Reconstructing granular particles from X-ray computed tomography using the TWS machine learning tool and the level set method

X-ray computed tomography (CT) has emerged as the most prevalent technique to obtain three-dimensional morphological information of granular geomaterials. A key challenge in using the X-ray CT technique is to faithfully reconstruct particle morphology based on the discretized pixel information of CT images. In this work, a novel framework based on the machine learning technique and the level set method is proposed to segment CT images and reconstruct particles of granular geomaterials. Within this framework, a feature-based machine learning technique termed Trainable Weka Segmentation is utilized for CT image segmentation, i.e., to classify material phases and to segregate particles in contact. This is a fundamentally different approach in that it predicts segmentation results based on a trained classifier model that implicitly includes image features and regression functions. Subsequently, an edge-based level set method is applied to approach an accurate characterization of the particle shape. The proposed framework is applied to reconstruct three-dimensional realistic particle shapes of the Mojave Mars Simulant. Quantitative accuracy analysis shows that the proposed framework exhibits superior performance over the conventional watershed-based method in terms of both the pixel-based classification accuracy and the particle-based segmentation accuracy. Using the reconstructed realistic particles, the particle-size distribution is obtained and validated against experiment sieve analysis. Quantitative morphology analysis is also performed, showing promising potentials of the proposed framework in characterizing granular geomaterials.

     

Fine classification of construction land using high-resolution remote sensing images: a case study in planning restricted zone of nuclear power plant

Detailed construction land information plays a significant role in monitoring planning restricted zone of nuclear power plant and ecological environment protection. This study focuses on developing fine classifying method of construction land in planning restricted zone of nuclear power plant using high spatial resolution GF(GaoFen)-1 remote sensing images. The object-oriented classification method is used in this study; the important process of which is image segmentation and classification. Multi-scale segmentation method, rule-based decision tree, and the nearest neighbor classifier are used in classifying construction land classes, i.e., road, industrial, and residential. An optimal segmentation scale is crucial to image segmentation in object-oriented classification. Instead of laborious trial-and-error experiments for optimal image segmentation, the change rates of the local variance in the homogeneous region are calculated to get the optimal segmentation scales. Multi-level classification strategy is used in the following classification. Rule-based decision tree is used to classify road and water, vegetation and non-vegetation, and industrial and residential. And the nearest neighbor classifier is used to classify cropland and forest within the vegetation land use type. The accuracy assessment result shows that the overall accuracy is 89.67% and Kappa coefficient is 0.85 for object-oriented classification, which is much higher than pixel-based maximum likelihood classifier (overall accuracy is 79.17% and Kappa coefficient is 0.74) and support vector machine classifier (overall accuracy is 74.16% and Kappa coefficient is 0.68).     

面向对象的遥感图像分类方法研究

影响遥感图像分类效果的主要因素之一是空间分辨率。通过融合多分辨率遥感图像,引入面向对象的思想,有效地克服了多光谱图像空间分辨率低的问题。该方法由图像分割和分类等一系列技术组成,首先用基于区域分割法则对正射校正SPOT图像进行分割,然后把它作为参考用最大似然法分类器和其他一些经验规则对TM图像进行分类。对土地覆盖图分类进行精度测试,取得了良好的应用效果。     

Automated Kerogen Classification in Microscope Images of Dispersed Kerogen Preparation

We develop the classification part of a system that analyses transmitted light microscope images of dispersed kerogen preparation. The system automatically extracts kerogen pieces from the image and labels each piece as either inertinite or vitrinite. The image pre-processing analysis consists of background removal, identification of kerogen material, object segmentation, object extraction (individual images of pieces of kerogen) and feature calculation for each object. An expert palynologist was asked to label the objects into categories inertinite and vitrinite, which provided the ground truth for the classification experiment. Ten state-of-the-art classifiers and classifier ensembles were compared: Naïve Bayes, decision tree, nearest neighbour, the logistic classifier, multilayered perceptron (MLP), support vector machines (SVM), AdaBoost, Bagging, LogitBoost and Random Forest. The logistic classifier was singled out as the most accurate classifier, with an accuracy greater than 90. Using a 10 times 10-fold cross-validation provided within the Weka software, we found that the logistic classifier was significantly better than five classifiers (p<0.05) and indistinguishable from the other four classifiers. The initial set of 32 features was subsequently reduced to 6 features without compromising the classification accuracy. A further evaluation of the system alerted us to the possible sensitivity of the classification to the ground truth that might vary from one human expert to another. The analysis also revealed that the logistic classifier made most of the correct classifications with a high certainty.     

An unsupervised multi-scale segmentation method based on automated parameterization

The continuous improvement of the launched satellites’ spatial and spectral resolutions has brought new challenges for remote sensing image segmentation technology. The traditional supervised methods greatly depend on artificial interpretation and reduce the degree of automation and robustness of image segmentation. Therefore, the article proposes a novel unsupervised multi-scale segmentation method for high-resolution remote sensing images based on automated parameterization and it mainly includes three steps, adaptive selection of scale parameter (SP) based on local area homogeneity index J-value, multi-scale segmentation based on the inter-scales boundaries constraint strategy, and region merging based on multi-features. The article makes experiments by multi-group high-resolution remote sensing images of different launched satellites and compares the proposed method with the well-known commercial software eCognition and a traditional supervised method. The results show that the proposed method can locate the object edges more accurately and extract the object outlines more completely, and needs no human intervention in segmentation process, so it can provide a generic and effective unsupervised solution for high-resolution remote sensing image segmentation.     

