Projection Classification Based Iterative Algorithm

3rd International Conference on Manufacturing, Optimization, Industrial and Material Engineering

Zhang, Ruiqiu^1 ; Li, Chen^1 ; Gao, Wenhua^2

School of Design, South China University of Technology, Guangzhou, Guangdong

510640, China^1

School of Mathematics, South China University of Technology, Guangzhou, Guangdong

510640, China^2

关键词: 3D image reconstruction;    Classification based methods;    Convergence speed;    Fast convergence speed;    Iterative algorithm;    Projection image;    Reconstruction image;    Reconstruction speed;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/83/1/012022/pdf
DOI  :  10.1088/1757-899X/83/1/012022

来源: IOP

【 摘 要 】

Iterative algorithm has good performance as it does not need complete projection data in 3D image reconstruction area. It is possible to be applied in BGA based solder joints inspection but with low convergence speed which usually acts with x-ray Laminography that has a worse reconstruction image compared to the former one. This paper explores to apply one projection classification based method which tries to separate the object to three parts, i.e. solute, solution and air, and suppose that the reconstruction speed decrease from solution to two other parts on both side lineally. And then SART and CAV algorithms are improved under the proposed idea. Simulation experiment result with incomplete projection images indicates the fast convergence speed of the improved iterative algorithms and the effectiveness of the proposed method. Less the projection images, more the superiority is also founded.