Hossein Boroumand Noghabi - Department of Computer Engineering Islamic Azad University of Zanjan, Iran
Fatemeh Abdollahi - Department of Computer Engineering Islamic Azad University of Zanjan, Iran

In this article we are want to introduce a new method to detect the fault in mechanical equipments by using thermal images. Infrared thermography will be used in order to supervise the situations in various industrial backgrounds. It is being used in many different fields including physics, electrical engineering , metalogy and engineering , mechanical engineering, biomedical engineering and structural engineering.these methods are used for helping to anticipate the faults or possible failures in machines.their accurate detection is a key issue in computer aided detection schame.to improve the performance of detection , we propose a boosting based twin support vector machine to detect the faults in mechanical equipment. The algorithm is composed of five modules: taking a picture, normalizing, segmentation, feature extraction component, and the boosting TWSVM module. And for distinguishing the heat condition of the mechanical parts (correct/faulty) we used Boosting TWSVM classification with NETA standard, then tested it on 160 thermal images of mechanical installations.

Boosting TWSVM,TWSVM, thermography, mechanical equipment, image processing