In this paper, PSO-OMRC model is proposed for heating ventilation and air conditioning (HVAC) system. The PSO-OMRC predicts SPT in HVAC for energy saving and thermal comfort of persons in room. PSO-OMRC used data collected from temperature, humidity, PIR, and thermal camera sensors in HVAC installed rooms for SPT prediction. In existing HVAC systems, SPT is constant and consumes more energy. Moreover, SPT prediction using traditional machine learning algorithms considers various parameters such as room temperature, humidity and person count in room. Existing algorithms considers the above parameters for energy saving in HVAC and thermal comfort for persons, whereas other parameters such as room walls and ceiling temperature, human body temperature need to be considered for SPT prediction and for energy saving in HVAC. The PSO-OMRC model predicts the SPT based on number of occupants and microclimatic conditions such as environment temperature, humidity, room wall temperature and ceiling temperature. From experimental results, the utilization of the thermal image occupant model provides significant energy savings in the PSO-OMRC model. In nano coated room, PSO-OMRC model is evaluated with the implementation of a thermal image occupant model to improve energy savings, resulting in approximately 25.2% of energy savings while maintaining the level of thermal comfort.