Prediction of Strain Rate of Aluminum/Silicon Carbide Using Gray Level Run-Length Matrices  

作　　者：Ahmad E. Eladawi SaadA. A. Sayed Hammad T. Elmetwally Tamer O. Diab 

机构地区：[1]Department of Mechanical Engineering, Benha University, Banha 13511, Egypt [2]Department of Mechanical Design, Helwan University, Helwan 11731, Egypt [3]Department of Mechanical Engineering, Beni-Suef University, Beni-Suef 62 511, Egypt [4]Department of Electrical Engineering, Benha University, Banha 13511, Egypt

出　　处：《Journal of Energy and Power Engineering》2017年第5期355-361,共7页能源与动力工程（美国大卫英文）

摘　　要：The microstructural processes occurring in metals and alloys during hot deformation are： DRX （dynamic recrystallization）, superplastic deformation, dynamic recovery, wedge cracking, void formation, inter-crystalline cracking, prior particle boundary （FFB） cracking, and flow instability processes. Deformation characteristics of materials are interpreted as follows： in the low temperature （T≤0.25 of melting temperature） and high strain rate regime of 10 to 100 s-1, void formation occurs at hard particles and leads to ductile fracture. Many researchers used the currently defined statistical approaches to characterize images and extract useful information from the captured images. For more suitable of specific tasks, some researchers are introducing new texture features. HOS （higher-order statistics） estimate properties of three or more pixels occurring at specific locations relative to each other. GLRLMs （gray level run-length matrices） are popular method of HOS to extract texture features. This paper deals with texture features of GLRLM to predict strain rate values for Aluminum/Silicon Carbide.

关 键 词：Image processing computer vision GLRLMs texture features strain rate. 

分 类 号：TK[动力工程及工程热物理]