Geometric Adjustment
Image processing algorithms adjust for the apparent shift in a component’s position caused by the viewing angle of off-axis cameras. This computational adjustment, known as parallax correction, ensures that components near the edge of a wide-angle lens are not misidentified as being misaligned. In automated optical inspection, cameras positioned directly above the center of the board view peripheral parts at an angle.
Without adjusting for this shift, the system would calculate incorrect placement coordinates and generate false defect reports.
Calculation Process
Mathematical correction relies on the known height of the component and the distance from the optical axis. Using the sensor’s focal length and the board’s design files, the algorithm calculates the offset for each part on the assembly. Parallax correction becomes increasingly important as components become taller, because taller bodies produce larger shift errors on the image sensor.
The software applies a coordinate transformation to shift the inspected features back to their true positions. This process enables high-speed single-camera systems to inspect large printed circuit boards with high accuracy. Proper computation ensures that automated systems maintain tight tolerance bands even on mixed-technology boards.
Process Impact
Implementing accurate correction reduces the rate of false failures at the optical inspection stage. Without this adjustment, the machine would reject acceptable boards, requiring operators to spend time manually verifying false defects. High-mix assembly lines benefit most because they handle a wide range of component heights on the same board surface.
Proper calibration ensures that the correction remains stable across different production runs.