Determination of Decarburization Depth Base on Deep Learning Methods

In the heat treatment of steel, decarburization is a serious issue that leads to poor wear resistance and low fatigue life. At present, depth was determined using visual estimation by human eye, software through traditional image analysis. Therefore, analysis remains limited in experts algorithms. Artificial intelligence general-purpose technology has multitude applications. This paper uses concept deep learning propose layer detector (DLD) can determine decarburized layers. DLD system boasts high performance, real-time, learning, computation costs. addition, we used several kinds layers images compare proposed method with other network architectures. The experimental results show yields detection accuracy 92.97%, which higher than existing methods computational demands are far lower novel for automatic determination as an application metallographic