关键词: 压缩感知, 正交匹配追踪, 稀疏表示, 压缩重构

Abstract:

Compressed sensing theory is a new sampling theory based on the sparsity of signal in recent years. Orthogonal matching pursuit algorithm is one of the typical reconstruction methods, aiming at the problems in speech recognition, using orthogonal matching pursuit algorithm for signal reconstruction of speech signal reconstruction algorithm, compared to the traditional compressed sensing is more suitable for the noisy speech, reconstructed speech quality will be higher, the denoising effect it will be more obvious. The basic research on the performance of speech signal CS is of great practical value and significance.

Key words: compressed sensing, orthogonal matching pursuit, sparse representation, compression reconstruction