Microcurrent tactile parameter automatic generation method and system based on microscopic geometric feature mapping
By generating microcurrent tactile parameters based on micro-geometric feature mapping, the problems of poor adaptability to unknown textures and high computational resource consumption in existing technologies are solved, and lightweight and interpretable tactile feedback effects are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SOUTH CHINA UNIV OF TECH
- Filing Date
- 2026-01-09
- Publication Date
- 2026-05-08
AI Technical Summary
Existing microcurrent haptic rendering technology has poor adaptability to unknown virtual textures, and deep learning methods consume large computational resources and suffer from modal mismatch, making it difficult to achieve lightweight and highly interpretable haptic feedback.
By analyzing the microscopic geometric features of virtual textures, microcurrent stimulation control parameters are generated using a visual feature mapping model, including explicit feature extraction and linear regression analysis. A visual-tactile roughness and softness mapping model is constructed to drive the finger electrode array for tactile feedback.
It achieves efficient and lightweight tactile feedback for unknown textures, is highly adaptable, and can directly map visual features into electrotactile physical control parameters, making it suitable for operation on mobile VR devices.
Smart Images
Figure CN121999243A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of virtual reality and human-computer interaction technology, and in particular to a method and system for automatically generating microcurrent tactile parameters based on micro-geometric feature mapping. Background Technology
[0002] With the popularization of metaverse and virtual reality (VR) technologies, haptic feedback technology has become key to enhancing immersion. Current microcurrent haptic rendering technology, such as the invention patent CN119292469A published on January 10, 2025, proposes a rendering model based on a three-stage process of "contact-sliding-separation" and establishes a method for modulating the stimulation frequency using the physical roughness parameter λ.
[0003] However, existing data-driven rendering methods typically rely on pre-collected databases of real objects. This means that for unknown or computer-generated new textures in virtual scenes (such as random terrain or sci-fi materials in games), the system cannot know their roughness and softness due to the lack of pre-measured physical parameters, resulting in haptic feedback failure or regression to a single vibration mode.
[0004] While some existing technologies utilize deep learning networks (such as GANs) to directly generate vibration waveforms from images, these methods have significant drawbacks:
[0005] 1. High computing power requirements; complex neural networks are difficult to run in real time on lightweight VR headsets or embedded chips.
[0006] 2. Modal mismatch results in the generation of mechanical vibration acceleration waveforms, which cannot be directly converted into discrete control parameters (such as pulse width, frequency, and number of activated electrodes) required for microcurrent stimulation.
[0007] 3. Lack of interpretability: Black-box models are difficult to accurately adapt to the electrotactile coding rules based on physiological characteristics (such as Weber's Law).
[0008] Therefore, there is an urgent need for a lightweight, highly interpretable method that can directly map the visual features of virtual textures to electrotactile physical control parameters. Summary of the Invention
[0009] This invention provides a method and system for automatically generating microcurrent tactile parameters based on micro-geometric feature mapping. By analyzing the visual features of virtual textures, it automatically generates microcurrent stimulation control parameters, solving the problems of existing technologies that rely heavily on pre-stored databases and have poor adaptability to unknown textures.
[0010] On the one hand, the automatic generation method for microcurrent tactile parameters based on microscopic geometric feature mapping adopted in the embodiments of the present invention includes the following steps:
[0011] S1. Real-time acquisition of microscopic geometric image data of the surface of the virtual object currently in contact with the user in the virtual reality scene;
[0012] S2. Explicitly extract features from the micro-geometric image data and calculate the first visual feature value representing the degree of texture undulation and the second visual feature value representing the sharpness of texture edges.
[0013] S3. Input the first visual feature value into the preset visual-tactile roughness mapping model to calculate the surface spatial wavelength parameter of the virtual object; input the second visual feature value into the preset visual-tactile softness mapping model to calculate the softness coefficient of the virtual object.
[0014] S4. Transmit the surface spatial wavelength parameters and softness coefficient to the microcurrent tactile rendering controller to drive the finger electrode array to generate corresponding tactile feedback.
[0015] Preferably, step S3 includes:
[0016] S31. Constructing a mapping model:
[0017] A variety of representative real material samples were selected, and the surface of each real material sample was pressed using a visual-tactile sensor to collect its microscopic three-dimensional morphology data.
