基于对数相似度约束Otsu的自然场景病害果实图像分割

针对自然场景下,由于复杂背景以及多变环境,水果病害果实图像分割难的问题,提出了一种基于对数相似度约束Otsu和水平集活动轮廓的近椭圆形病害果实图像分割方法。考虑背景的复杂多变,提出对数相似度约束Otsu分割来区分病害果实与背景;由于水平集活动轮廓模型的局部最优性,提出采用自适应膨胀系数的改进距离规则水平集活动轮廓模型来精确演化轮廓。先对病害果实区域样本的颜色进行混合高斯建模,获得整个病害果实图像与样本模型的对数相似度;对对数相似度进行约束Otsu阈值分割以及形态学滤波;采用最小二乘法对滤波后的曲线轮廓进行椭圆拟合,对拟合后的椭圆采用自适应膨胀系数的距离规则水平集活动轮廓演化,得到病害果实完整轮廓。对18个不同场景的病害果实进行分割,平均误判率和漏判率分别为1.77%和1.6%,实验结果表明,该方法可以从复杂自然场景图像中分割出病害果实。     

最优控制问题的Tikhonov正则化

主要研究了约束线性二次最优控制问题.通过一阶最优性条件将它等价地转化为单调变分不等式问题,并利用变分不等式的Tikhonov正则化方法研究了约束线性二次最优控制问题的正则化,证明了扰动问题的解收敛到原问题的最小范数解.     

Application of soft sensor in the on line measurement of theiron ore’s surface temperature

It is difficult to measure the surface temperature of iron ore directly. A method is put forward to handle this problem by using soft sensing technique. This on line measurement method is used to replace the Lagrange interpolation off line method to estimate surface temperature. The method used L M optimum algorithm to build up ANN soft sensor model combined with off line learning neural network to establish the correlation between input variables and target variables, to achieve the surface temperature on line detection. The results of simulation and experimentation indicate that the method is reasonable and feasible.     

基于EM38的土壤剖面电导率预测研究

【目的】以海涂围垦区盐碱土为研究对象,利用EM38大地电导率仪在地表不同高度测量的土壤表征电导率预测土壤不同深度土层剖面的电导率。【方法】利用EM38电导率线性响应模型结合Tikhonov正则化能较好解决病态矩阵的线性反演问题。利用该方法来预测土壤剖面电导率,并对预测结果进行误差分析,最后通过偏差变化法增加噪声利用局部灵敏度分析法来评价模型的灵敏度。【结果】研究发现,土壤剖面平均电导率和地表不同高度平均表征电导率具有极显著相关性,能利用表征电导率较好的预测剖面平均盐分。线性模型不仅能较好的预测土壤剖面电导率的变化趋势,而且在数值上也能较好的预测电导率的大小,平均预测误差在40%左右。相对预测误差较大的土层,模型灵敏度越大,因此可通过提高EM38数据测量的稳定性来提高预测精度。【结论】利用EM38表征电导率采用线性模型结合Tikhonov正则化的方法能够较好的反演土壤剖面电导率,预测结果可以为样区土壤管理提供科学的决策依据。     

Computation method of magnetic field inverse problem on grounding grids fault diagnosis

Grounding grids are the guarantee of substation running safety. The fault of grounding grids is a threat to people and equipment in substations. A method to diagnose the fault of grounding grids based on magnetic field inverse problem. The inverse problem of magnetic field is established by injecting and extracting a low frequency current. The morbid of inverse problem is solved through regularization and Newton iteration method to obtain the distribution of the mesh currents. The approximate distribution of the magnetic field is calculated to diagnose the fault of the grounding grid. A 4×4 substation grounding grids is taken as an example. By comparison of the forward problem result, the error is under 3%, so the prposed method of inverse problem is verified. The selection of measurement point is discussed, and 40% of the number of meshes is accounted to approximate the real distribution of the magnetic field on the grounding grids.     

Image denoising based on adaptive graph regularization

Adaptive regularization can select different parameters based on the features of local areas in an image, which can differentiate the edges and noise in an image flexibly. An adaptive graph regularization is proposed based on graph spectral theory and adaptive regularization, which uses the Non local means to generate the weighting function of graph. The adaptive graph regularization equation is used to filter the noisy image. Simulation results show that the proposed method can effectively remove the noise and is superior to other graph theory based partial differential equation methods.     

离散广义系统的正则化及其在经济中的应用

A regular system for soluble generlized systems is obtained using the regulariiation method. It is proved that the state variables of soluble generalized systems must satisfy certain constraint conditions. The solution of soluble generalized systems can b     

Temperature dependence of relaxation spectra for highly hydrated gluten networks

In the present investigation, the temperature dependence (0–50 °C) of the relaxation spectrum of hydrated gluten was studied using novel numerical algorithms. Tikhonov regularization, in conjunction with the L-curve criterion for optimal calculation of the regularization parameter, was used to generate the relaxation spectrum from stress relaxation measurements on shear. The methodology used revealed six molecular events with baseline resolution that could be grouped into fast- and slow-relaxation regimes. The fast-relaxation regime exhibited strong temperature dependence whereas the slow one is temperature independent indicating on the whole two dominant mechanisms of interactions. The “loop and train” structural model for gluten interactions was found adequate to describe the relaxation events in this system, with the fast regime being assigned to interactions due to hydrogen bonding whereas the slow one to permanent cross-linking of the entire network. Findings of the present investigation provide fundamental understanding and give new insights into the complexity of interactions and relaxation modes of hydrated gluten.     

基于纹理差异度引导的DRLSE病虫害图像精准分割方法

为精准分割纹理背景下的病虫害目标,提出一种基于混合高斯模型的纹理差异度表示方法和基于纹理差异度引导的DRLSE农作物病虫害精准分割方法。采用基于混合高斯模型的纹理差异度表示方法得到纹理差异度图像,利用Otsu自动阈值分割和形态学后处理得到农作物病虫害DRLSE演化的初始区域,将基于改进纹理差异度的边缘停止函数应用到DRLSE演化中。实验结果表明该方法可以获得精准的害虫和病变区域轮廓。     

一种基于实测模态参数的结构损伤诊断方法

本文提出了一种利用实测结果固有频率与振型诊断结构损伤大小和位置的方法，并提出了一种能抑制测试误差对诊断准确性影响的正则化技术求解质量和刚度损伤诊断方程，以三根模型的损伤诊断实例说明了本方法的有效性。