Zoom Lens Distortion Correction Using Hexic Polynomial Data

The patent changes the parameter representation by using a hexic polynomial model with specifically constrained coefficients (k1, k2, k3) instead of storing complete correction data for each imaging condition. This parameter transformation reduces the data structure from comprehensive lookup tables to compact coefficient sets that can represent distortion characteristics across multiple focal lengths with fewer variables.

The patent segments the correction data storage by dividing the zoom range into specific focal length intervals (fm1 to fm2) and using different polynomial coefficient sets for different segments. This segmentation allows accurate representation of distortion characteristics in each segment while avoiding the need to store data for every possible imaging condition, thus reducing overall data amount while maintaining correction accuracy.

The patent applies dynamics by making the polynomial coefficients dynamic rather than static. Instead of using fixed approximation coefficients across the entire zoom range, the system updates coefficients (k1, k2, k3) at specific focal length intervals (fm1 to fm2) to adapt to the peculiar changes in distortion characteristics during zooming. This dynamic coefficient adjustment maintains correction accuracy while preserving data reduction benefits.

The patent performs preliminary action by pre-calculating and storing polynomial coefficients at key focal length intervals before actual imaging operations. By anticipating the need for correction at different zoom levels and preparing appropriate coefficient sets in advance, the system ensures accurate correction is available when needed without requiring real-time computation or excessive data storage during operation.