[0018] Each set of collected microscopic three-dimensional morphology data is processed in two ways. On the one hand, visual feature values are calculated, the morphology data is converted into a grayscale texture map, and the first visual feature value representing the degree of texture undulation and the second visual feature value representing the sharpness of texture edges are extracted. On the other hand, physical true values are calculated, the depth data is analyzed in the frequency domain, the frequency peak with the largest power spectral density is extracted, the true average spatial wavelength is calculated, and the normalized true softness coefficient is determined based on the deformation depth of the elastic body.
[0019] Using the first and second visual feature values corresponding to real material samples as independent variables X and the physical truth value as the dependent variable Y, linear regression analysis was performed to construct visual-tactile roughness mapping model and visual-tactile softness mapping model, respectively.
[0020] S32. Generate tactile parameters:
[0021] Substitute the first visual feature value calculated in real time in step S2 into the visual-tactile roughness mapping model to output the predicted surface spatial wavelength parameter.
[0022] The second visual feature value calculated in real time in step S2 is substituted into the visual-tactile softness mapping model to output the predicted softness coefficient.
[0023] Preferably, the finger electrode array adopts a vertical elliptical multilayer electrode array, including three layers (inner, middle, and outer) with multiple electrode points; the driving logic of the microcurrent haptic rendering controller for the finger electrode array in step S4 includes modulation in two dimensions:
[0024] S41. During the contact pressing stage, the finger electrode array is driven by electrode diffusion control logic.
[0025] S42. During the sliding interaction phase, frequency modulation logic is used to drive the finger electrode array.
[0026] Furthermore, in step S41, when virtual contact is detected, the electrode activation strategy is determined based on the softness coefficient and the preset softness threshold.
[0027] When the softness coefficient is less than the preset softness threshold, the virtual object is determined to be hard, and only the inner central electrode of the finger electrode array is activated for high-intensity stimulation. Moreover, as the pressing depth increases, the activation area does not spread to the outer layer of the finger electrode array.
[0028] When the softness coefficient is greater than or equal to the preset softness threshold, the virtual object is determined to be soft, and the activated area is rapidly expanded from the inner layer of the finger electrode array to the middle and outer layers of the finger electrode array.
[0029] Furthermore, in step S42, the tangential sliding speed of the virtual hand along the surface of the virtual object is acquired in real time. The fundamental stimulation frequency is calculated based on the generated surface spatial wavelength parameters. :
[0030] ;
[0031] Introducing an upper limit threshold for stimulation frequency The final stimulation frequency is calculated using the following formula:
[0032] ;
[0033] Where k is a constant coefficient; the calculated final stimulation frequency generates the corresponding stimulation signal, driving all activated electrodes.
[0034] On the other hand, the automatic generation system for microcurrent tactile parameters based on microscopic geometric feature mapping used in this embodiment of the invention is used to implement the above-mentioned automatic generation method for microcurrent tactile parameters. The system includes:
[0035] The visual data acquisition module is used to extract surface texture images of interactive objects in a virtual scene;
[0036] The feature extraction module is used to calculate the statistical texture features of the image, including a first visual feature value that characterizes the degree of texture undulation and a second visual feature value that characterizes the sharpness of texture edges.
[0037] The parameter mapping calculation module is used to store the preset visual-tactile roughness mapping model and visual-tactile softness mapping model, and convert the first visual feature value and the second visual feature value into physical control parameters.
[0038] The haptic rendering driver module is used to receive calculated physical control parameters and send electrical stimulation control commands to the finger electrode array.
[0039] Compared with the prior art, the beneficial effects achieved by the present invention include:
[0040] 1. Strong generalization ability: No need to physically sample each virtual object; haptic feedback can be generated as long as there is a visual texture, realizing "what you see is what you touch".
[0041] 2. High computational efficiency: The computational cost of algorithms based on statistical features is much smaller than that of deep learning models, making them suitable for running on mobile VR devices.
[0042] 3. Good adaptability: The generated parameters directly correspond to the physical control variables of electrotactile sensing, which can perfectly drive the bionic electrode array for fine rendering. Attached Figure Description
[0043] Figure 1 This is a flowchart of the automatic generation method of microcurrent tactile parameters based on micro-geometric feature mapping in an embodiment of the present invention;
[0044] Figure 2 This is a data flow diagram of the mapping model construction process and the tactile parameter generation process in this embodiment of the invention;
[0045] Figure 3 This is a schematic diagram of the electrode array activation logic in the hard contact mode of this invention embodiment;
[0046] Figure 4 This is a schematic diagram of the electrode array activation logic in the soft contact mode of this invention embodiment;
[0047] Figure 5 This is a schematic diagram of the electrode array activation logic in the sliding mode in an embodiment of the present invention;
[0048] Figure 6 This is a diagram of the overall architecture of the system in this embodiment of the invention. Detailed Implementation
[0049] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.
