A General Method for Sensor Planning in Multi-Sensor Systems: Extension to Random Occlusion