论文部分内容阅读
针对传统变换基函数难以获得地震数据最优的稀疏表示,提出基于字典学习的随机噪声压制算法,将地震数据分块,每一块包含多个地震记录道在一定采样时间段内波形的信息,利用自适应字典学习技术,以地震数据块为训练样本,根据地震数据邻近块中记录道相似的特点,构造超完备字典,稀疏编码地震数据,从而恢复数据的主要特征,压制随机噪声.实验表明算法具有较高的PSNR值,并且能较好的保持地震数据纹理复杂区域的局部特征.
Aiming at the difficulty of obtaining the optimal sparse representation of seismic data by traditional transformation basis function, a random noise suppression algorithm based on dictionary learning is proposed. The seismic data is divided into blocks, each of which contains the information of waveforms of multiple seismic traces in a certain sampling time. Adaptive dictionary learning technology uses seismic data blocks as training samples and constructs an overcomplete dictionary based on the similar features of the recording tracks in adjacent blocks of seismic data to sparsely encode the seismic data so as to restore the main features of the data and suppress random noise.The experiment shows that the algorithm It has higher PSNR value and better preserves the local features of complex regions of seismic data texture.