文章基本信息

标题：BAG-OF-SHAPES DESCRIPTOR USING SHAPE ASSOCIATION BASED ON FREEMAN CHAIN CODE
本地全文：下载
作者：EMA RACHMAWATI ; IPING SUPRIANA ; MASAYU L. KHODRA 等
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：5
出版社：Journal of Theoretical and Applied
摘要：A novel bag-of-shapes descriptor constructed using shape association is presented in this paper. We believe that shape association has significant impact in constructing better shape representation of object, for the purpose of object recognition. In our proposed model, shape association is represented in the set of representative prototypes, which is generated through K-medoids clustering based on association likelihoods. The association likelihood is obtained through pairwise distance computation using Needleman-Wunsch algorithm, as the shape is represented in sequence of code of Freeman Chain Code. We evaluate our method on a set of 32 fruit subcategories captured in multi viewpoint. We show that our approach can reliably classify the shape of multi-class fruit with average accuracy of 82.96 % using nearest neighbor classifier.
关键词：Bag-of-Shapes; Freeman Chain Code; Shape Association; K-medoids clustering; Needleman-Wunsch Algorithm; Nearest Neighbor Classifier