Sparse Representation over Learned Dictionary for Symbol Recognition

Tiêu đềSparse Representation over Learned Dictionary for Symbol Recognition
Loại công bốJournal Article
Năm xuất bản2016
Tác giảDo, T-H, Tabbone, S, Ramos-Terrades, O
Tạp chíSignal Processing
Tập125
Trang36-47
Thời gian xuất bản1/2016
Tóm tắt

In this paper we propose an original sparse vector model for symbol retrieval task. More specifically, we apply the K-SVD algorithm for learning a visual dictionary based on symbol descriptors locally computed around interest points. Results on benchmark datasets show that the obtained sparse representation is competitive related to state-of-the-art methods. Moreover, our sparse representation is invariant to rotation and scale transforms and also robust to degraded images and distorted symbols. Thereby, the learned visual dictionary is able to represent instances of unseen classes of symbols.

DOI10.1016/j.sigpro.2015.12.020