Warping Calculation
Mathematical transforms used in machine vision systems correct for non-linear spatial distortions introduced by camera lenses and perspective angles. A polynomial warping matrix maps the pixel coordinates of a raw, distorted image to a corrected, rectilinear coordinate system. This matrix is calculated by comparing the imaged positions of a known reference target against its actual physical dimensions.
Once computed, the transformation is applied to every incoming frame to ensure that all measurements represent true physical distances.
Sensor Alignment
Lens distortion and sensor misalignment are common problems in automated optical inspection systems. By applying a polynomial warping matrix, the inspection software can compensate for barrel and pincushion distortions that would otherwise cause measurement errors. This correction is particularly important when inspect modules are mounted at an angle to capture three-dimensional solder fillet profiles.
The mathematical model utilizes higher-order polynomial equations to resolve complex, non-linear deformations across the entire field of view.
Metrology Accuracy
High-accuracy measurements of fine-pitch components require sub-pixel interpolation and precise spatial calibration. The calculation of the polynomial warping matrix is performed during the system setup and is validated using high-precision calibration plates. This matrix remains valid as long as the mechanical relationship between the camera and the sample stage does not change.
In assembly environments with high vibration or thermal variation, automated routines are scheduled to verify the matrix parameters periodically. The stability of this transformation ensures that the automated inspection algorithms can reliably measure component placement tolerances and detect solder joint skew. By maintaining this spatial correction, the inspection system prevented measurement drift that would otherwise trigger false reject cycles on high-volume production lines.