基于V型扫描的星载多光谱图像无损压缩算法

为了实现星载多光谱图像的无损压缩,深入研究了空间数据系统咨询委员会(CCSDS)建议的123.0-B-1标准;并结合CCD多光谱遥感成像特点,提出了利用V型扫描对预测方式进行改进;最后在VC6.0平台下实现了该算法。实验结果表明,该算法的平均压缩率比Z型扫描方式的CCSDS 123.0-B-1、CCSDS 121.0-B-1、CCSDS 122.0-B-1、JPEG-LS标准算法的平均压缩率分别高6.7%、68.9%、21.8%和12.9%。该算法复杂度低,适合运用于星上系统。     

Visibly Transparent Multifunctional Metascreen for Tunable Infrared Coloration and Microwave Transmission

Metascreens are remarkable optical materials that have attracted significant research interest owing to their ability to manipulate electromagnetic waves by selectively transmitting, reflecting, and absorbing specific wavelengths. The present study introduces a new metascreen concept called visibly transparent multifunctional metascreen (VTMM), which is capable of simultaneously transmitting visible light, selectively transmitting microwaves, and modulating infrared (IR) emissions, all through a single surface. The VTMM is fabricated by patterning a thin metal film in a square-loop pattern on a transparent substrate, resulting in frequency-selective transmission characteristics in the X-band and visible transparency. Before demonstrating independent control of the visible and IR coloration of the VTMM by varying the metal film thickness and using color backgrounds, the visible and IR coloration characteristics of the ultrathin metal film on glass (MFG) are analyzed. Based on the findings, the VTMM preserves optical images behind the metascreen while reducing the average IR signature from 89.0 to 69.3 W m⁻² sr^-1, with up to 23.5% reduction compared to a glass substrate, as the metal film thickness increases from 40 to 100 nm is confirmed. Furthermore, the designed selective transmission characteristics are shown to remain consistent across varying metal film thicknesses.     

环境卫星多光谱图像压缩算法

基于环境卫星多光谱图像特点的分析,提出了一种新的基于三维等级树集合划分算法(3D-SPI HT)和感兴趣区域(ROI)编码相结合的多光谱图像压缩算法。首先在谱间采用两种小波基相结合的三维离散小波变换(3D-DWT),去除多光谱图像在空间和谱间的冗余信息,减少恢复光谱的误差值,然后采用部分三维等级树集合划分算法和小波系数提升的感兴趣区域编码相结合的方法。该方法对小波系数从空间方向树上按对恢复光谱信息的重要性不同进行合理的码率分配,使得恢复光谱具有更好的分辨率,并依据比特平面层中重要系数的统计概率来自适应地进行3种编码模式的选择,提高了编码效率。实验数据结果表明,该算法比传统算法更好地保护了多光谱图像中的光谱信息,在压缩比为8∶1的情况下,满足了环境卫星多光谱图像压缩系统的要求。     

基于多光谱成像的光谱反射率重建

一些对颜色重现要求较高的应用领域需要获取目标表面上各点的光谱反射率。大多数天然物质表面的光谱反射率曲线比较平滑,可视为几个基向量的线性组合。基于这一原理的多光谱成像技术可以准确快速地重建目标表面的光谱反射率。通过对NCS色卡进行主成分分析,得出了彩色印刷品光谱反射率的基向量。建立了一个多光谱成像系统,用以比较基向量个数不同对反射率重建效果的影响。     

面向高保真再现的多光谱图像融合技术

面向高保真再现(高保真显示和高保真印刷)的多光谱图像融合是多光谱颜色再现的关键技术和核心环节.论文结合人类视觉系统的构成与特性,采用基于多分辨率分析理论的图像融合方法,并嵌入基于图像色貌模型的色彩转换方法,提出了面向高保真再现的多光谱图像融合方法,其核心为基于人类视觉系统的小波图像融合方法,并设计了融合框架、融合算法和融合效果评价指标.最后通过高分辨率图像与多光谱图像的融合试验,并通过融合指标的分析计算验证了此技术方法的有效性,它为颜色视觉的阶段理论学说提供了新的理论解释.     

Multi-Analytical Characterization and Radiocarbon Dating of a Roman Egyptian Mummy Portrait

Fayum mummy portraits, painted around 2000 years ago, represent a fascinating fusion of Egyptian and Graeco-Roman funerary and artistic traditions. Examination of these artworks may provide insight into the Roman Empire’s trade and economic and social structure during one of its most crucial yet still hazy times of transition. The lack of proper archaeological documentation of the numerous excavated portraits currently prevents their chronological dating, be it absolute or relative. So far, their production period has been defined essentially on the basis of the relevant differences in their pictorial style. Our study introduces the use of Accelerator Mass Spectrometry (AMS) to assess the age of a fragment of an encaustic painting belonging to the corpus of the Fayum portraits. The unexpected age resulting from ¹⁴C analysis suggests the need to reconsider previous assumptions regarding the period of production of the Fayum corpus. Furthermore, our multi-analytical, non-invasive approach yields further details regarding the fragment’s pictorial technique and constituting materials, based on spectral and morphological analysis and cross-sectional examination.     

