一种工业CT图像的分割算法

由于工业CT图像结构复杂,存在各种伪影,并且灰度分布呈现区域性质等特点,难以准确找出分割阈值,为此提出了一种适用于工业CT图像的分割算法。首先利用最大类间方差法和图像处理方法处理了外层伪影,然后利用聚类迭代的方法处理中心空气,得到感兴趣的区域。实验结果表明,此算法能够对具有先验知识的工业断层图像准确地提取感兴趣的区域。     

基于阈值分割与边缘检测的对象轮廓提取方法研究

为了提高智能视频监控系统中对象检测算法的检测准确性,实现对检测对象轮廓的准确提取,在分析目前常用于获取对象轮廓形态的对象检测方法不足的基础上,提出了基于阈值分割与边缘检测的对象轮廓提取方法。该方法需要阈值而又不依赖于阈值,选取任一阈值对检测对象进行阈值分割,再结合Sobel边缘检测以及经过定制的边界跟踪算法,实现对检测对象轮廓的提取。经实验得出的轮廓检测结果在不同阈值的条件下都呈现出较好的完整性与一致性。因此,方法具有较好的鲁棒性,实现了对检测对象轮廓的完整提取,提高了对象检测算法的检测准确性。     

高空间分辨率遥感影像多尺度分割优化组合算法

高空间分辨率遥感影像能够充分地描述地表覆盖空间异质性,可用于提取地面目标物。然而高空间分辨率在像元尺度的目标提取时易产生"椒盐效应"问题,面向对象的小尺度影像分割也受此效应影响;而大尺度的影像分割造成较小目标的遗漏。本文提出了一种针对高空间分辨率遥感影像的多尺度分割优化组合算法MOCA(Multi-scale-segmentation Optimal Composition Algorithm),基于后验概率信息熵指标选择影像中每个地面目标的最优分割尺度并集成组合,获得高空间分辨率遥感影像的多尺度分割优化组合结果。本文使用F指标和BCI(Bidirectional Consistency Index)两种指标评估地面目标物提取精度,并将MOCA与同类多尺度分割方法进行比较。实验结果表明,本文提出的MOCA算法可实现多个分割尺度的最优组合,并获得较高的地面目标物提取精度。     

结合均值漂移分割与聚类分析的遥感影像变化检测

针对传统遥感影像变化检测方法对前后期影像数据质量要求高、适应范围窄、检测精度较低等问题,本文在引入异常点检测思想的基础上,提出了一种结合均值漂移分割与聚类分析的遥感影像变化检测方法。首先将前期地理国情矢量数据与待监测的遥感影像进行地理配准;然后在地理国情矢量数据的基础上对遥感影像作均值漂移算法二次细分割,获得的矢量图斑继承了原有父级类属性,并对同一父级类的影像图斑进行光谱、空间、纹理、指数等特征提取;最后采用高斯混合模型与最大期望值算法聚类获得多个类别,在父级类下找出异常类别的图斑。通过试验对比分析,表明了本文方法的有效性和可靠性,为遥感影像变化检测提供了新思路。     

Defining a model of 3D seismogenic sources for Seismic Hazard Assessment applications: The case of central Apennines (Italy)

Geology-based methods for Probabilistic Seismic Hazard Assessment (PSHA) have been developing in Italy. These methods require information on the geometric, kinematic and energetic parameters of the major seismogenic faults. In this paper, we define a model of 3D seismogenic sources in the central Apennines of Italy. Our approach is mainly structural-seismotectonic: we integrate surface geology data (trace of active faults, i.e. 2D features) with seismicity and subsurface geological–geophysical data (3D approach). A fundamental step is to fix constraints on the thickness of the seismogenic layer and deep geometry of faults: we use constraints from the depth distribution of aftershock zones and background seismicity; we also use information on the structural style of the extensional deformation at crustal scale (mainly from seismic reflection data), as well as on the strength and behaviour (brittle versus plastic) of the crust by rheological profiling. Geological observations allow us to define a segmentation model consisting of major fault structures separated by first-order (kilometric scale) structural-geometric complexities considered as likely barriers to the propagation of major earthquake ruptures. Once defined the 3D fault features and the segmentation model, the step onward is the computation of the maximum magnitude of the expected earthquake (M _max). We compare three different estimates of M _max: (1) from association of past earthquakes to faults; (2) from 3D fault geometry and (3) from geometrical estimate corrected by earthquake scaling laws. By integrating all the data, we define a model of seismogenic sources (seismogenic boxes), which can be directly used for regional-scale PSHA. Preliminary applications of PSHA indicate that the 3D approach may allow to hazard scenarios more realistic than those previously proposed.     

粗糙集理论在遥感影像分割中的应用

结合粗糙集理论和K—均值聚类算法,提出一种遥感影像的粗糙聚类分割方法。根据遥感影像中特征属性的相互依赖关系,应用粗糙集理论的等价关系,求出K—均值聚类所需要的初始类的个数和均值,然后采用聚类算法对图像进行分割。实验结果表明该方法比随机选取聚类的中心点和个数减少了运算量,提高了分类精度和准确性。     

基于直线和区域特征的遥感影像线状目标检测

针对高分辨率航空遥感影像中线状目标的特点,提出一种结合区域和直线特征识别线状目标的方法。在基于标记点分水岭变换进行初始分割的基础上,利用关于目标的知识和区域邻接图(RAG)对感兴趣区域进行合并,得到最终检测结果。实验结果表明,本文方法可以有效地从遥感影像中提取线状目标。     

用简化脉冲耦合神经网络对图像进行二值分割

结合图像二值分割的特点,对脉冲耦合神经网络算法进行了一定的简化和改进,通过将其应用到航空遥感影像的二值分割中,发现其具有良好的分割效果,特别是在具有强噪声和阴影影响的影像分割中更显示其优于传统的算法。     

Automated Sizing of Coarse-Grained Sediments: Image-Processing Procedures

This is the first in a pair of papers in which we present image-processing-based procedures for the measurement of fluvial gravels. The spatial and temporal resolution of surface grain-size characterization is constrained by the time-consuming and costly nature of traditional measurement techniques. Several groups have developed image-processing-based procedures, but none have demonstrated the transferability of these techniques between sites with different lithological, clast form and textural characteristics. Here we focus on image-processing procedures for identifying and measuring image objects (i.e. grains); the second paper examines the application of such procedures to the measurement of fluvially deposited gravels. Four image-segmentation procedures are developed, each having several internal parameters, giving a total of 416 permutations. These are executed on 39 images from three field sites at which the clasts have contrasting physical properties. The performance of each procedure is evaluated against a sample of manually digitized grains in the same images, by comparing three derived statistics. The results demonstrate that it is relatively straightforward to develop procedures that satisfactorily identify objects in any single image or a set of images with similar sedimentary characteristics. However, the optimal procedure is that which gives consistently good results across sites with dissimilar sedimentary characteristics. We show that neighborhood-based operations are the most powerful, and a morphological bottom-hat transform with a double threshold is optimal. It is demonstrated that its performance approaches that of the procedures giving the best results for individual sites. Overall, it out-performs previously published, or improvements to previously published, methods.     

基于云模型的图像分割方法研究

分析了图像信息的不确定性、云模型的理论和方法,提出了一种基于云模型的图像分割新方法,实现了图像分割,取得了良好效果。