In this paper a novel, cost-effective and robust occupancy monitoring system is presented, which is based on a fuzzy confidence voting algorithm utilizing spatial height histograms. Spatial height histograms are invariant to rotations and translations, providing this way a desirable feature to occupancy measurement systems, and when combined with distance coefficients can fix an occupancy feature vector, which is the main source for the fuzzy confidence voting occupant tracking algorithm. The proposed occupancy extraction system can be efficiently applied to multi-space environments using a privacy preserving multi-camera cloud. Statistics per building, space and occupant can be finally extracted by the system. The experimental results will illustrate its robustness, accuracy and efficiency on occupancy extraction.