Adaptive wide-angle image correction method and device based on segmented mapping of single mesh

The invention discloses a self-adaptive wide-angle image correction method and device based on single mesh map segmentation mapping, the invention not only realizes the automatic extraction of controlpoint pairs, but also adapts to the optimal segmentation correction according to the lens distortion degree, realizes the automatic correction processing and improves the local and global correctionaccuracy. The method comprises the following steps: preprocessing the distorted mesh template image to obtain a single pixel binary distorted mesh template image; the neighborhood features of the intersection points are used to extract the control point pairs from the distorted mesh template graph. The minimum subsection optimization model of radial distortion is established. The inverse correction mapping table is established to obtain the coordinate correspondence between the correction result map and the original distorted image. Coordinate inverse mapping process is performed to obtain a first correction result ima