Prediction of the Number of Defects in Image Sensors by VM Using Equipment QC Data