A robust approach for object matching and classification using Partial Dominant Orientation Descriptor

详细信息    查看全文

• 作者：Abdessamad Elboushaki ; ^{abdessamad.elboushaki@gmail.com}Author Vitae ; Rachida HannaneAuthor Vitae ; Karim AfdelAuthor Vitae ; Lahcen KouttiAuthor Vitae
• 关键词：Object matching ; Object classification ; Image classification ; Partial Dominant Orientation Descriptor ; K-Nearest Neighbors ; Adaptive object distance
• 刊名：Pattern Recognition
• 出版年：2017
• 出版时间：April 2017
• 年：2017
• 卷：64
• 期：Complete
• 页码：168-186
• 全文大小：3171 K
• 卷排序：64

文摘

We propose a novel descriptor for complex object matching and classification. The proposed descriptor combines both local and global properties of interest points. The proposed descriptor is highly distinctive and insensitive to common image transformations. Adaptive object distance is proposed to measure the similarity between objects. We compare our approach with many existing methods and we achieve the state-of-the-art performances on several benchmarks.