全息图的小波域BP神经网络压缩实现

摘要

计算全息图的有效存储和快速传输对于实现真正意义上的动态三维全息显示有着十分重要的意义，然而计算全息图信息量庞大，不利于传输和存储，这就迫切需要对大数据量的全息图进行快速高效的压缩。基于此，提出一种基于小波域BP神经网络的全息图压缩技术，即先用小波变换对全息图进行预处理，通过将小波基与全息图的内积进行加权和来实现全息图的特征提取，然后将提取的特征向量代入神经网络以完成函数逼近、分类，实现全息图的压缩。该方法可获得124.52∶1的压缩比且仍能获得较清晰的再现像，实验结果很好地证明了该方法的可行性和有效性，且算法结构简单，运算速度快，能在较大压缩比下恢复出令人满意的再现像。%Holographic storage and fast transmission are great significance for realizing the true dynamic 3 -D holo-graphic display,but the huge amount of information calculation of computer-generated hologram (CGH)is not condu-cive to the transmission and storage.So there is an urgent need for fast and efficient compression method aiming at holograms with large amounts of data.Based on this,a BP neural network algorithm of hologram compression in wave-let domain is proposed.Firstly,computer-generated hologram pretreatment is carried out by wavelet transform.Sec-ondly,the inner product of wavelet and holograms are weighted and used to implement the feature extraction of holo-gram.Then the extracted feature vectors are substituted into neural network so as to implement the function approxi-mation,classification and hologram compression.The compression ratio can reach 1 24.52 ∶1 and still get a clear im-age reproduction.The experimental results clearly show the feasibility and effectiveness of the method.The proposed algorithm has the advantages of simple structure and fast calculation speed,and can recover satisfied reconstructed im-age at the larger compression ratio.