Soft robotic grippers facilitate contact-rich manipulation, including robust grasping of varied objects. Yet the beneficial compliance of a soft gripper also results in significant deformation that can make precision manipulation challenging. We present visual pressure estimation & control (VPEC), a method that uses a single RGB image of an unmodified soft gripper from an external camera to directly infer pressure applied to the world by the gripper. We present inference results for a pneumatic gripper and a tendon-actuated gripper making contact with a flat surface. We also show that VPEC enables precision manipulation via closed-loop control of inferred pressure. We present results for a mobile manipulator (Stretch RE1 from Hello Robot) using visual servoing to do the following: achieve target pressures when making contact; follow a spatial pressure trajectory; and grasp small objects, including a microSD card, a washer, a penny, and a pill. Overall, our results show that VPEC enables grippers with high compliance to perform precision manipulation.

翻译：软机器人抓手器可以促进接触丰富的操纵,包括强力捕捉各种物体。 但是,软抓手的有利合规性也导致显著变形,使得精密操作具有挑战性。 我们展示了视觉压力估计和控制(VPEC),这种方法使用外部摄像头未变软抓手的单一 RGB 图像,直接推断握手对世界的压力。 我们展示了气压抓手的推论结果和与平坦表面接触的有倾向活性握手的推论结果。 我们还显示, VPEC 能够通过闭路控制推断压力来进行精密操作。 我们展示了移动操纵器(Hello Robot的Strach RE1)的结果,我们使用视觉推力进行以下操作:在接触时达到目标压力; 跟踪空间压力轨迹; 抓住小物体,包括微SD卡、 异体、 一便、 一便士和一药丸。 我们的结果显示, VPEC 能够让高度合规的抓手来进行精确操纵。