基于非线性扩散的纹理分割(英文)

A texture image segmentation based on nonlinear diffusion is presented. The scale of texture can be measured during the process of nonlinear diffusion. A smooth 5-channel vector image with edge preserved, which is composed of intensity, scale and orientation of texture image, can be achieved by coupled nonlinear diffusion. A multi-channel statistical region active contour is employed to segment this vector image. The method can be seen as a kind of unsupervised segmentation because parameters are not sensitive to different texture images. Experimental results show its high efficiency in the semiautomatic extraction of texture image.     

An Operational Method for Fast Detecting Abnormal Channels in Imaging Spectrometers

Data from abnormal Channels in an imaging spectrometer almost always exerts an undesired impact on spectrum matching,chassification,pattern recognition and other applications in hyperspectral remote sensing.To solve this problem.researchers should get rid of the data acquired by these channels.Selecting abnormal channels just in the way of visually examining each band image in a imaging data set is a conceivably hard and boring job.To relieve the burden,this paper proposes a method which exploits the spatial and spectral autocorrelations inherent in imaging spectrometer data,and can be used to speed up and ,to a great degree,automate the detection of abnormal channels in an imaging spectrometer.This method is applied easily and successfully to one PHI data set and one Hymap data set ,and can be applied to remotely sensed data from other hyperspectral sensors.     

The study on automatic registration of InSAR images

Interferometric Synthetic Aperture Radar (InSAR) allows production of high resolution DEM and detection of small earth motions using multiple pass SAR data sets obtained by remote sensing satellite. But the whole process has not yet reached sufficient robustness to warrant automated DEM production as commonly produced by stereovision with optical images. The automatic algorithm for precision registration is one of the bottlenecks in InSAR data processing. In this paper, an automatic approach with multi-step image matching algorithm is presented. All procedures are automatically implemented. The experiment is carried out successfully with SIR-C/L-band InSAR data. The triangular piecewise rectification is also advanced in reducing local distortion between the images and processing the large scene image. The primary result has prospect in the precision registration for the repeat-track InSAR data and reveals the potential of the presented automatic strategy.     

New mathematical model based on affine transformation for remote sensing image with high resolution

This paper calculates the parameters of image position and orientation,proposes a mathematical model and adopts a mew method with three steps of transformations based on parallel ray projection.Every step of the model is strict,and the map function of each transformation is the first order polynomials and other simple function.The final calculation of the parameters is for the linear equation with good status.As a result,the problem of the relativity of image parameter calculation is solved completely.Some experiments are carried out.     

SAR image classification based on its texture features

SAR images not only have the characteristics of all-ay, all-eather, but also provide object infor-mation which is different from visible and infrared sensors. However, SAR images have some faults, such as more speckles and fewer bands. The au-thors conducted the experiments of texture statistics analysis on SAR im-age features in order to improve the accuracy of SAR image interpretation.It is found that the texture analysis is an effective method for improving the accuracy of the SAR image interpreta-tion.     

Rank-Defect Adjustment Model for Survey-Line Systematic Errors in Marine Survey Net

In this paper,the structure of systematic and random errors in marine survey net are discussed in detail and the adjustment method for observations of marine survey net is studied,in which the rank-defect haracteristic is discovered first up to now.On the basis of the survey-line systematic error model,the formulae of the rank-defect adjustment model are deduced according to modern adjustment theory.An examle of calculations with really observed data is carried out to demonstrate the efficiency of this adjustment model.Moreover,it is proved that the semi-systematic error correction method used at present in marine gravimetry in China is a special case of the adjustment model presented in this paper.     

Fast algorithm for maneuvering target detection in SAR imagery based on gridding and fusion of texture features

Designing detection algorithms with high efficiency for Synthetic Aperture Radar(SAR) imagery is essential for the operator SAR Automatic Target Recognition(ATR) system.This work abandons the detection strategy of visiting every pixel in SAR imagery as done in many traditional detection algorithms,and introduces the gridding and fusion idea of different texture fea-tures to realize fast target detection.It first grids the original SAR imagery,yielding a set of grids to be classified into clutter grids and target grids,and then calculates the texture features in each grid.By fusing the calculation results,the target grids containing potential maneuvering targets are determined.The dual threshold segmentation technique is imposed on target grids to obtain the regions of interest.The fused texture features,including local statistics features and Gray-Level Co-occurrence Matrix(GLCM),are investigated.The efficiency and superiority of our proposed algorithm were tested and verified by comparing with existing fast de-tection algorithms using real SAR data.The results obtained from the experiments indicate the promising practical application val-ue of our study.     

