

[0065] S2. Construct an undirected spatiotemporal graph based on skeleton information.

[0066] Specifically, the undirected spatiotemporal graph based on skeleton information is expressed using Formula 1:

[0067] G=(V,E) Formula 1

[0068] Among them, the undirected space-time graph G is composed of a vertex set V and an edge set E;

[0069] The vertex set V is represented by Formula 2:

[0070] V={Vti|t=1,...,T,i=1,...,N} Equation 2

[0071] Among them, T represents the number of source video frames corresponding to the skeleton information, and N represents all the joint points in the skeleton information;

[0072] The edge set E includes connections between different joint points in the same frame of the corresponding source video, and connections between the same joint points between different frames of the corresponding source video.

[0073] The obtained undirected space-time graph is Figure 3 The style shown includes connections between different joints in the same frame, and connections between the same joints in different frames.

[0074] S3, encoding the skeleton information to obtain an RGB image containing motion information, specifically including the following steps:

[0075] S31, applying a depth-first tree traversal to the skeleton information to generate a chain order sequence j, thereby best preserving the spatial relationship between the joints in the original skeleton structure;

[0076] S32, connecting the chain sequence j row by row according to the time frame t to form a matrix M, that is, each column is the time evolution of the joint in the sequence J, and the dimension of the matrix M is J×T×3, where J is the number of joint points in the skeleton information after the depth-first tree traversal order, T is the total number of frames of the source video, and 3 represents the three-dimensional coordinates x, y, z of each joint point in the skeleton information;

[0077] S33, calculate the motion structure using equation 4 according to the matrix M,

[0078] N M,t =M J,t+d -M J,t Formula 4

[0079] Among them, the matrix N M,t The difference is calculated by two matrices M with a difference of d frames, and its dimension is J×(Td)×3;

[0080] S34, according to the motion structure, use formula 5 to calculate the magnitude of the motion, and use formulas 6-1 to 6-4 to calculate the direction of the motion,

[0081]

[0082]

[0083]

[0084]

[0085]

[0086] Among them, A M,t Indicates the magnitude of the movement, D M,t Represents the stacking of three directions of motion, represents the direction of movement of coordinates x and y, represents the direction of movement of coordinates y and z, Indicates the direction of motion of coordinates z and x; the dimension of motion magnitude A is J×(Td)×1; the dimension of motion direction D is J×(Td)×3, similarly, 3 represents the three-dimensional direction (x, y, z);

[0087] S35, combining the size of the motion and the direction of movement to obtain an RGB image containing motion information.

[0088] Preferably, since estimating the direction value of each joint may generate noise values ​​for joints without any movement, the movement size A can be filtered according to Formula 7,

[0089]

[0090] Where m is the threshold of motion size, D′ M,t is the filtered motion direction.

[0091] S4. Use a pre-trained GCN network to process the undirected spatiotemporal graph to obtain a first vector, where the first vector includes spatial and temporal features under natural connection of the human skeleton.

[0092] Preferably, the network definition formula of the GCN network is as shown in Formula 3,

[0093]

[0094] Among them, f in Indicates the vertex set V corresponding to the input feature, f out represents the output feature, Λ represents the diagonal matrix, A represents the connection between the same joint points between different frames corresponding to the single-frame human skeleton connection, the unit matrix I represents the connection between different joint points in the same frame corresponding to the joint self-connection, and the weight W represents the weight parameters of multiple outputs. The network effectively obtains spatial and temporal information through the natural connection of the human skeleton, and then obtains the first vector through the graph convolution network.

[0095] S5. Use a pre-trained CNN network to process the RGB image to obtain a second vector, where the second vector contains size and direction features of the human skeleton movement.

[0096] The action structure of the human skeleton is modeled for each action, and the size and direction features are calculated according to the above network framework. Finally, the two are fused to obtain a skeleton image with rich spatiotemporal information. The generated RGB image is sent to the pre-trained network model to generate the second vector.

[0097] S6. Fusing the features of the first vector and the second vector to obtain a third vector.

[0098] Specifically, the undirected spatiotemporal graph is input into the GCN network and then pooled, and the first vector f is obtained through the fully connected layer. gcn =[x1,x2,…,x n ]; The RGB image is input into the CNN network and then pooled, and the second vector f is obtained through the fully connected layer cnn =[y1,y2,…,y n ] The two eigenvectors are combined to obtain the third vector f fusion =[x1+y1,x2+y2,…,x n +y n ].

[0099] S7. Use the third vector to predict the human action category. In particular, input the third vector into a number of binary SVM classifiers and perform classification tasks to obtain a human action category prediction result.

