ACCELERATED LENS DISTORTION CORRECTION WITH NEAR-CONTINUOUS WARPING OPTIMIZATION

摘要

A digital image processing technique, such as an image warping operation, is stored in a pre-computed lookup table (LUT) prior to image processing. The LUT represents a pixel-to-pixel mapping of pixel coordinates in a source image to pixel coordinates in a destination image. For vectors containing only inlier pixels, a fast remap table is generated based on the original LUT. Each SIMD vector listed in the fast remap is indexed to the coordinates of one of the source pixels that maps to one of the destination pixels in the vector. Other SIMD vectors that contain at least one outlier pixel are listed in an outlier index. For each vector indexed in the fast remap, linear vector I/O operations are used to load the corresponding source pixels instead of using scatter/gather vector I/O load operations via the LUT. The remaining outlier pixels are processed using scatter/gather I/O operations via the LUT.

主权项

1. A computer-implemented image processing method, comprising:
receiving, by a computer processor, a first lookup table representing a pixel-to-pixel mapping of source pixel coordinates of an input image to destination pixel coordinates of an output image; generating, by the computer processor, a second lookup table representing a vector-to-pixel mapping of a first plurality of vectors of the destination pixel coordinates to a first plurality of the source pixel coordinates based on the first lookup table, wherein all of the destination pixel coordinates in each of the first plurality of vectors correspond to pixel-to-pixel mappings in the first lookup table that are linear with slope 1 with respect to each other; generating, by the computer processor, a third lookup table representing a listing of a second plurality of vectors of the destination pixel coordinates, wherein at least one of the destination pixel coordinates in each of the second plurality of vectors correspond to a pixel-to-pixel mapping in the first lookup table that is not linear with slope 1 with respect to other pixel-to-pixel mappings; receiving, by the computer processor, source image data; applying, by the computer processor, a linear remapping operation from the source image data to each of the destination pixel coordinates in the first plurality of vectors in the second lookup table to obtain a first portion of destination image data; and applying, by the computer processor, a non-linear remapping operation from the source image data to each of the destination pixel coordinates in the second plurality of vectors in the third lookup table to obtain a second portion of the destination image data.