[0050] Example
[0051] This embodiment provides a method for automatically generating microcurrent tactile parameters based on microscopic geometric feature mapping, such as... Figure 1 As shown, the specific steps include:
[0052] S1. Real-time acquisition of microscopic geometric image data of the surface of the virtual object currently in contact with the user in the virtual reality scene.
[0053] The system uses the raycasting module of a VR engine (such as Unity) to monitor the contact and collision between the virtual hand and virtual objects in real time. When the virtual hand makes contact with a virtual object, it samples the normal map or height map at the contact point. Unlike ordinary RGB textures, this type of image data directly contains the microscopic geometric information (bump and concavity) of the object's surface, providing a data foundation for the subsequent deduction of tactile parameters.
[0054] S2. Explicitly extract features from the micro-geometric image data and calculate the first visual feature value representing the degree of texture undulation and the second visual feature value representing the sharpness of texture edges.
[0055] This embodiment is based on the principle of cross-modal consistency in human perception. In human daily experience, the microscopic geometry of an object's surface is highly correlated with its mechanical properties: hard materials (such as wood and stone) are not easily deformed, so their microscopic edges are usually sharp, resulting in larger image gradient values; while soft materials (such as fabrics and leather) are prone to curling or deformation, with relatively smooth edges, resulting in smaller image gradient values. This embodiment utilizes this strong correlation between visual cues and tactile perception, using the mean edge gradient as a proxy variable to map the softness of the material. There is a strong negative correlation between the sharpness of the microscopic geometric edges of an object's surface and its macroscopic physical softness; hard objects have sharp edges, large gradients, and small softness (S).
[0056] In this step, a microscopic image region (e.g., 64×64 pixels) around the contact point between the virtual hand and the virtual object is captured by the host computer processing terminal, and an explicit image processing algorithm is used for real-time calculation:
[0057] The first visual feature value, used to characterize the roughness of the texture, is denoted as... The contrast of the image is calculated using the Gray-Level Co-occurrence Matrix (GLCM) algorithm. The specific calculation formula is as follows:
[0058] ;
[0059] in In the gray-level co-occurrence matrix The probability value of location. Physically, the coarser the texture, the greater the micro-geometric fluctuations, and the higher the calculated first visual feature value. The larger the value.
[0060] The second visual feature value, used to characterize the hardness of the texture, is denoted as... The mean gradient of the image edges is obtained by using the Sobel operator. Specifically, the gradients in the horizontal and vertical directions are calculated first, and then the mean of the gradient magnitude (the maximum rate of change of the function value along the gradient direction) is taken as the mean. In a physical sense, hard objects have sharp edges and drastic pixel gradient changes, resulting in calculated second visual feature values. The value is relatively large.
[0061] S3. Input the first visual feature value into the preset visual-tactile roughness mapping model to calculate the surface spatial wavelength parameter of the virtual object; input the second visual feature value into the preset visual-tactile softness mapping model to calculate the softness coefficient of the virtual object.
[0062] This step is the core of achieving cross-modal conversion from vision to touch, and its execution relies on a pre-built mapping model. In this embodiment, the visual-tactile roughness mapping model and the visual-tactile softness mapping model are collectively referred to as the visual-tactile feature mapping model; for example... Figure 2 As shown, based on the data flow, this step includes two logical stages: mapping model construction and haptic parameter generation.
[0063] S31. Construct a mapping model.
[0064] To establish the mathematical relationship between visual features and physical tactile parameters, this embodiment performed calibration work before system operation, including data acquisition, feature and truth value extraction, and linear regression fitting. Figure 2 The left side is shown below:
[0065] Data acquisition: Twenty representative real material samples (including cotton, sandpaper, wood, leather, plastic, etc.) were selected, and their microscopic three-dimensional morphology data were collected by pressing the surface of each real material sample with a high-precision visual tactile sensor (such as Gelsight).
[0066] Feature and Truth Value Extraction: Each set of acquired microscopic 3D topographic data undergoes dual processing. Firstly, visual feature values are calculated, converting the topographic data into a grayscale texture map. The first visual feature value (contrast) and the second visual feature value (gradient mean) are extracted using the same algorithm as in step S2. Secondly, physical truth values are calculated. Frequency domain analysis (2D-FFT) is performed on the depth data to extract the frequency peak with the highest power spectral density, and the true average spatial wavelength is calculated. Simultaneously, the normalized true softness coefficient was determined based on the elastomer deformation depth. .
