Pick and Place Machine Vision Calibration for Fine Pitch Components
Fine pitch machine vision calibration requires quartz grid targets, telecentric optics, and closed-loop thermal drift compensation to hold 5 µm accuracy.
Pixel Scale
Spatial Relationship
Ratio between the physical dimensions of an object and its representation on a digital sensor determines the measurement accuracy of an inspection system. Establishing the pixel scale involves measuring a known reference standard to calculate the distance each pixel covers in millimeters or microns. Calculated values allow the software to convert image coordinates into real-world units for quality checks.
Sensor Geometry
Changes in the working distance between the lens and the board will alter the size of the features on the sensor. Because the pixel scale is sensitive to these fluctuations, fixed-mount cameras are preferred for consistent automated inspection. Any shift in the camera position requires a new calibration to update the conversion factor.
Measurement Precision
Precision in the assembly process relies on the ability of the vision system to detect small deviations in component placement. If the pixel scale is too large, the system will miss errors that fall below the detection threshold. A balance must be struck between the field of view and the resolution to ensure that the entire board is inspected with enough detail to catch small defects.
Systems that use multiple cameras can achieve a finer resolution by splitting the board into smaller segments.