[0100] Preferably, the SVM classifier formula is shown in Formula 8 to Formula 10:

[0101]

[0102]

[0103]

[0104] Where: w is the normal vector, which determines the direction of the hyperplane; θ is the displacement term: y i ∈{a,b},i=1,2,…,n;The function φ converts the input low-dimensional sample f fusion Mapped to high space; ε i is a slack variable, C is a regularization coefficient, a is a certain category, and b is other categories. After training n classifiers, if only one classifier outputs a positive result, mark the action as the classification result; if multiple classifiers identify it as positive, select the action with the largest confidence interval as the classification result.

[0105] Another aspect of the present invention also relates to a human action recognition system based on a dual-stream network, the structure of which is as follows: Figure 2 As shown, including:

[0106] The spatiotemporal graph processing module is used to construct an undirected spatiotemporal graph based on skeleton information;

[0107] An RGB image processing module is used to encode the skeleton information to obtain an RGB image containing motion information;

[0108] A GCN network processing module, used for processing the undirected spatiotemporal graph using a pre-trained GCN network to obtain a first vector;

[0109] A CNN network processing module, used for processing the RGB image using a pre-trained CNN network to obtain a second vector;

[0110] A vector fusion module, used for fusing the features of the first vector and the second vector to obtain a third vector;

[0111] The action recognition module is used to predict the human action category using the third vector.

[0112] By using this system, the above-mentioned operation processing method can be executed and the corresponding technical effects can be achieved.

[0113] An embodiment of the present invention also provides a computer-readable storage medium capable of implementing all the steps in the method in the above embodiment, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, all the steps in the method in the above embodiment are implemented.

[0114] An embodiment of the present invention also provides an electronic device for executing the above method. As an implementation device of the method, the electronic device has at least a processor and a memory, and in particular, the memory stores data and related computer programs required for executing the method, such as skeleton information, undirected space-time graphs, and RGB images, etc., and the processor calls the data and programs in the memory to execute all the steps of the implementation method and obtains the corresponding technical effects.

[0115] Preferably, the electronic device may include a bus architecture, the bus may include any number of interconnected buses and bridges, and the bus will include various circuits linked together by one or more processors and memories. The bus may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are all well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the receiver and transmitter. The receiver and transmitter may be the same component, i.e., a transceiver, providing a unit for communicating with various other systems over a transmission medium. The processor is responsible for managing the bus and general processing, while the memory may be used to store data used by the processor when performing operations.

[0116] In addition, the electronic device may further include components such as a communication module, an input unit, an audio processor, a display, and a power supply. The processor (or controller, operation control) used may include a microprocessor or other processor device and / or a logic device, which receives input and controls the operation of various components of the electronic device; the memory may be one or more of a cache, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory, or other suitable devices, which may store the above-mentioned related data information, and may also store a program for executing related information, and the processor may execute the program stored in the memory to realize information storage or processing, etc.; the input unit is used to provide input to the processor, for example, it may be a key or a touch input device; the power supply is used to provide power to the electronic device; the display is used to display display objects such as images and text, for example, it may be an LCD display. The communication module is a transmitter / receiver that sends and receives signals via an antenna. The communication module (transmitter / receiver) is coupled to the processor to provide input signals and receive output signals, which may be the same as the case of a conventional mobile communication terminal. Based on different communication technologies, multiple communication modules may be provided in the same electronic device, such as a cellular network module, a Bluetooth module and / or a wireless local area network module, etc. The communication module (transmitter / receiver) is also coupled to a speaker and a microphone via an audio processor to provide an audio output via the speaker and receive an audio input from the microphone, thereby realizing a common telecommunication function. The audio processor may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor is also coupled to a central processor, so that recording can be performed on the machine through the microphone, and the sound stored on the machine can be played through the speaker.

[0117] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0118] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or boxes.