[0067] Linear regression fitting: Extract the first and second visual feature values corresponding to the real material samples, and use the two visual feature values as independent variables X and the physical truth value as the dependent variable Y, respectively, to perform linear regression analysis, thereby establishing the mapping function relationship and constructing the visual-tactile roughness mapping model and the visual-tactile softness mapping model. Specific configuration is as follows:
[0068] Roughness mapping model: configured such that the larger the image contrast value, the smaller the generated surface spatial wavelength parameter λ;
[0069] Softness mapping model: The larger the mean gradient of the image edges, the smaller the generated softness coefficient S.
[0070] S32, Generate tactile parameters.
[0071] like Figure 2 As shown on the right side, during online interaction, the system directly calls the constructed visual-tactile roughness mapping model and visual-tactile softness mapping model:
[0072] The first visual feature value calculated in real time in step S2 Substitute the data into the visual-tactile roughness mapping model and output the predicted surface spatial wavelength parameters. ;
[0073] The second visual feature value calculated in real time in step S2 Substitute the data into the visual-tactile softness mapping model to output the predicted softness coefficient. .
[0074] In this embodiment, linear regression analysis can be performed using the formula Y=kX+b, and the two visual feature values calculated in real time can be collectively referred to as... Surface spatial wavelength parameters and softness coefficient These can be collectively referred to as prediction parameters.
[0075] S4. Transmit the surface spatial wavelength parameters and softness coefficient generated in step S3 to the microcurrent tactile rendering controller to drive the finger electrode array to generate corresponding tactile feedback.
[0076] The microcurrent tactile rendering controller converts the generated physical parameters (i.e., surface spatial wavelength parameters and softness coefficient) into electrical control signals, driving the finger electrode array to generate tactile feedback.
[0077] In this embodiment, the finger electrode array adopts a vertical elliptical multilayer electrode array, including three layers (inner, middle, and outer) with a total of 13 electrode points. Figures 3-5 The diagram illustrates the electrode array activation logic under different tactile parameters, where blank electrodes indicate inactivity, solid black electrodes indicate connection to a cathode stimulation signal (pressure sensation), and black wavy stripe electrodes indicate connection to an anode stimulation signal (vibration sensation). The specific driving logic includes modulation in the following two dimensions:
[0078] S41. During the contact pressing stage, the finger electrode array is driven by electrode diffusion control logic.
[0079] When virtual contact is detected, based on the softness coefficient The electrode activation strategy is determined by a preset softness threshold. In this embodiment, the generated softness coefficient... Normalization is performed to make its value range 0-1, and a softness threshold is preset. =0.3.
[0080] when At times, such as Figure 3 As shown, the system determines that the virtual object has a hard texture, and only activates the inner central electrode of the finger electrode array for high-intensity stimulation. Furthermore, as the pressure depth increases, the activation area does not spread to the outer layer of the finger electrode array, thus simulating the stress concentration sensation when contacting a hard object. This is the mode of pressing in contact with a hard object.
[0081] when At times, such as Figure 4 As shown, the system determines that the virtual object has a relatively soft texture, and controls the activation area to rapidly expand from the inner layer of the finger electrode array to the middle and outer layers of the finger electrode array. Furthermore, the diffusion rate of the activation area is related to the generated softness coefficient. The ratio is directly proportional; that is, the greater the softness coefficient, the shorter the time required for the activation area to diffuse to the outer layer of the finger electrode array, simulating the sensation of the finger quickly sinking into the soft material. This is the mode of pressing in contact with a soft material.
[0082] S42. During the sliding interaction phase, frequency modulation logic is used to drive the finger electrode array.
[0083] The system acquires the tangential sliding speed of the virtual hand along the surface of the virtual object in real time. Based on the generated surface spatial wavelength parameters Calculate the basic stimulus frequency :
[0084] ;
[0085] Considering human perception threshold and hardware security, this embodiment introduces an upper limit threshold for stimulation frequency. This upper limit threshold can be set to 200Hz-300Hz. The formula for calculating the final stimulation frequency is:
[0086] ;
[0087] Where k is a constant coefficient, and in this embodiment, k=1. The calculated final stimulation frequency generates a corresponding stimulation signal, driving all activated electrodes to produce a dynamic vibration sensation at the user's fingertip that varies with the texture roughness. For example... Figure 5 As shown.
