CN-121999243-A - Micro-current touch parameter automatic generation method and system based on micro-geometric feature mapping

CN121999243ACN 121999243 ACN121999243 ACN 121999243ACN-121999243-A

Abstract

The invention relates to the technical field of virtual reality and man-machine interaction, in particular to a micro-current haptic parameter automatic generation method and system based on micro-geometrical feature mapping. The method comprises the steps of obtaining microscopic geometric image data of the surface of a virtual object currently contacted by a user in real time, carrying out explicit feature extraction on the microscopic geometric image data, calculating to obtain a first visual feature value representing fluctuation degree of texture and a second visual feature value representing sharpness degree of texture edges, respectively inputting the first visual feature value and the second visual feature value into a visual touch roughness and softness mapping model to respectively obtain surface space wavelength parameters and softness coefficients of the virtual object, transmitting the surface space wavelength parameters and the softness coefficients to a micro-current touch rendering controller, and driving an electrode array of a finger to generate corresponding touch feedback. The micro-current stimulation control parameters are automatically generated by analyzing the visual characteristics of the virtual textures, so that the problems of strong dependence on a pre-stored database and poor adaptability to unknown textures in the prior art are solved.

Inventors

JIN JIANXIU
LIU JIALONG
MA YUANJUN
YAO HUILIN
XIONG QIWEI
SHU LIN

Assignees

华南理工大学

Dates

Publication Date: 20260508
Application Date: 20260109

Claims (10)

1. The automatic generation method of the micro-current touch parameters based on the micro-geometrical feature mapping is characterized by comprising the following steps of: s1, acquiring microscopic geometric image data of the surface of a virtual object currently contacted by a user in a virtual reality scene in real time; s2, performing explicit feature extraction on the microscopic geometric image data, and calculating to obtain a first visual feature value representing the fluctuation degree of the texture and a second visual feature value representing the sharpness degree of the texture edge; S3, inputting the first visual characteristic value into a preset visual sense roughness mapping model, and calculating to obtain a surface space wavelength parameter of the virtual object; And S4, transmitting the surface space wavelength parameter and the softness coefficient to a micro-current touch rendering controller, and driving the finger electrode array to generate corresponding touch feedback.
2. The method for automatically generating micro-current haptic parameters as recited in claim 1, wherein step S3 includes: s31, constructing a mapping model: selecting a plurality of representative real material samples, pressing the surface of each real material sample by using a visual touch sensor, and collecting microscopic three-dimensional morphology data of the real material samples; Performing double processing on each group of acquired microscopic three-dimensional morphology data, on one hand, calculating visual characteristic values, converting the morphology data into a gray texture map, and extracting a first visual characteristic value representing the fluctuation degree of texture and a second visual characteristic value representing the sharpness degree of texture edges; on the other hand, calculating a physical true value, carrying out frequency domain analysis on the depth data, extracting a frequency peak value with the maximum power spectrum density, calculating a real average spatial wavelength, and simultaneously determining a normalized real softness coefficient according to the deformation depth of the elastic body; Respectively taking a first visual characteristic value and a second visual characteristic value corresponding to the real material sample as independent variables X, taking a physical true value as dependent variable Y, and carrying out linear regression analysis to respectively construct a visual touch roughness mapping model and a visual touch softness mapping model; S32, generating a touch parameter: Substituting the first visual characteristic value obtained by real-time calculation in the step S2 into a visual tactile roughness mapping model, and outputting predicted surface space wavelength parameters; Substituting the second visual characteristic value obtained in the step S2 through real-time calculation into the visual sense touch softness mapping model, and outputting a predicted softness coefficient.
3. The method according to claim 2, wherein in step S31, the roughness map model is configured to have a larger image contrast value, a smaller surface space wavelength parameter is generated, and the softness map model is configured to have a larger image edge gradient average value, and a smaller softness coefficient is generated.
4. The automatic generation method of micro-current touch parameters according to claim 1, wherein the finger electrode array is a vertical elliptic multilayer electrode array and comprises a plurality of electrode points of inner, middle and outer layers, and the driving logic of the micro-current touch rendering controller to the finger electrode array in the step S4 comprises two-dimensional modulation: S41, driving the finger electrode array by adopting electrode diffusion control logic in a contact pressing stage; s42, driving the finger electrode array by adopting frequency modulation logic in the sliding interaction stage.
5. The method according to claim 4, wherein in step S41, when a virtual contact is detected, an electrode activation strategy is determined according to a softness coefficient and a preset softness threshold; When the softness coefficient is smaller than a preset softness threshold value, judging that the virtual object is harder in texture, only activating the inner layer center electrode of the finger electrode array to perform high-intensity stimulation, and enabling the activation area not to diffuse to the outer layer of the finger electrode array along with the increase of the pressing depth; When the softness coefficient is larger than or equal to a preset softness threshold value, the virtual object is judged to be softer physically, and the activation area is controlled to be rapidly expanded from the inner layer of the finger electrode array to the middle layer and the outer layer of the finger electrode array.
6. The method of claim 5, wherein the diffusion rate of the active region is proportional to the generated softness factor when the softness factor is greater than or equal to a predetermined softness threshold.
7. The method according to claim 4, wherein in step S42, the tangential sliding speed of the virtual hand along the surface of the virtual object is obtained in real time Calculating basic stimulus frequency according to the generated surface space wavelength parameter : ; Introducing an upper threshold of stimulation frequency The calculation formula of the final stimulation frequency is: ; where k is a constant coefficient, and generating a corresponding stimulation signal at the calculated final stimulation frequency to drive all activated electrodes.
8. The automatic generation method of micro-current haptic parameters according to claim 1, wherein step S1 monitors the contact and collision of the virtual hand with the virtual object in real time, and samples a normal map or a height map at a contact point when the virtual hand is in contact with the virtual object.
9. The automatic generation method of micro-current touch parameters according to claim 1, wherein step S2 calculates the contrast of the image by using a gray level co-occurrence matrix algorithm to obtain a first visual characteristic value, and calculates the edge gradient mean value of the image by using a Sobel operator to obtain a second visual characteristic value.
10. A micro-current haptic parameter automatic generation system based on micro-geometric feature mapping for implementing the method of any one of claims 1-9, the system comprising: the visual data acquisition module is used for extracting a surface texture image of an interactive object in the virtual scene; the feature extraction module is used for calculating statistical texture features of the image, and comprises a first visual feature value representing the fluctuation degree of texture and a second visual feature value representing the sharpness degree of texture edges; the parameter mapping calculation module is used for storing a preset visual touch roughness mapping model and a visual touch softness mapping model and converting the first visual characteristic value and the second visual characteristic value into physical control parameters; The touch rendering driving module is used for receiving the calculated physical control parameters and sending an electric stimulation control command to the finger electrode array.