A new micro programmable blazed grating (μPBG) and its application to multispectral imaging

A new micro programmable blazed grating (μPBG) was designed, fabricated and characterized using the two-layer polysilicon surface micromachining process. The μPBG presented is simple in structure and has the capability of continuous spectral tuning for a wide wavelength range. To maximize the operational blazed angle, dimple structure was adopted as the key component for grating slats, and its height was close to the air gap. By the measurements, the developed μPBG sample can reach a maximum blazed angle of ~5.19° at a driving voltage of 112 V. Accordingly, the blazed wavelength can reach a mid-infrared wavelength of 4.88 μm. Then, a preliminary experiment to investigate the feasibility of multispectral imaging by using the developed μPBG as the core dispersing element was performed, and the results are encouraging. The advantages of the proposed method for multi- spectral imaging are discussed, and some suggestions for further research are also put forward.     

基于近地面多光谱的黄河三角洲典型地区土壤含盐量估算研究

选择山东省无棣县“渤海粮仓”项目核心示范区为研究区,利用ADC便携式多光谱相机和EC110便携式盐分计,采集该区近地多光谱相片和土壤表层含盐量数据,通过NDVI, SAVI, GNDVI三种植被指数分别与实测土壤含盐量构建线性、指数、对数、乘幂、二次和三次函数共18种模型,进而优选土壤盐分含量最佳估测模型,反演和分析研究区土壤盐分状况。结果显示,各模型均可有效估测土壤盐分含量,以SAVI为因变量构建的各模型估测效果较好,其中以SAVI的线性模型(Y=-0.524x+0.663,n=70)为最佳,显著检验水平下的F检验值最高,为141.347,估测R2为0.797,精度达到93.36%;研究区的土壤盐分含量集中在2.5‰～3.5‰之间,呈现从西南向东北逐渐升高的明显分布规律。探索了基于近地面多光谱数据的土壤含盐量估测方法,为研究区乃至整个黄河三角洲滨海盐碱土的盐分含量估测提供了一种快速有效的技术方法。     

A rate control approach to distributed source coding for interferential multispectral image compression

Distributed source coding (DSC) is applied to interferential multispectral image compression owing to strong correlation among the image frames. Many DSC systems in the literature use feedback channel (FC) to control rate at the decoder, which limits the application of DSC. Upon an analysis of the image data, a rate control approach is proposed to avoid FC. Low-complexity motion compensation is applied first to estimate side information at the encoder. Using a polynomial fitting method, a new mathematical model is then derived to estimate rate based on the correlation between the source and side information. The experimental results show that our estimated rate is a good approximation to the actual rate required by FC while incurring a little bit-rate overhead. Our compression scheme performs comparable with the FC based DSC system and outperforms JPEG2000 significantly.     

微型快照式窄带多光谱成像宫颈癌筛查方法

白光阴道镜图像对比度较低,不利于医生鉴别不同病变程度组织,也不利于自动化宫颈癌筛查。利用癌变组织富含血红蛋白成分及血红蛋白具有特征波段这一特性,与传统高光谱空间扫描成像及分时获取不同波段多光谱成像方法相反,利用快照式多光谱窄带成像来加速光谱图像获取过程,提升不同病变程度组织之间灰度对比度同时,降低后续图像分析处理算法难度,实现对宫颈组织病变类型高帧率自动化分类。首先,使用微型快照式窄带多光谱摄像方法,在血红蛋白的两个强吸收峰(415±10)和(525±10) nm、一个反射带(620±10) nm和一个背景波段(450±10) nm共四个波段对宫颈组织进行快照式零时差获取四幅窄带光谱图像。而后,对所获取的光谱图像进行简单代数加减,以生成突显病变组织的融合图像,提高不同病变程度组织之间的对比度。最后,使用欧式距离分类算法,对光谱融合图像中不同病变级别进行分类,建立计算机辅助宫颈癌筛查方法。创新点在于实现了高帧率计算机辅助光学病理诊断方法。分别采用临床常规白光阴道镜及微型快照式窄带多光谱摄像对宫颈癌手术切下的新鲜组织进行彩色图像及光谱融合图像的高帧率采集,并使用同一个欧式距离分类算法对两种图像进行自动分类,分类结果都以组织病理诊断作为标准来计算正确率。通过对比两种分类结果正确率来检验光谱融合图像相对于彩色图像是否提升对比度,及其是否可以实现与组织病理诊断(金标准)结果一致的诊断。欧式距离分类算法对光谱融合图像分类准确率接近100%,远高于对白光阴道镜图像约50%的准确率。多位临床医生对基于微型快照式多光谱摄像头光谱融合图像的计算机自动分类结果表示接受。微型快照式窄带多光谱成像方法能有效提升光谱融合图像获取帧率及不同病变程度组织之间灰度对比度,能有效快速地将宫颈组织划分为与组织病理诊断结果一致的病变类型。由于诊断客观、无创伤、结果立等可得,该方法将有助于实现落后地区宫颈癌筛查的普及以及图像导航下的宫颈癌精准治疗手术。