Optical emission spectrum processing using discrete-wavelet transform compression
Taikang Ning, Chien-Hsing Huang, J. Jensen, V. Wong, H. Chan · 2016
This paper describes the application of Discrete-Wavelet Transform (DWT) image compression to optical emission spectrum images collected during wafer fabrication. The goal of the current study is twofold. We are seeking not only a computationally efficient image compression algorithm that can significantly reduce the data storage requirement but also is capable of retaining sufficient data authenticity which is critical for diagnostic purposes. The Set Partitioning In Hierarchical Trees (SPIHT) algorithm was employed in our study to evaluate its performance when processing different types of wafer fabrication image data. We found that with a proper choice of the processing parameters in SPIHT it is achievable to obtain better than 99.5% data reduction while maintaining sufficient authenticity of the original image.