Object registration for remote sensing images using robust kernel pattern vectors

Ding, Min; Jin, Zi; Tian, Zheng; Duan, Xinhui; Zhao, Wei; Li-juan, Yang

doi:10.1007/s11432-011-4478-2

Cited by 5 publications

(3 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where every Euclidean distance is measured between two ROIs' coordinates of the center of gravity given by formula (5). (ii) Line-feature based similarity measure: first construct the independent similarity, SSR(structural similarity ratio), to get the structural changing scales, i.e.,…”

Section: Roi Matching and Outlier Removalmentioning

confidence: 99%

“…The feature-based methods extract and match similar features from pairs of two or more images. Therefore effective feature detection and accurate feature matching methods play important roles in this process [2][3][4][5][6]. The area-based methods implement registration directly by utilizing the information correlation between reference and float images, which can be regarded as a template matching process [7][8][9].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A coarse-to-fine image registration method based on visual attention model

Feng

et al. 2014

Sci. China Inf. Sci.

View full text Add to dashboard Cite

A variable threshold visual sensing and image reconstruction method based on pulse sequence SCIENCE CHINA Information Sciences A progressive remote sensing image registration method based on adaptive scale SCIENTIA SINICA Informationis 41, 55 (2011); A model for visual image-background discrimination by relative movement Science in China Series C-Life Sciences 40, 79 (1997);. RESEARCH PAPER. SCIENCE CHINA Information Sciences

show abstract

Section: Roi Matching and Outlier Removalmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A coarse-to-fine image registration method based on visual attention model

Feng

et al. 2014

Sci. China Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…The past decade has witnessed promising progress on this change detection task [6] and other visual tasks [7][8][9][10][11][12] in remote sensing , by benefiting from the strong representative capability of deep neural networks [13,14]. However, most of these studies tackle the BCD problem with near-nadir images, which are taken nearly from a directly top-down view by aerial vehicles or satellites [15,16].…”

Section: Introductionmentioning

confidence: 99%

Detecting Building Changes with Off-Nadir Aerial Images

Pang¹,

Wu²,

Ding³

et al. 2023

Preprint

View full text Add to dashboard Cite

The tilted viewing nature of the off-nadir aerial images brings severe challenges to the building change detection (BCD) problem: the mismatch of the nearby buildings and the semantic ambiguity of the building facades. To tackle these challenges, we present a multi-task guided change detection network model, named as MTGCD-Net. The proposed model approaches the specific BCD problem by designing three auxiliary tasks, including: (1) a pixel-wise classification task to predict the roofs and facades of buildings; (2) an auxiliary task for learning the roof-to-footprint offsets of each building to account for the misalignment between building roof instances; and (3) an auxiliary task for learning the identical roof matching flow between bi-temporal aerial images to tackle the building roof mismatch problem. These auxiliary tasks provide indispensable and complementary building parsing and matching information. The predictions of the auxiliary tasks are finally fused to the main building change detection branch with a multi-modal distillation module. To train and test models for the BCD problem with off-nadir aerial images, we create a new benchmark dataset, named BANDON * . Extensive experiments demonstrate that our model achieves superior performance over the previous state-of-the-art competitors.

show abstract