Scale-Adaptive NN-Based Similarity for Robust Template Matching

Zhang, Zhihao; Yang, Xianqiang; Jia, Xiaogang

doi:10.1109/tim.2020.3028401

Cited by 7 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experiments: The efficacy of the proposed method was evaluated using the MS COCO dataset, and the proposed method was compared with the recently developed SDIS and WSDIS algorithms [3]. To evaluate the matching performance of the proposed method under various rotations and scales, 5400 experiments were conducted.…”

Section: Similarity Matchingmentioning

confidence: 99%

See 1 more Smart Citation

A dual evaluation multi‐scale template matching algorithm based on wavelet transform

Zhu

et al. 2021

Electronics Letters

View full text Add to dashboard Cite

It is difficult to achieve the requirements simultaneously of real‐time operation and accuracy when the template matching method encounters the problems of rotation and scale invariance. This letter proposes a dual‐evaluation multiscale template‐matching algorithm based on wavelet transform. First, the image grids generated from a strengthened edge image based on the wavelet transform are defined to reduce the region for detecting feature points. Then, an evaluation strategy based on gradient direction entropy is proposed to evaluate the local pixel for detecting local candidate points that contain rich information. Another evaluation strategy, based on the dominant gradient direction and the uniform LBP model, has been proposed to extract rotation‐invariant feature pixels. To conduct fast and robust matching, the proposed strategies extract fewer feature points with rich local information. The experimental results demonstrate that the proposed method is robust to rotation and scale invariance and achieves real‐time accuracy.

show abstract

Section: Similarity Matchingmentioning

confidence: 99%

“…Zhang et al. [3] proposed a weighted scalable DIS (WSDIS) measure by extending diversity similarity (DIS).…”

Section: Introductionmentioning

confidence: 99%

A dual evaluation multi‐scale template matching algorithm based on wavelet transform

Zhu

et al. 2021

Electronics Letters

View full text Add to dashboard Cite

show abstract

“…Template matching is one of the frequently used machine vision techniques and is useful for a range of applications including image matching, logo recognition, 3D reconstruction and more [1][2][3][4]. The template matching task can be decomposed into two problems: how to create a more accurate similarity measure, and how to create a more efficient feature representation.…”

Section: Introduction and Reviewmentioning

confidence: 99%

Template matching via bipartite graph and graph attention mechanism

Zhao

Wei

et al. 2022

IET Image Processing

View full text Add to dashboard Cite

Most existing template matching algorithms are global matching between the template target and the search region, which makes the matching process retain a lot of unfavourable background information and ignore the structure and local information of the template target. To address this problem, a template matching algorithm based on bipartite graph and graph attention mechanism is proposed in this paper. The algorithm models the similarity matching problem between template features and search region features as a complete bipartite graph, realises local‐to‐local information transfer between the two, and uses the graph attention mechanism to apply weights between local information to obtain a learnable embedding network module. In addition, in terms of feature representation, a multi‐level feature fusion module based on CNN is introduced, which improves the representation of a target by fusing features with different representational meanings of the target. Experimental results on several typical datasets show that the proposed algorithm achieves leading performance in terms of accuracy and efficiency compared to the two state‐of‐the‐art CNN‐based template matching algorithms, Deep‐DIM and QATM.

show abstract

CLaSP: Cross‐view 6‐DoF localisation assisted by synthetic panorama

Zhu,

Yan,

Cheng

et al. 2024

IET Computer Vision

View full text Add to dashboard Cite

Despite the impressive progress in visual localisation, 6‐DoF cross‐view localisation is still a challenging task in the computer vision community due to the huge appearance changes. To address this issue, the authors propose the CLaSP, a coarse‐to‐fine framework, which leverages a synthetic panorama to facilitate cross‐view 6‐DoF localisation in a large‐scale scene. The authors first leverage a segmentation map to correct the prior pose, followed by a synthetic panorama on the ground to enable coarse pose estimation combined with a template matching method. The authors finally formulate the refine localisation process as feature matching and pose refinement to obtain the final result. The authors evaluate the performance of the CLaSP and several state‐of‐the‐art baselines on the Airloc dataset, which demonstrates the effectiveness of our proposed framework.

show abstract

Scale-Adaptive NN-Based Similarity for Robust Template Matching

Cited by 7 publications

References 28 publications

A dual evaluation multi‐scale template matching algorithm based on wavelet transform

A dual evaluation multi‐scale template matching algorithm based on wavelet transform

Template matching via bipartite graph and graph attention mechanism

CLaSP: Cross‐view 6‐DoF localisation assisted by synthetic panorama

Contact Info

Product

Resources

About