CAJ | 학술논문

针对二维光纤光谱数据结构的特殊性,先分块在其波长方向上进行整数H变换去除波长方向相邻像素点间的冗余,再将波长方向的直流分量在空间方向上进行整数H变换作为整块数据的低频部分,然后根据变换后系数的特点采用嵌入式零树(EZT)编码和算术编码相结合的方法进行编码.理论分析表明,该压缩编码方法可实现二维光纤光谱数据的无损压缩;实验结果也表明,该方法具有较好的压缩效果.
침대이유광섬광보수거결구적특수성,선분괴재기파장방향상진행정수H변환거제파장방향상린상소점간적용여,재장파장방향적직류분량재공간방향상진행정수H변환작위정괴수거적저빈부분,연후근거변환후계수적특점채용감입식령수(EZT)편마화산술편마상결합적방법진행편마.이론분석표명,해압축편마방법가실현이유광섬광보수거적무손압축;실험결과야표명,해방법구유교호적압축효과.
Based on the structural characteristics of 2D fiber-fed spectra data, integer H-transform is calculated on each block of the original image in wavelength direction to remove the redundancy of adjacent pixels. Then integer H-transform is calculated in space direction on the DC component of wavelength direction and the transform coefficients are taken as the low-frequency part of the block. Finally, both Embedded Zero-tree(EZT) coding and arithmetic coding are adopted to code the coefficients. Theoretical analysis shows that the coding method can be applied to lossless compression for 2D fiber-fed spectra data. The experimental results demonstrate the better compression ratio of the proposed method.