A method and system for automatically extracting the position and contour of landmark points in an image

Abstract

The invention relates to a method and a system for automatically extracting the position and profile of a mark point in an image. The method comprises the steps of obtaining an original image; performing median filtering processing on the original image; performing ellipse fitting processing on the mark point subjected to the filtering processing to obtain a preliminary extraction result; converting image point coordinates of points of an ellipse boundary of the preliminary extraction result into polar coordinates; and uniformly selecting the plurality of the points on the ellipse boundary to perform accurate extraction processing, and finally obtaining accurate data of the mark point. According to the method and the system, a mark point area in the image can be extracted automatically and more accurately based on preliminary extraction processing through accurate extraction processing, and a parameter equation of the mark point area is fitted out; and the extracted profile of the mark point can reach the sub-pixel level, so that the accuracy and stability of an existing method are greatly improved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method and system for automatically extracting the position and outline of circular marker points in an image. Background technique [0002] Using stereo vision for 3D reconstruction is an important means of close-range 3D reconstruction. According to the image point coordinates of the corresponding points with the same name in the left and right images with known internal and external orientation elements and the internal and external orientation elements of the image, the three-dimensional coordinates of the changed point can be obtained by using the principle of forward intersection. In order to facilitate the search for corresponding points with the same name, the method of pasting visual marker points on the surface of the object is usually used. This method is easy to operate and requires less calculation. To be accurate and reliable. Since the imaging of circular landma...

AI Technical Summary

Problems solved by technology

[0003] Most of the current methods for extracting marker points in images cannot completely remove the interference such as laser lines or other bright spots. Most of the extracted marker point contours are at the pixel level, and the accuracy and correctness need to be improved.

Images