Filter Method and Its Consistency of Double-Star Position/SINS Integrated System

According to the characteristic of Beidou Double-star positing system (for short: Double-star position), the optimal predication model of Double-star position/SINS integrated system is put forward, which can make use of the delayed position in-formation from Double-star positioning system to predicate optimally for the integrated system, and then to correct SINS, and af-fords integrated results of some navigation parameters. In order to validate the consistency of the filter, the criteria for consistency of a filter is also studied, and the tested statistics are given, the experiment based on practical measured data shows that the filtering method is consistent with the integrated system.     

Deblurring Texture Extraction from Digital Aerial Image by Reforming “Steep Edge” Curve

Texture extract from digital aerial image is widely used in three-dimensional city modeling to generate “photo-realistic” views. In this paper, a method based on reforming “Steep edge” curve, which clearly explains how the diffraction of the sunlight makes digital aerial image blurring, is proposed to deblur the texture extraction from digital aerial image, and the experiment shows a good result in visualization and automation.     

Principal component analysis in application to object orientation

This paper proposes a new method based on principal component analysis to find the direction of an object in any pose.Experiments show that this method is fast,can be applied to objects with any pixel distribution and keep the original properties of objects invariant.It is a new application of PCA in image analysis.     

一种运用纹理和光谱特征消除投影差影响的建筑物变化检测方法

针对不同时期高分辨率遥感影像变化检测中城区建筑物因投影差差异所产生的误检测现象,提出了一种综合应用光谱和纹理特征的建筑物变化检测方法。以变化和未发生变化地物影像的散度作为可分性依据,首先对光谱差分影像在混合高斯密度模型下建模,并采用马尔可夫最小错误概率准则提取初始变化区域,往往含有错判的建筑物。然后将误判建筑物影像类和真实变化影像类构成训练集,通过引入多通道Gabor滤波器,提取训练集的纹理差分特征,并采用分类别PCA变换实施纹理差分特征的选择。最后对选择出的纹理差分特征依据高斯混合密度模型建模,并用马尔可夫最小错误概率提取真变化区域,即可去除光谱信息检测所产生的伪变化。试验表明,本文方法能够较好地解决建筑物变化的错判问题,提高了影像变化检测的精度。     

结合光谱和纹理的高分辨率遥感图像分水岭分割

针对遥感图像分割时仅利用光谱信息容易造成过分割和边缘定位不准的问题,提出一种结合光谱强度和纹理信息的遥感图像分水岭分割算法。首先分别提取图像的光谱梯度和纹理梯度,提出一种改进双边滤波模型,滤除图像中的噪声的周时,采用了一种局部的平滑尺度,能够有效消除纹理信息,借助于滤波算法,分别对原图像和Gabor纹理特征图像进行平滑处理,利用边缘检测算子得到光谱梯度和纹理梯度。最后利用形态学膨胀方法进行融合融合,使用分水岭变换对图像分割。用三幅高分辨率彩色遥感图像数据进行实验,并与JSEG（Joint Systems Engineering Group）和多分辨率分割方法进行比较,结果表明该方法具有较高的边界定位准确性,同时降低了过分割和欠分割现象。     

一种基于概率潜在语义模型的高分辨率遥感影像分类方法

针对高分辨率遥感影像中"同谱异物","同物异谱"现象对影像分类过程造成的干扰,将文本分析中的概率潜在语义模型应用于高分辨率遥感影像分类,提出一种无监督的遥感影像分类新方法.该方法首先利用均值漂移分割方法对影像进行分割构建图像区域集合,然后提取集合各区域中每个像元的Gabor纹理特征,并对这些特征进行聚类形成视觉词汇,最...     

