During the egress phase of Mars exploration, the egress mechanism image segmentation is the basis of the egress mechanism vision measurement analysis. In this paper, an egress mechanism image segmentation method based on region and feature fusion is proposed. Firstly, preliminary image segmentation based on color feature is carried out in the scene image. Then, according to the difference between the reference image and the real time scene image in image scale, rotation and 3D view angle, the rotation, scale and affine invariant feature points are extracted and used as the seed pixels to complete egress mechanism image segmentation using region growing. Lastly, by constructing environment for experiments, the method of egress mechanism image segmentation has been verified. The method is robust to the change of the view angle, the change of the egress mechanism slope and the change of the illumination. The experimental result shows that it can be used in the egress mechanism image segmentation in the Mars exploration.
[1]
G LoweDavid,et al.
Distinctive Image Features from Scale-Invariant Keypoints
,
2004
.
[2]
Andrew Zisserman,et al.
MLESAC: A New Robust Estimator with Application to Estimating Image Geometry
,
2000,
Comput. Vis. Image Underst..
[3]
Binoy Pinto,et al.
Speeded Up Robust Features
,
2011
.
[4]
Luc Van Gool,et al.
Speeded-Up Robust Features (SURF)
,
2008,
Comput. Vis. Image Underst..
[5]
Jiri Matas,et al.
Robust wide-baseline stereo from maximally stable extremal regions
,
2004,
Image Vis. Comput..