Description

Micro-current touch parameter automatic generation method and system based on micro-geometric feature mapping Technical Field The invention relates to the technical field of virtual reality and man-machine interaction, in particular to an automatic generation method and system of micro-current touch parameters based on micro-geometrical feature mapping. Background With the popularity of meta-universe and Virtual Reality (VR) technologies, haptic feedback technology has become a key to enhancing immersion. The current micro-current haptic rendering technology, such as the invention patent CN119292469A published in 2025, 1 and 10, proposes a rendering model based on three stages of contact-sliding-separation, and establishes a method for modulating the stimulation frequency by using the physical roughness parameter lambda. However, existing data-driven rendering methods typically rely on pre-acquired real object databases. This means that for completely new textures (e.g. random topography in games, science fiction materials) in virtual scenes that are not known or generated by computers, the system cannot learn their roughness and softness due to lack of pre-measured physical parameters, resulting in haptic feedback failure or rollback to a single vibration mode. In the prior art, there are solutions that utilize deep learning networks (e.g., GAN) to directly generate vibration waveforms from images. However, this type of method has obvious drawbacks: 1. The computational power requirement is high, and the complex neural network is difficult to run in real time on a lightweight VR head display or an embedded chip. 2. The mode mismatch generates an acceleration waveform of mechanical vibration, which cannot be directly converted into discrete control parameters (such as pulse width, frequency and electrode activation number) required by micro-current stimulation. 3. Lacking interpretability, the black box model is difficult to precisely adapt to the rule of electrotactile encoding based on physiological properties (e.g., weber's law). Therefore, a method that is lightweight, interpretable, and capable of directly mapping visual features of virtual textures to electrotactile physical control parameters is needed. Disclosure of Invention The invention provides a micro-current touch parameter automatic generation method and system based on micro-geometric feature mapping, which automatically generate micro-current stimulation control parameters by analyzing visual features of virtual textures, and solve the problems of strong dependence on a pre-stored database and poor adaptability to unknown textures in the prior art. On one hand, the micro-current tactile parameter automatic generation method based on micro-geometric feature mapping adopted by the embodiment of the invention comprises the following steps: s1, acquiring microscopic geometric image data of the surface of a virtual object currently contacted by a user in a virtual reality scene in real time; s2, performing explicit feature extraction on the microscopic geometric image data, and calculating to obtain a first visual feature value representing the fluctuation degree of the texture and a second visual feature value representing the sharpness degree of the texture edge; S3, inputting the first visual characteristic value into a preset visual sense roughness mapping model, and calculating to obtain a surface space wavelength parameter of the virtual object; And S4, transmitting the surface space wavelength parameter and the softness coefficient to a micro-current touch rendering controller, and driving the finger electrode array to generate corresponding touch feedback. Preferably, step S3 includes: s31, constructing a mapping model: selecting a plurality of representative real material samples, pressing the surface of each real material sample by using a visual touch sensor, and collecting microscopic three-dimensional morphology data of the real material samples; Performing double processing on each group of acquired microscopic three-dimensional morphology data, on one hand, calculating visual characteristic values, converting the morphology data into a gray texture map, and extracting a first visual characteristic value representing the fluctuation degree of texture and a second visual characteristic value representing the sharpness degree of texture edges; on the other hand, calculating a physical true value, carrying out frequency domain analysis on the depth data, extracting a frequency peak value with the maximum power spectrum density, calculating a real average spatial wavelength, and simultaneously determining a normalized real softness coefficient according to the deformation depth of the elastic body; Respectively taking a first visual characteristic value and a second visual characteristic value corresponding to the real material sample as independent variables X, taking a physical true value as dependent variable Y, and carrying out lin