METHOD FOR DETECTING FEATURE POINTS WIDE ANGLE CAMERA

摘要

The present invention relates to a method for detecting a feature point of an image of a wide angle camera, which comprises the following steps: obtaining a grid image by photographing a grid pattern from at least one wide angle camera having a radial distortion; obtaining a histogram for all pixel lines of X and Y axes with respect to the grid image, and performing equalization of the histogram; generating a circular mask having a reference radius by using a mid-point-based algorithm, and performing a feature point detection algorithm which inspects a feature point by pixel unit in the entire area of the grid image by using the generated circular mask; and calculating a gray scale response value by using a variation amount of a gray scale value with respect to candidate feature points of the grid image or peripheral pixels of the candidate feature points by the feature point detection algorithm, and determining the corresponding candidate feature point as the feature point when the gray scale response value is greater than or equal to a predetermined threshold. Accordingly, the present invention can automatically detect the feature point by pixel unit over the entire area of the grid image photographed by the wide angle camera, and does not need to directly obtain coordinate values of crossing points of the grid image through a manual operation, so a time required to detect the feature point can be significantly reduced.