This paper studies the optimization model and area coverage enhancement issues in wireless multimedia sensor networks (WMSNs). First, based on the directional perception characteristics of the multimedia sensor nodes, a modified sector-ring sensing model is provided to explore issues in video acceptable sharpness and regional coverage enhancement in WMSNs. Next, a mixed virtual potential field is proposed to enhance the area coverage in WMSNs through centroids movement under the effect of gravity and repulsion from the virtual potential field. To overcome the deficiency of local minimum and its reduced optimizing effect of coverage with traditional virtual potential field, the proposed algorithm improves the function of repulsion force by taking the common coverage rate of neighbor nodes into consideration to assist the adjustment of perception directions of the nodes. It also transmits border-dynamic repulsion force to sensor nodes after the iteration is completed to further optimize utilization of boundary nodes. Finally, a series of experiments and related simulations are performed to demonstrate the effectiveness of the algorithm.