AIFS—A Tool for Biomorphic Fractal Modeling

A considerable class of fractal sets can be represented by using the attractors of Iterated Function Systems (Barnsley, 1988), with affine contractive mappings of a metric space . The modeling capabilities of such systems are heavily limited however. For example, it is not easy to predict the location of the attractor nor its global shape. Then, Iterated Systems are not affinely invariant (affine mappings of the elements of the system do not result in affine image of its attractor). In this paper a new setting, the affine invariant Iterated Function System is described in such a way that it removes the mentioned shortcomings and can be used for shape-predictable modeling of fractal based forms. The stress is put on modeling of biological forms and their atributes such as: continuous deformation of the attractor in desired way (like in growing), branching (plants, vascular or alveolar network), gradual changing of fractal dimension from smooth to space-filling fractals. The last is useful for creating images of tissues in different stages of development, symmetry, gradual transformation from one to another form, etc. The fractal images obtained by AIFS are merely to gain resemblance to bio-forms.     

基于改进U-Net的量子点STM图像分割

为提升量子点图像分割精度,降低特征识别误差,提出一种基于改进U-Net的量子点图像分割方法.首先,在预处理阶段,设计了以色彩通道为权值的灰度化算法,以提升后续分割效果.其次,在STM图像分割部分,在原始U-Net结构上引入中间过渡层以均衡网络各层特征.而后,建立数据集,并通过实验对比不同分割算法的精确度、召回率、F-measure.最后,将分割算法应用于量子点的特征识别,并测试了不同分割方式对应用的影响.实验结果显示,改进灰度化方法保留细节信息丰富,明显提升了量子点分割精度;改进U-Net的平均精确率、召回率、F-measure相较原始网络分别提升了13.83%、2.16%、8.13%.同时,实验数据表明由于分割精度的提升,量子点数量、纵横比等特征参数的识别更加精确.     

An Interpretation Architecture for Deep Learning Models with the Application of COVID-19 Diagnosis

The Coronavirus disease 2019 (COVID-19) has become one of the threats to the world. Computed tomography (CT) is an informative tool for the diagnosis of COVID-19 patients. Many deep learning approaches on CT images have been proposed and brought promising performance. However, due to the high complexity and non-transparency of deep models, the explanation of the diagnosis process is challenging, making it hard to evaluate whether such approaches are reliable. In this paper, we propose a visual interpretation architecture for the explanation of the deep learning models and apply the architecture in COVID-19 diagnosis. Our architecture designs a comprehensive interpretation about the deep model from different perspectives, including the training trends, diagnostic performance, learned features, feature extractors, the hidden layers, the support regions for diagnostic decision, and etc. With the interpretation architecture, researchers can make a comparison and explanation about the classification performance, gain insight into what the deep model learned from images, and obtain the supports for diagnostic decisions. Our deep model achieves the diagnostic result of 94.75%, 93.22%, 96.69%, 97.27%, and 91.88% in the criteria of accuracy, sensitivity, specificity, positive predictive value, and negative predictive value, which are 8.30%, 4.32%, 13.33%, 10.25%, and 6.19% higher than that of the compared traditional methods. The visualized features in 2-D and 3-D spaces provide the reasons for the superiority of our deep model. Our interpretation architecture would allow researchers to understand more about how and why deep models work, and can be used as interpretation solutions for any deep learning models based on convolutional neural network. It can also help deep learning methods to take a step forward in the clinical COVID-19 diagnosis field.     

Separable convolution template (SCT) background prediction accelerated by CUDA for infrared small target detection

This paper presents a novel background prediction method for infrared small target detection (ISTD). Using a separable convolution template (SCT) to accelerate the traditional background prediction by graphic processing unit (GPU), the new method provides a significant improvement in the prediction speed, which enables the prediction process in real time. And experimental results show its high efficiency and practical application over previous work. The mathematical approach proposed here could be extended to accelerate the applications referred to image convolutions not only to the infrared field.     

