Shape retrieval by using multi‐scale angle‐based representation and dynamic label propagation. Issue 4 (11th November 2020)

• Record Type: Journal Article
• Authors: Yu, Yanxia; Zheng, Danchen; Zhao, Liang; Sun, Chuang; Li, Xiang; Zhuang, Yan
• Is Part Of: IET cyber-systems and robotics. Volume 2:Issue 4(2020)
• Page Start: 197
• Page End: 204
• Publication Date: 2020-11-11
• Subjects: learning (artificial intelligence) -- shape recognition -- image matching -- image representation -- dynamic programming -- feature extraction -- distance learning -- image retrieval; dynamic label propagation -- robustness -- discrimination power -- triangle‐area representation -- shape matching method -- multiscale angle representation -- different sample points -- shape contour -- shape descriptors -- space angles -- different scale levels -- shape representation -- multiscale information -- dynamic programming -- shape distance -- feature matching step -- shape retrieval results -- pairwise shape distances -- previous distance learning methods -- authors method -- different shape databases -- multiscale angle‐based representation; Robotics -- Periodicals; Cybernetics -- Periodicals; Cybernetics; Robotics; Periodicals; 629
• DOI: 10.1049/iet-csr.2020.0044
• Languages: English
• ISSNs: 2631-6315

Abstract

Abstract : To improve the robustness and discrimination power of the triangle‐area representation, a novel shape matching method based on multi‐scale angle representation is proposed in this study. By analysing the configurations of different sample points from each shape contour, shape descriptors are constructed by using space angles at different scale levels. With the proposed shape representation, the multi‐scale information of shape contours is efficiently described, and the dynamic programming is further used to determine the correspondence between samples from different shapes and calculate the shape distance in the feature matching step. Moreover, to improve the shape retrieval results based on pairwise shape distances, the dynamic label propagation is introduced as the post‐processing step. Unlike previous distance learning methods learning the database manifold implicitly, the authors method retrieves relative objects on the shortest paths from near to far explicitly, and the underlying structure can be effectively captured. The proposed method tested on different shape databases provides the performances superior to many other methods, and it can be applied to visual data processing and understanding of the internet of things.