Detection and research of non-metallic inclusion based on artificial intelligence
FU Xiang-kui1,FU Yi-di2
1. College of Mechanical Engineering， Wuhan Polytechnic University， Wuhan 430023, Hubei, China 2. College of Computer Science, Shanghai Fudan University, Shanghai 200433, China
Abstract：Using the artificial intelligence techniques and neural network technology, the image recognition of each metallographic image photo was conducted. Starting with non-metallic inclusion，proposed a new method to detect non-metallic dot inclusions, overcame the interference term of the image to find out the real non-metallic inclusion, and formulated a kind of computer software model which can detect the non-metallic inclusion automatically. Analyzed the detection method and key technology of the non-metallic inclusion according to the phase diagram of the steel specific operating procedures to observe the test results. Finally, the accuracy of the model was tested to get practical solutions, and made useful attempt on the automatic detection and modular standardized discriminate of the non-metallic inclusion.
傅向葵，傅一迪. 基于人工智能非金属夹杂物的检测与研究[J]. 中国冶金, 2014, 24(11): 14-17.
FU Xiang-kui,FU Yi-di. Detection and research of non-metallic inclusion based on artificial intelligence. China Metallurgy, 2014, 24(11): 14-17.
2014, Vol. 24 
