Tactile Rendering And Performance Analysis Of Mobile Terminal Gesture Interaction

By utilizing touch screens and a graphical user interface which can be manipulated directly,gesture-based interaction with mobile devices offers a simpler and more natural means of human-machine interaction than those utilizing a mouse and keyboard.However,in the process of touch gesture interaction,the user generally cannot rely on familiar haptic cues to manipulate digital content,which could lead to a visual or auditory overload.Electrostatic tactile feedback proposes a possible solution to such a problem.By controlling electrostatic friction between users' fingers and a touch surface,it enables virtual elements to reproduce physical properties such as roughness and texture,thus improving the user experience.Despite this,most existing tactile rendering methods for touch gesture interactions only embrace certain types of gestures in the context of particular scenarios,in which there is a lack of generality in the methods for tactile rendering as well as development tools for electrostatic tactile feedback.This limits the application of electrostatic tactile feedback for interactions performed on mobile devices.Given the above problems,this paper attempts to build a variety of electrostatic friction patterns with a certain degree of generality.It also provides specific electrostatic tactile rendering processes combined with multiple types of scenarios,which can serve as a guide for designing and implementing electrostatic tactile feedback.The main contents consist of building reusable electrostatic friction patterns,analyzing physical parameters which could affect perceived differences among friction patterns,choosing basic gestures suitable for electrostatic tactile feedbacks,extracting feature parameters,presenting electrostatic tactile rendering procedure in the context of typical gesture-based interaction scenarios,and investigating the influence of friction patterns on interaction performance.The contributions of this paper are as follows:(1)Ten reusable electrostatic friction patterns are built,and the impact of three physical parameters on the distinction of the friction patterns is validated,namely the continuity of the force profiles,the location of the feedback force,and the duty cycle of the feedback force area.Ten electrostatic friction patterns are specified concerning four spatial distribution features,including the location of the feedback force,the duty cycle of the feedback force area,the continuity of the force profiles,and the shape of the force profiles.The mapping relationship between the perceived electrostatic friction force and the characteristic parameters in the space domain is established.An experiment is conducted to investigate whether the user can effectively discriminate friction patterns adjusted by the same parameter and then assign them to the correct group.The results show that there is a clear difference between the three types of friction patterns: "Bump","Slope" and "Flat".The continuity of the force profiles,the location of the feedback force,and the duty cycle of the feedback force area have a direct influence on the perception of friction patterns,while the perceived difference between friction patterns modulated by the shape of the force profiles or the driving signal waveform is ambiguous.Additionally,users have different preferences for friction patterns,therefore,multiple friction patterns provide a wider space for them to choose a suitable one.(2)Five gestures used for dragging,pinching,etc.,are selected as the basic gestures for interactive operations on mobile devices,and methods for extracting features from them are given.Tactile rendering processes for three scenarios,including object alignment in grid-based layout,to-do list,and image zoom,are established.Five basic gestures are selected as basic gestures,including drag,flick,spread,pinch,and rotate.This is due to the fact that electrostatic tactile feedback can only be perceived if there is a relative motion between the finger and the surface.Characteristic parameters are extracted,such as sliding displacement,velocity,and distance between two fingers.The mapping relationship between characteristic parameters and the driving signal amplitude is established.Based on the state of gesture operations in the context of three different scenarios,such as object alignment in a grid-based layout,to-do list,and image zoom,an electrostatic friction pattern is determined.In response to the characteristic parameters extracted from gesture operations,electrostatic tactile rendering is completed according to the selected friction pattern and the mapping between the perceived friction force and the applied amplitude.Additionally,a generic procedure for the tactile rendering of gestures is explored based on ten friction patterns.(3)Experiments with drag gesture tasks in three representative friction pattern conditions are carried out to examine the impact of friction patterns with different subjective evaluations on completion time and success rate.Two friction patterns,which are respectively selected as the favorite friction pattern and the most disliked one by most users,are selected as variable values in this experiment.No tactile feedback was used as a baseline control.The completion time and success rate are measured based on target acquisition for dragging tasks performed by twelve participants.One-way repeated measures ANOVA is used to investigate the statistical effects of friction patterns on the efficiency and accuracy of drag gestures.The experimental results show that personal preference for friction patterns doesn't have a significant influence on interaction performance.The goodness of fit to Fitts' law is near to 0.9 for all conditions.Although the completion time of the dragging task performed by participants under the most preferred friction pattern(vibrating plane-sawtooth)increases slowly with the increasing index of difficulty,the difference with the least preferred friction pattern and no tactile feedback remains small.