An Efficient and Robust Image Quality Assessment Method for On Line Weld Pool Monitoring

Image quality characterizes the perceived image degradation and is of great importance in many industrial applications which use image quality assessment as their first step to select images. The accuracy and efficiency of the image quality assessment method is critical in many real time applications, e.g. on-line monitoring. In this paper, an efficient and effective image quality assessment method is proposed for on line weld pool monitoring. The proposed method assesses the image quality with a better stability compared to other evaluated state of art methods. In addition, the computation speed of the proposed method is also fast and suitable for on line applications.