The invention provides a workpiece double-sided defect identification method and equipment based on multi-view
visual inspection, and relates to the technical field of
industrial machine vision, after a workpiece is migrated to a transparent detection table, a
light source is driven to be dynamically switched according to a
light source time sequence, and a multi-angle
image sequence is synchronously captured; and segmenting an
image sequence by taking a double-sided area of the workpiece as a constraint, generating an array type multi-angle segmentation sub-image of a
light source parameter index, and then outputting a double-sided defect panoramic distribution thermodynamic diagram through defect identification. The technical problems that in the prior art, due to the fact that workpiece overturning detection efficiency is low, synchronous imaging of double-face defects is difficult, and the defects are obviously interfered by ambient light, the defects lack space quantification basis, the
false detection rate is high, and sorting decision is interfered are solved. The technical effects of sub-
millimeter spatial positioning precision, multi-dimensional risk quantitative mapping and full-process automatic sorting decision
closed loop of double-sided defects of the workpiece are achieved, and the reliability and
processing efficiency of industrial quality inspection are remarkably improved.