Currently, the scanning laser Doppler vibrometer (SLDV) has become a standard laboratory tool for high spatial resolution vibration measurements. Though the quality of the measurement result is very high in well-controlled experiments, there are still some limitations on the application in harsh industrial environments. Firstly, the user interaction of available SLDV systems is too high, requiring trained personnel. But more important, laser drop-out can cause outliers in the measurements when non-treated (and often dark) surfaces are measured. These limitations prevent the real breakthrough of the SLDV in industry, notwithstanding the fact that a high potential exists for the use of the SLDV for quality control. In this article, a procedure will be developed to make the SLDV measurement process fully automatic and more robust. Firstly, an automatic laser calibration and component recognition method -- that is based on newly developed image processing algorithms -- is described. Secondly, a robust singular value decomposition (SVD) is introduced to eliminate laser drop-out from the measurements. The merits of the proposed methodology (measurement automation and robustification) for quality control purposes will be validated on SLDV measurements of a mask of a picture tube.