[0119] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0120] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps of the functions specified in a box or multiple boxes. Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0121] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A human action recognition method based on a two-stream network, characterized in that: include: S1, obtaining skeleton information corresponding to human body action behavior; S2, construct an undirected spatiotemporal graph based on skeleton information; S3, encoding the skeleton information to obtain an RGB image containing motion information; S4. Processing the undirected spatiotemporal graph using a pre-trained GCN network to obtain a first vector, wherein the first vector includes spatial and temporal features of a natural connection of a human skeleton; S5. Process the RGB image using a pre-trained CNN network to obtain a second vector, where the second vector includes size and direction features of a human skeleton motion; S6, fusing the features of the first vector and the second vector to obtain a third vector; S7, predicting the human action category using the third vector; The step S3 comprises the following sub-steps: S31, applying depth-first tree traversal to the skeleton information to generate a chain order sequence j; S32, connecting the chain sequence j row by row according to the time frame t to form a matrix M, wherein the dimension of the matrix M is J×T×3, wherein J is the number of joint points in the skeleton information after traversal in the depth-first tree, T is the total number of frames of the source video, and 3 represents the three-dimensional coordinates x, y, z of each joint point in the skeleton information; S33, calculate the motion structure using equation 4 according to the matrix M, N M,t =M J,t+d -M J,t Formula 4 Among them, the matrix N M,t The difference is calculated by two matrices M with a difference of d frames, and its dimension is J×(Td)×3; S34, according to the motion structure, use formula 5 to calculate the magnitude of the motion, and use formulas 6-1 to 6-4 to calculate the direction of the motion, Among them, A M,t Indicates the magnitude of the movement, D M,t Represents the stacking of three directions of motion, represents the direction of movement of coordinates x and y, represents the direction of movement of coordinates y and z, Indicates the direction of motion of coordinates z and x; the dimension of motion magnitude A is J×(Td)×1; the dimension of motion direction D is J×(Td)×3; S35, combining the size of the motion and the direction of movement to obtain an RGB image containing motion information.

2. The method according to claim 1, characterized in that The undirected spatiotemporal graph based on skeleton information is expressed using Formula 1: G=(V,E) Formula 1 Among them, the undirected space-time graph G is composed of a vertex set V and an edge set E; The vertex set V is represented by Formula 2: V={Vti|t=1,...,T,i=1,...,N} Equation 2 Among them, T represents the number of source video frames corresponding to the skeleton information, and N represents all the joint points in the skeleton information; The edge set E includes connections between different joint points in the same frame of the corresponding source video, and connections between the same joint points between different frames of the corresponding source video.

3. The method according to claim 2, characterized in that The network definition formula of the GCN network is shown in Formula 3: Among them, f in Indicates the vertex set V corresponding to the input feature, f out Represents the output feature, Λ represents the diagonal matrix, A represents the single-frame human skeleton connection corresponding to the connection between the same joint points between different frames, the unit matrix I represents the joint self-connection corresponding to the connection between different joint points in the same frame, and the weight W represents the weight parameters of multiple outputs.

4. The method according to claim 1, characterized in that The step S3 further comprises: According to formula 7, the motion size A is filtered. Where m is the threshold of motion size, D' M,t is the filtered motion direction.

5. The method according to claim 1, characterized in that The step S7 comprises: The third vector is input into several binary SVM classifiers respectively and the classification task is performed to obtain the human action category prediction result.

6. A human action recognition system based on a dual-stream network, characterized in that: include: The spatiotemporal graph processing module is used to construct an undirected spatiotemporal graph based on skeleton information; An RGB image processing module is used to encode the skeleton information to obtain an RGB image containing motion information; A GCN network processing module, used for processing the undirected spatiotemporal graph using a pre-trained GCN network to obtain a first vector; A CNN network processing module, used for processing the RGB image using a pre-trained CNN network to obtain a second vector; A vector fusion module, used for fusing the features of the first vector and the second vector to obtain a third vector; an action recognition module, for predicting a human action category using a third vector; The RGB image processing module is used for the following steps: S31, applying depth-first tree traversal to the skeleton information to generate a chain order sequence j; S32, connecting the chain sequence j row by row according to the time frame t to form a matrix M, wherein the dimension of the matrix M is J×T×3, wherein J is the number of joint points in the skeleton information after traversal in the depth-first tree, T is the total number of frames of the source video, and 3 represents the three-dimensional coordinates x, y, z of each joint point in the skeleton information; S33, calculate the motion structure using equation 4 according to the matrix M, N M,t =M J,t+d -M J,t Formula 4 Among them, the matrix N M,t The difference is calculated by two matrices M with a difference of d frames, and its dimension is J×(Td)×3; S34, according to the motion structure, use formula 5 to calculate the magnitude of the motion, and use formulas 6-1 to 6-4 to calculate the direction of the motion, Among them, A M,t Indicates the magnitude of the movement, D M,t Represents the stacking of three directions of motion, represents the direction of movement of coordinates x and y, represents the direction of movement of coordinates y and z, Indicates the direction of motion of coordinates z and x; the dimension of motion magnitude A is J×(Td)×1; the dimension of motion direction D is J×(Td)×3; S35, combining the size of the motion and the direction of movement to obtain an RGB image containing motion information.

7. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

8. An electronic device, characterized in that: including a processor and a memory; The memory is used to store skeleton information, undirected spatiotemporal graph and RGB image; The processor is used to execute the method according to any one of claims 1 to 5 by calling skeleton information, an undirected spatiotemporal graph and an RGB image.

9. A computer program product comprising a computer program and / or instructions, characterized in that: When the computer program and / or the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images