Summary: A visual trajectory planning method was proposed based on hybrid artificial potential field for a nonholonomic mobile robot equipped with a monocular camera in consideration of field-of-view (FOV) constraints. A hybrid model was designed based on the concept of conventional artificial potential field. The model consisted of two parts. One part only affects the angular velocity and involves the terms for target deflection and field of view constraints, and the other part affects both the linear velocity and the angular velocity. The potential field for field of view constraints was defined based on the feature point coordinates in the image space. The potential fields for target deflection and the relative pose were calculated based on the scaled three-dimensional reconstruction results obtained from the decomposition of homography. The robot can be regulated to the desired pose with the targets kept in the field of view under the hybrid artificial potential field. Simulation results were provided in some representative circumstances to show the effectiveness of the proposed method.