纹理图像统计模型与纹理图像分割

本文详细介绍了一种新的纹理图像统计模型－Ｇｉｂｂｓ模型，给出了两种常用的Ｇｉｂｂｓ分布。在此基础上作者详细推导出了两种纹理图像的分割算法，并用一幅遥感图像进行了实验，获得了较好的分割结果。     

高分辨率遥感影像变化检测的频域分析预处理方法

目前,大多数高分辨率遥感影像变化检测方法的效率不高,精度也难以达到实际应用要求,本文引入频域分析的方法进行变化检测预处理,达到提高变化信息检测精度和效率的目标。首先,对影像频域能量的分布特征进行分析,并以此为基础,利用累加距离匹配函数,完成试验影像的合理分块;其次,通过变化性纹理规则度的定义和计算,实现对同能量下变化信息的精确描述,缩小变化区域搜索范围,有效抑制背景信息对目标信息的干扰。最后,通过3种常用的变化检测方法的验证,发现变化检测预处理方法可以使检测算法在现有精度和效率的基础上有进一步提高。     

A New Approach to Improve the Cluster-based Parallel Processing Efficiency of High-Resolution Remotely Sensed Image

Object-oriented remotely sensed images processing method has been accepted by more and more experts of remote sensing. To advance the efficiency of data processing, parallel image computing is a good choice since large volumes of data need be analyzed efficiently and rapidly. This paper presents the information extraction method based on per-parcel extraction of high-resolution remotely sensed image; to extract efficiently different information from remotely sensed image, this paper gives the research idea of image rough-classification based on large-scale and subtle-segmentation based on small-scale; to improve the efficiency of image processing, we adapt parallel computing method to solve this problem by presenting an new data-partition method. At last this paper gives the implementation of the research idea based on Message Passing Interface (MPI) and analyzes our experimental system efficiency, and the results show that the new methods can improve the efficiency of high-resolution remotely sensed image data processing efficiently and have a good application.     

基于频域滤波的高分辨率遥感图像城市河道信息提取

提出一种基于频域滤波的城市河道信息提取方法。首先对高分辨率遥感图像进行傅里叶变换得到频谱图, 并利用径向和角向分布图分析城市河道的频谱特征。其次, 基于城市河道的双线型特点, 将其分为边缘特征和低频信息两个部分, 并根据周期性纹理的频谱模型和地物频谱能量分布规律确定两个部分的频域识别标志。然后设计相应的扇环形带通log Butterworth滤波器和低通Butterworth滤波器分别对城市河道的边缘特征和低频信息进行提取, 并根据该两部分信息实现城市河道信息提取。最后对城市河道信息提取结果进行定量评价, 结果表明, 本文方法可以有效地实现城市河道的信息提取。     

利用SVM相关反馈和语义挖掘的遥感影像检索

针对语义鸿沟问题,将基于语义特征挖掘模型与支持向量机相关反馈方法相结合,建立了基于支持向量机相关反馈的人机交互遥感影像语义检索系统。实验结果表明,该方法利用高层语义特征及人机交互反馈信息缩小了语义鸿沟,提高了影像检索的精度。     

一种基于相位一致的高分辨率遥感图像特征检测方法

精确检测图像边缘特征是进行高分辨率遥感图像分割和识别的关键。空域特征检测算子以解决阶跃形边缘为主,得到的边缘特征对图像的亮度和对比度敏感。本文引入了一种基于频域相位一致的图像特征检测方法,该方法对遥感图像亮度和对比度具有不变性,同时适用于多种边缘特征的检测。使用Log Gabor小波计算IKONOS Pan图像的相位一致多尺度梯度,对农田、道路和厂房等典型地物进行特征检测的结果表明,相位一致算法对图像局部亮度和对比度不敏感;并且对线形物体产生单线响应,不似空域检测算子产生双线响应。最后考察滤波器尺度和方向参数变化及添加高斯噪声对检测结果的影响,发现相位一致算法无需先使用低通滤波去除噪声,因而具有更稳定的特征定位精度;并且抗噪声干扰的能力强,检测结果不会因为噪声而出现波动。基于相位一致的遥感图像不变特征提取,为高分辨率遥感图像的分割和对象识别提供了基础。     

粗糙集高分辨率遥感影像面向对象分类

面向对象的高分辨率遥感影像分类已受到研究者们的广泛关注。本文提出一种基于粗糙集理论的面向对象分类方法以区分高分辨率遥感影像上的不同地物。首先,利用基于相位一致梯度与前景标记的分水岭变换进行影像分割,提取图像斑块;然后,利用Gabor小波提取斑块的纹理特征,进而根据粗糙集理论提取纹理分类规则;最后,在对象光谱特征的初步分类结果,根据纹理分类规则得到最终结果基础上。依据粗糙集理论只能处理离散属性数据,本文重点提出一种适用于面向对象分类的连续区间属性离散化方法。实验表明本文方法可取得较好分类结果与较高分类精度。