Thermal distribution monitoring of the container data center by a fast infrared image fusion technique

This paper develops a thermal distribution monitoring (TDM) method for a container data center (CDC). First, the image is divided into hot and non-hot regions and their corner and edge features are extracted. Then, the regions that are blended by fast infrared image fusion (FIIF) technique, named as hot spot fusion; otherwise, non-hot spot fusion. In the experiments, the study tests seven benchmarks from Octets, SPi Corporation, and Kimura’s literature. The results show the proposed method achieves good performance for recognizing physical position of hot spots. We also test five actual Inventec CDC images which consist of angles from 0 to 45 degree in different field of view (FOV). The method is validated to be applicable to the container data center or ordinary data center with high density of equipment. The experimental results show that the image fusion is efficient to find hot spot position for the management information system (MIS).