Abstract | ||
---|---|---|
In order to better resolve the conflict between the full use and effective description of the Linear Feature information in the study of free form linear feature (FFLF) matching, this paper proposed a remote sensing image matching method using the hierarchical matching strategy. First the edge features of the image were detected and tracked, in order to extract the free-form sub-pixel linear features with better continuity; in the coarse matching process, the closed linear features (CLF), linear feature intersection (LFI) and corner (LFC) were selected as conjugated entity, and then the false match was gradually eliminated based on the area, angle and other geometry information as well as the parameter distribution features of the model determined by the conjugate features combination to be selected, finally the initial value of accurate matching was determined by the conjugate features; in the accurate matching process, based on multi-level two-dimensional iterative closest point (ICP) algorithm, sub-pixel edge points were orderly used with the sampling rate from low to high for matching. Experimental results show that this method has stable performance for the coarse matching; high accuracy and precision of the coarse matching can provide the initial matching parameters of high precision for accurate matching; accurate matching can reach the sub-pixel level precision equal to the least square matching and can better achieve stable and accurate matching for the images with smaller affine deformation. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1007/978-981-10-3005-5_18 | Communications in Computer and Information Science |
Keywords | Field | DocType |
Free-form linear feature (FFLF),Image matching,Hierarchical matching strategy,Closed linear feature (CLF),Linear feature corner (LFC),Linear feature intersection (LFI),Iterative closest point (ICP) algorithm | Template matching,Least squares,Affine transformation,Pattern recognition,Computer science,Image matching,Sampling (signal processing),Artificial intelligence,Initial value problem,Accuracy and precision,Iterative closest point | Conference |
Volume | ISSN | Citations |
663 | 1865-0929 | 0 |
PageRank | References | Authors |
0.34 | 3 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xiaowei Chen | 1 | 0 | 0.34 |
Haitao Guo | 2 | 147 | 20.07 |
Chuan Zhao | 3 | 4 | 1.47 |
Baoming Zhang | 4 | 0 | 0.34 |
Yuzhun Lin | 5 | 0 | 0.34 |