[0088] Based on the same inventive concept, this embodiment also provides an automatic generation system for microcurrent tactile parameters based on microscopic geometric feature mapping, such as... Figure 6 As shown. The microcurrent tactile parameter automatic generation system of this embodiment mainly includes a virtual VR interaction terminal (such as Unity, a head-mounted display), a host computer processing terminal, and a slave computer control terminal in its hardware architecture; wherein, the host computer processing terminal includes a feature extraction module and a parameter calculation module, used for feature extraction and parameter calculation; the slave computer control terminal includes a microcurrent tactile rendering controller and a finger electrode array. From the perspective of data acquisition and processing, the hardware architecture of the system of this embodiment can be divided into the following modules:
[0089] The visual data acquisition module is used to extract surface texture images of interactive objects in the virtual scene. Located in the virtual VR interaction layer, this module acquires the normal map or height map of the target object's surface in the virtual scene. Unlike ordinary RGB photos, normal maps and height maps include the microscopic geometric information of the virtual object's surface, i.e., the surface unevenness, which is highly consistent with the physical source of tactile perception.
[0090] The feature extraction module is used to calculate the statistical texture features of the image, including a first visual feature value representing the degree of texture undulation and a second visual feature value representing the sharpness of texture edges. This module is located on the host computer processing end and belongs to the feature layer. It uses explicit image processing algorithms (rather than neural networks) to extract features, calculates the image contrast using the Gray-Level Co-occurrence Matrix (GLCM) to characterize surface roughness, and uses a gradient algorithm to calculate the sharpness of image edges to characterize surface hardness, thereby achieving texture feature extraction.
[0091] The parameter mapping calculation module stores preset visual-tactile roughness mapping models and visual-tactile softness mapping models, and converts the first visual feature value and the second visual feature value into physical control parameters. This module is located on the host computer processing end and belongs to the mapping layer. Based on the pre-built Gelsight visual-tactile database, it establishes a regression model between visual features and physical parameters, i.e., a mapping model, which converts image features into surface spatial wavelength and softness coefficient required for microcurrent rendering in real time.
[0092] The haptic rendering driver module receives calculated physical control parameters and sends electrical stimulation control commands to the finger electrode array. Located at the lower-level control unit and belonging to the output layer, this module substitutes the generated parameters into the micro-current frequency control formula to achieve real-time realistic rendering of unknown textures.
[0093] As can be seen from the above, in the technical solution of this invention, the surface micro-geometric image data of the target object in the virtual interactive scene is first acquired; image processing algorithms are used to extract a first visual feature value representing the degree of texture undulation and a second visual feature value representing the gradient of texture edges; based on a pre-constructed visual-tactile roughness and softness mapping model, the visual feature values are converted into roughness parameters and softness parameters required for electro-tactile rendering, respectively; finally, the generated parameters are input into a micro-current tactile rendering controller to drive the finger electrode array. This invention can deduce tactile parameters in real time from visual features for virtual textures without pre-stored physical data, and has the advantages of low computational load and strong physical interpretability, solving the problems of high computational resource consumption and difficulty in adapting to electro-tactile encoding mechanisms in existing deep learning generation methods.
[0094] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.
Claims
1. A method for automatically generating microcurrent tactile parameters based on microscopic geometric feature mapping, characterized in that, Includes the following steps: S1. Real-time acquisition of microscopic geometric image data of the surface of the virtual object currently in contact with the user in the virtual reality scene; S2. Explicitly extract features from the micro-geometric image data and calculate the first visual feature value representing the degree of texture undulation and the second visual feature value representing the sharpness of texture edges. S3. Input the first visual feature value into the preset visual-tactile roughness mapping model to calculate the surface spatial wavelength parameters of the virtual object. The second visual feature value is input into the preset visual-tactile softness mapping model to calculate the softness coefficient of the virtual object. S4. Transmit the surface spatial wavelength parameters and softness coefficient to the microcurrent tactile rendering controller to drive the finger electrode array to generate corresponding tactile feedback.
