作者: Yao Yao , Xin Xin , Ping Guo
DOI: 10.1007/978-3-642-34500-5_10
关键词:
摘要: In image fusion techniques based on joint sparse representation (JSR), the composite is calculated from of features, which are represented with coefficients. Orthogonal matching pursuit (OMP) and basis (BP) main candidates to estimate Previously OMP utilized for advantage low complexity. However, noticeable errors occur when dictionary JSR cannot ensure coefficients enough. Alternatively, BP more robust than in such cases (though suffered larger complexity). Unfortunately, it has never been studied tasks. this paper, we investigate fusion. The target verify that 1) what extent can outperform OMP; 2) trade-off between OMP. Finally, conclude, some cases, obviously outperforms one under an affordable computational