This paper studies the dynamic coverage control problem for cooperative region reconnaissance where a group of agents are required to reconnoitre a given region. The main challenge of this problem is that the sensing region of each agent is an ellipse. This modeling results in asymmetric(directed) interactions among agents. First, the region reconnaissance is formulated as a coverage problem, where each point in the given region should be surveyed until a preset level is achieved. Then, a coverage control law is designed that minimizes coverage performance index by finite switches between nominal control laws and perturbation control law. Finally, numerical simulations are provided to indicate the efficiency of the proposed control law.