Fig 1 - uploaded by Yang Yang
Content may be subject to copyright.
An example which SUSAN operator detects the road contour and extracts the road feature points. (a) is a unstructured road image, (b) is road contour, the red dots in (c) are SUSAN feature points
Source publication
In this paper, we present a mixture feature based multi-temproal unstructured road image registration approach, which consists the following four steps. (i) SUSAN algorithm is used to detect the contour of the road and to extract the road image feature points; (ii) Using the mixture feature to guide the correspondence estimation on the extracted ro...
Contexts in source publication
Context 1
... smaller the region of SUSAN, the larger the response of the source feature point, so that the feature point information in the image is enhanced, the road contour is obtained (see Fig.1 (b)). Where the threshold is empirical, assuming that the maximum value of v is max ...
Similar publications
The traditional line matching method has the problems of inaccurate endpoint location, small matching quantity and long time-consuming for images with large viewpoint variation, and the present point matching technology for the same type of image is facing the challenge of improving the matching quantity and accuracy. To solve these problems, a fea...
Citations
... It is an essential problem in computer vision and pattern recognition applications and widely applied in remote sensing image processing. [3][4][5][6][7][8] The current nonrigid point set registration methods have two major types of classification according to their methodological differences: (a) iterative or noniterative methods and (b) learning or nonlearning methods. Since our work is mainly concerned with iterative methods and parameter adjustment problems, we introduce and discuss the current methods along the iterative and learning perspectives. ...
... We test the performances of our methods under three parameter adjustment approaches in a relatively wide range of random variation (i.e., D degrees from 1 to 8, R from −90 deg to 97.9% (5) 98.6% (4) 98.1% (5) 99.6% (5,9) 99.3% (5,12) 98.3% (5,10) 98.3% (5,12) 95.2% Bird 86.9% (7) 82.7% (5) 87.1% (7) 87.1% (5) 91.3% (2,14) 80,4% (5,12) 98.9% (6,10) 76.7% (5,12) 89.7% ...
... Hand 95.4% (4) 94.6% (5) 99.8% (4) 97.2% (5) 98.1% (5,10) 95.3% (5,12) 99.8% (6,10) 97.2% (5,12) 91.5% ...
Nonrigid point set registration is a key technology in the field of remote sensing image registration, which is widely used in military and civil fields such as natural disaster damage assessment, agricultural and urban land-use planning, environmental quality monitoring, and ground target identification. We present a multifeature energy optimization framework and parameter adjustment-based nonrigid point set registration that has three contributions: (1) an energy optimization framework is designed to freely combine multiple features for estimating correspondences between two point sets, (2) the thin-plate spline and Gauss radial basis function transformation models can be optionally implemented for solving two-dimensional, three-dimensional, or higher-dimensional registration problems, and (3) the free parameters can be adjusted by the proposed three parameter adjustment approaches to yield higher registration accuracy in varied registration patterns. We test the performances of our method in contour point sets, sequence images, remote sensing images, medical images, and real images and compare with 10 state-of-the-art methods, where our method shows favorable performances in most scenarios. © 2018 Society of Photo-Optical Instrumentation Engineers (SPIE).
