Visual Servoing

Visual servoing uses visual information to control robot movements, aiming for precise and autonomous manipulation. Current research emphasizes robust control strategies in challenging scenarios, such as handling occlusions, target tracking in dynamic environments, and adapting to low-rigidity robots, often employing model predictive control, Kalman filtering, deep learning (including neural networks and diffusion models), and various visual feature extraction techniques. This field is crucial for advancing robotics in diverse applications, including autonomous aerial vehicles, on-orbit servicing, minimally invasive surgery, and industrial automation, by enabling more reliable and adaptable robot control.