2. The method for automatically generating microcurrent tactile parameters according to claim 1, characterized in that, Step S3 includes: S31. Constructing a mapping model: A variety of representative real material samples were selected, and the surface of each real material sample was pressed using a visual-tactile sensor to collect its microscopic three-dimensional morphology data. Each set of collected microscopic three-dimensional morphology data is processed in two ways. On the one hand, visual feature values are calculated, the morphology data is converted into a grayscale texture map, and the first visual feature value representing the degree of texture undulation and the second visual feature value representing the sharpness of texture edges are extracted. On the other hand, physical true values are calculated, the depth data is analyzed in the frequency domain, the frequency peak with the largest power spectral density is extracted, the true average spatial wavelength is calculated, and the normalized true softness coefficient is determined based on the deformation depth of the elastic body. Linear regression analysis was performed using the first and second visual feature values corresponding to real material samples as independent variables X and the physical truth value as the dependent variable Y, respectively, to construct visual-tactile roughness mapping models and visual-tactile softness mapping models. S32. Generate tactile parameters: Substitute the first visual feature value calculated in real time in step S2 into the visual-tactile roughness mapping model to output the predicted surface spatial wavelength parameter. The second visual feature value calculated in real time in step S2 is substituted into the visual-tactile softness mapping model to output the predicted softness coefficient.
3. The method for automatically generating microcurrent tactile parameters according to claim 2, characterized in that, In step S31, the roughness mapping model is configured such that the larger the image contrast value, the smaller the generated surface spatial wavelength parameter; and the softness mapping model is configured such that the larger the mean gradient of the image edge, the smaller the generated softness coefficient.
4. The method for automatically generating microcurrent tactile parameters according to claim 1, characterized in that, The finger electrode array adopts a vertical elliptical multilayer electrode array, including three layers (inner, middle, and outer) with multiple electrode points; the driving logic of the microcurrent haptic rendering controller for the finger electrode array in step S4 includes modulation in two dimensions: S41. During the contact pressing stage, the finger electrode array is driven by electrode diffusion control logic. S42. During the sliding interaction phase, frequency modulation logic is used to drive the finger electrode array.
5. The method for automatically generating microcurrent tactile parameters according to claim 4, characterized in that, In step S41, when virtual contact is detected, the electrode activation strategy is determined based on the softness coefficient and the preset softness threshold. When the softness coefficient is less than the preset softness threshold, the virtual object is determined to be hard, and only the inner center electrode of the finger electrode array is activated for high-intensity stimulation. Moreover, as the pressing depth increases, the activation area does not spread to the outer layer of the finger electrode array. When the softness coefficient is greater than or equal to the preset softness threshold, the virtual object is determined to be soft, and the activated area is rapidly expanded from the inner layer of the finger electrode array to the middle and outer layers of the finger electrode array.
6. The method for automatically generating microcurrent tactile parameters according to claim 5, characterized in that, When the softness coefficient is greater than or equal to the preset softness threshold, the diffusion rate of the activated region is proportional to the generated softness coefficient.
7. The method for automatically generating microcurrent tactile parameters according to claim 4, characterized in that, In step S42, the tangential sliding speed of the virtual hand along the surface of the virtual object is acquired in real time. The fundamental stimulation frequency is calculated based on the generated surface spatial wavelength parameters. : ； Introducing an upper limit threshold for stimulation frequency The final stimulation frequency is calculated using the following formula: ； Where k is a constant coefficient; the calculated final stimulation frequency generates the corresponding stimulation signal, driving all activated electrodes.
8. The method for automatically generating microcurrent tactile parameters according to claim 1, characterized in that, Step S1 monitors the contact and collision between the virtual hand and the virtual object in real time. When the virtual hand comes into contact with the virtual object, the normal map or height map at the contact point is sampled.
9. The method for automatically generating microcurrent tactile parameters according to claim 1, characterized in that, Step S2 uses the gray-level co-occurrence matrix algorithm to calculate the contrast of the image to obtain the first visual feature value, and uses the Sobel operator to calculate the mean gradient of the image edges to obtain the second visual feature value.
10. An automatic generation system for microcurrent tactile parameters based on microscopic geometric feature mapping, used to implement the method described in any one of claims 1-9, characterized in that, The system includes: The visual data acquisition module is used to extract surface texture images of interactive objects in a virtual scene; The feature extraction module is used to calculate the statistical texture features of the image, including a first visual feature value that characterizes the degree of texture undulation and a second visual feature value that characterizes the sharpness of texture edges. The parameter mapping calculation module is used to store the preset visual-tactile roughness mapping model and visual-tactile softness mapping model, and convert the first visual feature value and the second visual feature value into physical control parameters. The haptic rendering driver module is used to receive calculated physical control parameters and send electrical stimulation control commands to the finger electrode array.
Citation Information
Patent Citations
Virtual friction tactile rendering method and device based on micro-current stimulation
CN119292469A