The industrial tendency to increase complexity and density of VLSI circuits has necessitated accurate, reliable inspection of an IC wafer throughout its entire process lifetime. When inspecting in-process and in-situ, a larger number of defects are more readily detected. In particular, thin films which make up part of the IC structure and are present after several processing steps exhibit certain characteristic colors when illuminated with white light. Exploiting this color phenomenon, we introduce an inspection system which can measure critical film thicknesses, segment IC images, and detect an entire class of color defects that would be difficult or impossible to detect with typical grayscale imaging. We carry out this inspection on a unique multi-window parallel hardware architecture which allows the inspection process to be performed at high speeds. Based on image understanding techniques, the inspection is carried out in a two stage fashion where defects are first rapidly hypothesized and then verified in detail only within the salient regions of an image, thus eliminating a large amount of irrelevant data. The system has been tested on numerous IC images and shows promising results.