Satellite target extraction method

Abstract

The invention provides a satellite image target extraction method. The method comprises the following steps of: taking a central point of a target as a center, selecting a window with a certain size, fitting a cross vertical angle target black and white step conversion curve by adopting Boltzmann function curve, truly and accurately simulating the black and white gray value change characteristics on the image, accurately solving the central point position of the black and white intersection, fitting a straight line through least squares, and finely solving the high-precision central coordinates of the target in a mode of intersecting the column to straight line. Therefore, high-precision control data is provided for geometric calibration operation of the satellite.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, in particular to a satellite image target extraction method. Background technique [0002] In order to determine and verify the geometric accuracy calibration of the satellite, it is necessary to manually lay out a certain number of artificial marks (ie targets) on the ground, measure the geodetic coordinates of the center point of the target, and determine the image coordinates of the target center on the satellite image, which is consistent with the measured geodetic coordinates. The corresponding control data can be obtained. Therefore, it is necessary to extract the target center in the satellite image. On an image with a certain resolution, the position of the center point of the target cannot be manually determined to the sub-pixel level. In the past, the resolution of satellite images was not very high. The existing technology can reach the sub-pixel level by adopting gr...

AI Technical Summary

Problems solved by technology

However, due to the image of the modulation transfer function, the gray value of some pixels that deviate from the center of the target within the target range does not necessarily satisfy the regular change in gray value.

At this time, using polynomial fitting cannot satisfy all the points involved in the fitting calculation, but too few points cannot fit a suitable polynomial curve, and the extreme value of curve differentiation cannot well represent the black-white junction central location of

Images