新的基于分裂层树编码图像的传输方案_bmp图像编码怎么算

  摘要:提出了一种新的基于分裂层树编码(SPIHT)的实时图像传输方案。该方案首先将图像数据进行小波变换,然后将变换后的小波系数进行合理分块并分别采用 SPIHT编码,以防止在无线信道传输中产生误码扩散。最后为了提高重构图像的恢复质量,提取每一个块中最高一级的小波系数数据进行重复传输,同时采用最优的帧长,提高图像传输系统的吞吐量。理论分析和仿真结果表明,提出的方案比传统的基于SPIHT算法的图像传输方案具有更高的抗误码特性,重构图像的峰值信噪比提高约10�dB。
  
  关键词:分裂层树编码;图像无线传输;最优帧长;吞吐量;抗误码
  中图分类号: TP391.41;TN911.73文献标志码:A
  �
  New scheme for image transmission based on SPIHT
  
  �
  FU Yao��*, LIU Qing.li
  School of Information Science and Engineering, Shenyang Ligong University, Shenyang Liaonning 110159, China
  
  Abstract:
  In this paper, a novel real-time image transmission scheme based on SPIHT is presented. Firstly, the image data need to be transformed by wavelet. Secondly, in order to resist error pervasion when image is transmitted, the wavelet coefficients are separated into small blocks and encoded by SPIHT. Finally, in order to improve the quality of the restructure image, the wavelet coefficients of the highest level in every block are transmitted repeatedly. In order to improve throughput of image transmission system, the optimum frame length is proposed.. Both theoretical demonstration and simulation results here have validated that the proposed scheme provides higher character of resisted error than traditional scheme based on SPIHT, and can improve the peak signal to noise ratio of the restructure image about 10�dB.In this paper, a new real.time image transmission scheme based on Set Partitioning In Hierarchical Tree (SPIHT) was proposed. Firstly, the image data needed to be transformed by wavelet. Secondly, in order to resist error pervasion when image was transmitted, the wavelet coefficients were separated into small blocks and encoded by SPIHT. Finally, in order to improve the quality of the restructured image, the wavelet coefficients of the highest level in every block were transmitted repeatedly. In order to improve the throughput of the image transmission system, the optimum frame length was proposed. Both theoretical demonstration and simulation results here have validated that the proposed scheme provides stronger error resilience than traditional scheme based on SPIHT, and can improve the peak signal to noise ratio of the restructured image about 10�dB.
  �Key words:
  Set Partitioning In Hierarchical Tree (SPIHT); image wireless transmission; optimum frame length; throughput; resisted error
  
  �
  
  
  0 引言�
  实时图像在无线传输中,由于数据量非常大,需要进行压缩编码�[1]才能传输。图像压缩编码虽然减少了数据量,但压缩后的图像码流之间产生了依赖性,加上无线信道具有传输时延长、误码率高等特点�[2],通过无线信道传输图像将不可避免地出现差错和误码。如果有某个比特出错,那么对该比特后面数据的解码将产生巨大的影响,最终导致重构图像质量严重下降�[3]。因此,如何提高图像数据的抗误码性和重构图像的恢复质量成为待解决的问题。�
  针对该问题文献[4]将变换后的小波系数分成四块并分别进行Huffman编码,同时对低频数据采用回退�N�帧(Go.Back.�N�,GB�N�)重传方式�[5]。该方案虽然提高了重构图像的质量,但传输的总时延较大,不适应实时图像传输。文献[6-7]分别提出了一种基于JPEG2000编码的不等差错保护图像传输方案,但JPEG200的编码解码过程实现比较复杂,同时采用差错保护会增加数据的冗余。文献[8]提出了一种非均匀纠错图像传输方案,根据小波系数重要性的不同,分为不同等级进行保护。该方案虽然减小了由于信道误码带来的损失,但传输过程中增加了数据的冗余,降低了信道的利用率。文献[9]提出了一种基于水印的图像错误隐藏方案,由误码产生的图像损坏可以通过水印来恢复,提高恢复图像的峰值信噪比,但实现过程比较复杂,不利于实现。文献[10]将最高一级小波分解子带的码流数据嵌入到原压缩码流末端相同长度的比特位,增强了对低频数据的保护,但这样会损失高频数据,不适用于误码率较低的无线信道。�
  针对以上的问题,本文提出了一种新的基于层树分级编码�[11-12] (Set Partitioning In Hierarchical Tree,SPIHT)的实时图像传输方案。该方案首先进行提升小波变换,然后对小波系数进行分块并分别编码,提取每一个块中最高一级的小波系数码流放到压缩码流末端,最后根据不同的误码率封装成不同长度的数据帧在无线信道中传输。�1 图像传输总体框架�
  本文提出的图像传输方案总体框架如图1所示。
  发射端首先对遥感图像进行小波变换�[13-14],然后根据变换后的小波系数之间的关系将小波系数分解成若干各数据块。接着分别对各个数据块进行独立的SPIHT编码,防止数据在无线信道传输中由于误码而造成误码扩散,分别将各数据块的低频码流提取出来放到SPIHT编码的码流末端,提高低频数据的抗误码性。最后将所有的码流分别封装成数据帧在无线信道中传输。该方案不采用重传的方式,以减少传输时延,同时不采用信道编码,信道状态由误比特率(Bit Error Rate,BER)来度量。�
  
  2 小波系数分块�
  图像经过三次小波变换后被分成10个小波系数数据块,即LL3、HL3、LH3、HH3、HL2、LH2、HH2、HL1、LH1、HH1。左上角的LL3是低频数据对于恢复该图像来说是至关重要的,而其他的数据块是高频数据,主要是图像的边缘信息。如图2所示,分别提取LL3、HL3、LH3、HH3中4×4的小波系数数据块, 分别提取HL2、LH2、HH2中8×8的小波系数块,分别提取HL1、LH1、HH1中16×16的小波系数数据块,可以将256×256的图像分解成64个数据块,各数据块内小波系数坐标之间的关系如表1所示。�

推荐访问:分裂 传输 图像 编码