Photolithography is one of the most important processes in the production of integrated
circuits. Usually, attentive inspections are required after this process, but are
limited to the measurement of some physical parameters such as the critical dimension
and the line edge roughness. In this paper, a novel multiresolution multivariate technique
is presented to identify the abnormalities on the surface of a photolithographed
device and the location of defects in a sensitive fashion by comparing it to a reference
optimum, and generating fast, meaningful and reliable information. After analyzing the
semiconductor surface image in different levels of resolutions via wavelet decomposition,
the application of multivariate statistical monitoring tools allows the in-depth examination
of the imprinted features of the product. A two level nested PCA model is
used for surface roughness monitoring, while a new strategy based on ‘‘spatial moving
window’’ PCA is proposed to analyze the shape of the patterned surface. The effectiveness
of the proposed approach is tested in the case of semiconductor surface SEM
images after the photolithography process. The approach is general and can be
applied also to inspect a product through different types of images, different phases of
the same production systems, or different processes