Densely connected convolutional network block based autoencoder for panorama map compression

As a novel virtual reality (VR) format, panorama maps are attracting increasing attention, while the compression of panorama images is still a concern. In this paper, a densely connected convolutional network block (dense block) based autoencoder is proposed to compress panorama maps. In the proposed autoencoder, dense blocks are specially designed to reuse feature maps and reduce redundancy of features. Meanwhile, a loss function, which imports a position-dependent weight item for each pixel, is proposed to train and adjust network parameters, in order to make the autoencoder fit to properties of panorama maps. Based on the proposed autoencoder and the weighted loss function, a greedy block-wise training scheme is also designed to avoid gradient vanishing problem and speed up training. During training process, the autoencoder is divided into several sub-nets. After each sub-net is trained separately, the whole network is fine-tuned to achieve the best performance. Experimental results demonstrate that the proposed autoencoder, compared with JPEG, saves up to 79.69 % bit rates, and obtains 7.27dB gain in BD-WS-PSNR or 0.0789 gain in BD-WS-SSIM. The proposed autoencoder also outperforms JPEG 2000, HEVC and VVC in both BD-WS-PSNR and BD-WS-SSIM. Meanwhile, subjective results show that the proposed autoencoder can recover details of panorama images, and reconstruct maps with high visual quality.     

Dual pH‐ and thermal‐responsive nanocomposite hydrogels for controllable delivery of hydrophobic drug baicalein

Hydrogels have been widely used as mild biomaterials due to their bio‐affinity, high drug loading capability and controllable release profiles. However, hydrogel‐based carriers are greatly limited for the delivery of hydrophobic payloads due to the lack of hydrophobic binding sites. Herein, nano‐liposome micelles were embedded in semi‐interpenetrating poly[(N‐isopropylacrylamide)‐co‐chitosan] (PNIPAAm‐co‐CS) and poly[(N‐isopropylacrylamide)‐co‐(sodium alginate)] (PNIPAAm‐co‐SA) hydrogels which were responsive to both temperature and pH, thereby establishing tunable nanocomposite hydrogel delivery systems. Nano‐micelles formed via the self‐assembly of phospholipid could serve as the link between hydrophobic drug and hydrophilic hydrogel due to their special amphiphilic structure. The results of transmission and scanning electron microscopies and infrared spectroscopy showed that the porous hydrogels were successfully fabricated and the liposomes encapsulated with baicalein could be well contained in the network. In addition, the experimental results of response release in vitro revealed that the smart hydrogels showed different degree of sensitiveness under different pH and temperature stimuli. The results of the study demonstrate that combining PNIPAAm‐co‐SA and PNIPAAm‐co‐CS hydrogels with liposomes encapsulated with hydrophobic drugs is a feasible method for hydrophobic drug delivery and have potential application prospects in the medical field. © 2018 Society of Chemical Industry     

畔园记——一种聚落方法的探讨

论述了畔园的设计理念与方法，并探讨了一种当代 的聚落方法。提出了相关设计策略，包括通过场地结构与大单 元和间隙的有效结合以形成聚落；通过镶嵌术组织聚落，结合 景观与建筑；通过场地建造，竖向统合场地、设施、建筑，并 用中介物帮助揭示场地的质感；通过密度处理和空间开放，实 现聚落的亲疏有致。畔园的设计和建造，是探讨具有文野相 间、空间开放特征的聚落类型的一次尝试，并试图提出聚落中 的间隙美学。     

上海市遥感影像云服务平台设计与实现

针对上海市遥感影像管理和应用现状,采用镶嵌数据集为数据管理基本单元,云端影像处理服务为影像共享应用的基本手段,构建了上海市遥感影像云服务平台。解决了大规模遥感影像管理和应用存在的诸多问题,大幅缩短了影像应用的时间周期,并丰富了传统影像应用的手段,使遥感影像具有更强的适用性。     

济南市1∶10000数字高程模型的生成工艺

数字高程模型（ DEM）作为基础地理信息数据之一,在城市规划和管理中有着广泛的应用,本文简单介绍了利用适普软件公司的VirtuoZo软件制作济南市1∶10000数字高程模型（ DEM）的过程,并利用现有的各种地形图数据、二级导线点成果、控制测量成果对其精度进行分析,检测结果表明其精度满足国家相关规范要求,可为其他测绘单位生产数字高程模型（ DEM）提供参考。     

基于图像拼接的全景目标检测技术

针对战场指挥车等战地车辆内作战人员需要在车内实时了解外部战场动向的问题,提出一种基于图像拼接的全景目标检测技术研究。利用监控转台摄像头获得全景图像并检测运动物体,以此模拟战场实际情况。首先使转台进行水平360?旋转,同时使用安装在转台上的摄像头采集实时图像,利用转台转速与扫描场时间等参数对图像进行预处理,从而削弱前后两场信号叠加造成的边缘模糊。最后对前后两帧图像进行灰度值相关匹配,并检测出全景图像上的运动物体。实验结果表明,该技术可以实现全景图像拼接并在全景图中标出视场内运动物体的位置。     

Transient viscoelasticity study of tobacco mosaic virus/Ba2+ superlattice

Recently, we reported a new method to synthesize the rod-like tobacco mosaic virus (TMV) superlattice. To explore its potentials in nanolattice templating and tissue scaffolding, this work focused the viscoelasticity of the superlattice with a novel transient method via atomic force microscopy (AFM). For measuring viscoelasticity, in contrast to previous methods that assessed the oscillating response, the method proposed in this work enabled us to determine the transient response (creep or relaxation) of micro/nanobiomaterials. The mathematical model and numerical process were elaborated to extract the viscoelastic properties from the indentation data. The adhesion between the AFM tip and the sample was included in the indentation model. Through the functional equation method, the elastic solution for the indentation model was extended to the viscoelastic solution so that the time dependent force vs. displacement relation could be attained. To simplify the solving of the differential equation, a standard solid model was modified to obtain the elastic and viscoelastic components of the sample. The viscoelastic responses with different mechanical stimuli and the dynamic properties were also investigated.     

Fibre diffraction in the analysis of filamentous virus structure

In many plant and bacterial viruses, a filamentous nucleocapsid makes up the entire virus particle. These viruses usually form fibres naturally, but they do not crystallize. Fibre diffraction has therefore been the method of choice for structural studies, aided in recent years by cryo-electron microscopy (EM) and solid state nuclear magnetic resonance (ssNMR). Filamentous viruses, particularly tobacco mosaic virus (TMV), have been important in developing fibre diffraction methods, and fibre diffraction allowed TMV to be among the first virus structures determined. Structures of several viruses related to TMV and several filamentous bacterial viruses have been determined at resolutions of 3?Å or better, and lower resolution structures have been determined by fibre diffraction, sometimes in conjunction with other methods, for many other, unrelated, filamentous viruses.     

大屏幕显示系统的应用与分析

本文着重对大屏幕显示系统的构架组成进行了介绍,并针对大屏幕显示屏(大屏幕拼接墙)的种类及功能,适用场所等进行了对比分析.     

煤矿井下图像拼接的算法研究

研究煤矿井下图像拼接的算法,探讨井下环境光照不均、强噪声的图像拼接问题。用当前流行的四种特征点检测算法,如Moravec算子、Forstner算子、Harris角点检测算法和CSS角点检测算法提取井下图像的特征点,分析得出CSS角点检测算法更适合井下工作环境的结论。用CSS角点检测算法提取待拼接图片特征点,通过归一化互相关系数(NCC)提取特征点匹配对,并通过RANSAC删除伪匹配对,最终实现无缝拼接。