This paper presents a template-based vision system to detect and classify the nonuniformaties that appear on the semiconductor wafer surfaces. Design goals include detection of flaws and correlation of defect features based on semiconductor industry expert’s knowledge. The die pattern is generated and kept as the reference beforehand from the experts in the semiconductor industry. The system is...

This paper takes the large band sawing machine GB42500H as the research object, and establishes the dynamics models of its various parts. Respectively, the relative research is done from the theoretical modal analysis and the dynamic performance testing. According to the analysis results, the structure or geometry improvement scheme of the parts is proposed. The paper also studies preliminarily...

Abstract Sawing variation (SV), the degree of deviation from a specified target lumber size, is an unavoidable component sawing process. SV influenced by several factors such as machine, material, set works, feed and cutting parameters. To account for these resulting in deviations desired thickness cut must be increased that only minimal number boards less than thickness. Thus, greater amount S...

The planning of how hardwood logs can be sawn to improve recovery of high-value lumber has always been hampered by the limited information provided by external defects, and whatever internal defects are eventually revealed on the cut log faces by the sawing pattern. With expanded markets, lower-quality logs, increased competition from non-wood products, social pressures to manage public lands f...

The Canadian wood industry use sawing simulators to digitally break a log into basket of lumbers. However, those tend be computationally intensive. In some cases, this renders them impractical as decision support tools. Such case is the problem dispatching large volume several sawmills in order maximise total yield dollars. Fast machine learning metamodels were recently proposed address issue. ...