Server thermal fault monitoring and diagnosing method based on infrared images
作者： Liu Hang , Xie Ting
DOI:
关键词: Grayscale 、 Dimensionality reduction 、 Support vector machine 、 Artificial intelligence 、 Data center 、 Debugging 、 Infrared 、 Computer science 、 Entropy (information theory) 、 Image registration 、 Computer vision
摘要: The invention discloses a server thermal fault monitoring and diagnosing method based on infrared images. comprises acquiring images visible-light of in different states; carrying out tilt correction processing the each state by means standardization image registration, determining region-of-interest including manual segmentation method; converting determined into corresponding gray scale image, extracting an entropy feature from image; dimension reduction obtained global local row-column mean value PCA wherein main component after is used for training support vector machine classifier; to be detected, type trained SVM model. According invention, management efficiency personnel data center greatly improved, engineering debugging system maintenance are facilitated.
lens.org 参考文章(3)
Antony James Holliday, Thermal imager and transmitter ,(2012)
Tang Jinhong, Li Na, Liu Hang, Feng Zhongzheng, Wei Gang, Deng Xudong, Wu Chen, Infrared image automatic fault identifying method for high-voltage equipment ,(2014)
Duan Qichang, Duan Pan, Huang Xiaogang, Hu Bei, Chen Hongguang, Mao Mingxuan, Wang Hongshou, Chen Delin, Equipment fault diagnosis method and system of electric power system ,(2014)