基于不变像元灰度分类的自动辐射归化处理

在土地利用／覆盖变化监测中,通常要对多时相遥感影像进行辐射归化处理,使得影像间的成像差异减小,相同地物的光谱特征相似．本研究利用两时相ETM＋影像,精确选取不变像元,采用分段线性回归,探讨一种自动辐射归化处理方法．研究方法充分考虑了影像间成像条件的差异,通过影像对应波段差值直方图剔除变化像元来有效选取不变像元,将不变像元划分为暗、灰和亮3类并分级确定校正系数进行辐射归化处理．结果分析表明该方法具有一定优势．     

Structural evolution,growth mechanism and photoluminescence properties of CuWO4 nanocrystals

Copper tungstate (CuWO₄) crystals were synthesized by the sonochemistry (SC) method, and then, heat treated in a conventional furnace at different temperatures for 1 h. The structural evolution, growth mechanism and photoluminescence (PL) properties of these crystals were thoroughly investigated. X-ray diffraction patterns, micro-Raman spectra and Fourier transformed infrared spectra indicated that crystals heat treated and 100 °C and 200 °C have water molecules in their lattice (copper tungstate dihydrate (CuWO₄·2H₂O) with monoclinic structure), when the crystals are calcinated at 300 °C have the presence of two phase (CuWO₄·2H₂O and CuWO₄), while the others heat treated at 400 °C and 500 °C have a single CuWO₄ triclinic structure. Field emission scanning electron microscopy revealed a change in the morphological features of these crystals with the increase of the heat treatment temperature. Transmission electron microscopy (TEM), high resolution-TEM images and selected area electron diffraction were employed to examine the shape, size and structure of these crystals. Ultraviolet–Visible spectra evidenced a decrease of band gap values with the increase of the temperature, which were correlated with the reduction of intermediary energy levels within the band gap. The intense photoluminescence (PL) emission was detected for the sample heat treat at 300 °C for 1 h, which have a mixture of CuWO₄·2H₂O and CuWO₄ phases. Therefore, there is a synergic effect between the intermediary energy levels arising from these two phases during the electronic transitions responsible for PL emissions.     

Image sharing scheme based on combination theory

We present a simple algorithm for sharing and hiding secret image based on combination theory. The secret image is firstly encrypted by matrix multiplications and then shared into many shadow images by multiplying binary random sampling matrices. The sampling matrices randomly assign the pixel values to the shadow images which satisfy a specific combination rule as a constrain, so that the (t, n) threshold secret sharing scheme can be implemented. Numerical experiments have demonstrated the effectiveness of this image sharing algorithm.     

自适应双边滤波红外弱小目标检测方法

针对红外弱小目标检测,提出一种基于自适应双边滤波的背景预测算法.该算法利用空域低通滤波和图像灰度信息的非线性组合,自适应的对背景进行预测,达到提高弱小目标检测性能的目的.仿真和实验表明：与小波滤波的检测算法相比,该算法能够更加有效地从结构化背景中检测目标抑制背景.     

DISCRETE GRADIENT METHOD FOR IMAGE-BASED BIOMECHANICAL STRESS ANALYSIS 1)

介绍一种可用于计算生物力学的离散梯度方法,此方法可利用离散的点云模型直接进行数值模拟分析而不需要传统的几何模型。将离散梯度法应用于点云模型需要首先确定模型中点之间的相邻关系和每个点所分配的材料体积,然后通过用广义的有限差分的形式定义了梯度插值向量,并以此向量来近似函数在每个离散点上的梯度。从弱形式出发,推导建立了适用于弹性固体大变形问题的求解器,并具有和有限元法中双线性四边形单元一致的准确性和收敛性。着重描述了一种可以从医学图像中快速提取材料点并建立点云模型的方法,以及利用三角划分和重心划分确定材料点之间的相邻关系和每个材料点体积的具体过程。通过腹主动脉瘤膨胀的静力学模拟分析,展示了离散梯度法的实用性和准确性。该算法实现了基于医学图像进行生物力学分析的过程自动化,为病体特异性的研究和治疗提供便利和实用的工具。