A Control Barrier Function Approach for Observer-based Visually Safe Pursuit Control with Spherical Obstacles

Pursuing a target in the presence of obstacles requires that an autonomous mobile robot keeps sight of the target in a way robust to the target's unknown behavior. This paper presents visually safe pursuit control, which keeps a target with the unknown motion inside the camera's field of view while preventing occlusion caused by spherical obstacles. Framed as forward invariance of sets in the SE(3) state space, visual safety is ensured by the Control Barrier Functions (CBFs) approach. Concretely, by showing the Input-to-State stability of the vision-based observer that estimates the target's motion, we design safety certificates for visual safety that accommodate uncertainties in the target's motion. This enables us to synthesize a safe controller as a Quadratic Programming problem. Finally, the theoretical results are verified via a simulation of a visual pursuit scenario. Copyright (c) 2023 The Authors.