Control/Display Mapping And Pointing Precison Analysis Based On Visual Gesture Interaction

With the features of flexible operation,open space and comfortable interaction,gesture interaction on machine-vision is becoming the mainstream way of realizing the natural interaction between human and computers.Due to the lack of tactile feedback from the input device,the user needs to rely more on visual feedback from the screen to achieve control of the cursor.However,due to the principle of the camera imaging,control/display gain of gesture interaction for machine vision will often unexpectedly change because the user's movement.We discuss when user's locations relative to display space change,how the instability of CD gain affect user's object performance by Fitts click comparative experimental tests.In order to reduce the instability of CD gain,this paper presents an adaptive adjustment method of the control/display gain is proposed according to user's interaction features.According to ergonomics,we design multi-directional Fitts Click tests,to study the influence of moving direction and the CD gain 's transfer function on the pointing precison.The results show that setting a appropriate transfer function is helpful to improve the pointing precison and the performance of interaction.In interaction system design,the size of interface icon and the Corrective Phase of the transfer function should be referenced to each other,and the Ballistic Phase of the transfer function should refer to the screen resolution.Users are more accustomed to moving their hands in the horizontal direction in machine-vision interaction,so the task of moving in other directions should be reduced.It is also proved by experiments that the proposed transfer function for visual gesture interaction is an ideal mapping method,which is beneficial to maintain the stability of interactive performance.The research in this paper can provide effective guidance to improve the user experience in non-contact gesture human-computer interaction.These conclusions can provide important reference value to the design and applications of user-centered gesture-based interaction.