基于国产高分辨率光学遥感影像的水体提取

遥感图像中地表水体同山体、建筑物等地物产生的阴影在光谱特征上存在较高的类间相似性,导致提取过程中容易出现混淆和错分的情况。针对此问题,提出一种基于面向对象和人工蜂群的地表水体提取方法。该方法首先对遥感图像进行分割以获取分割对象的光谱、比率、几何形状等统计特征,以弥补高分遥感图像波段数目少,信息量不足的缺陷;并借助人工蜂群算法在解决复杂问题最优化方面的优势,选取水体同阴影二值分类的几何平均正确率作为算法的适应度函数,最终获取地表水体的最优化提取规则。选取厦门市大嶝岛和湖南省资兴市部分区域,基于国产高分一号、二号遥感数据进行水体提取,并与传统SVM分类结果进行比较。实验结果表明本算法提取水体的总体精度和Kappa系数均优于传统SVM分类器,表明该方法可应用于高分遥感图像的地表水体提取。     

A new algorithm using image colour system transformation for rock grain segmentation

Summary The present paper is concerned with the development and application of an effective automatic algorithm of image analysis in order to detect grain boundaries on microscope images of Redziny dolostone and Wisniowka quartzite. The algorithm utilises sets of 6 colour images for each measurement field on thin sections, which are recorded using an optical polarizing microscope in different polarization set-ups. The proposed method is based on an image pre-processing procedure that is focused on colour system transformation, followed by rock grain boundaries segmentation using the image analysis methods for each colour image. For the image pre-processing procedure, several colour system transformations were selected and compared. By using the alternative colour systems that concentrate on colour information we are able to minimise the effects of internal micro-structures in the grain boundaries segmentation procedure. The grain boundary maps obtained confirm that the use of an approximately perceptually uniform colour system as an image pre-processing procedure can significantly improve the rock grain segmentation. This newly-developed method may facilitate petrographical and stereological studies of rock structures.     

Porous Structure Reconstruction Using Convolutional Neural Networks

The three-dimensional high-resolution imaging of rock samples is the basis for pore-scale characterization of reservoirs. Micro X-ray computed tomography (µ-CT) is considered the most direct means of obtaining the three-dimensional inner structure of porous media without deconstruction. The micrometer resolution of µ-CT, however, limits its application in the detection of small structures such as nanochannels, which are critical for fluid transportation. An effective strategy for solving this problem is applying numerical reconstruction methods to improve the resolution of the µ-CT images. In this paper, a convolutional neural network reconstruction method is introduced to reconstruct high-resolution porous structures based on low-resolution µ-CT images and high-resolution scanning electron microscope (SEM) images. The proposed method involves four steps. First, a three-dimensional low-resolution tomographic image of a rock sample is obtained by µ-CT scanning. Next, one or more sections in the rock sample are selected for scanning by SEM to obtain high-resolution two-dimensional images. The high-resolution segmented SEM images and their corresponding low-resolution µ-CT slices are then applied to train a convolutional neural network (CNN) model. Finally, the trained CNN model is used to reconstruct the entire low-resolution three-dimensional µ-CT image. Because the SEM images are segmented and have a higher resolution than the µ-CT image, this algorithm integrates the super-resolution and segmentation processes. The input data are low-resolution µ-CT images, and the output data are high-resolution segmented porous structures. The experimental results show that the proposed method can achieve state-of-the-art performance.     

Automatic extraction of sparse trees from high-resolution ortho-images

Obtaining information about tree species distribution in agricultural lands is a topic of interest for various applications, such as tree inventory, forest management, agricultural land management, crop estimation, etc. This information can be derived from images obtained from modern remote sensing technology, which is the most economical way as compare to field surveys covering large geographic areas. Therefore, in this study, a new method is proposed for extraction and counting of sparse and regular distributed individual pistachio trees from agricultural areas on large scale from high-resolution digital ortho-photo maps, which were obtained using an airborne sensor (Ultracam-X). The input images were first smoothed by applying Gaussian filter to reduce the impact of noise. Normalized difference vegetation indices (NDVI) were then derived to obtain vegetation areas followed by Otsu’s global thresholding algorithm to obtain candidate tree areas. Further, connected component (CC) analysis was applied to segregate each object. Morphological processing was performed to fill holes within tree objects and get smooth contours, which were obtained by using the Moore-neighbor tracing method (MNTM) for each CC, while geometrical constraints were applied to undermine possible non-tree elements from output image. To further improve the segmentation results for sparse trees, a new method was applied, called quadratic local analysis (QLA). QLA helped to segment the trees, which were missed by the Otsu method due to low contrast and resulted in improved accuracy (3–6%). The obtained results were compared with well-known support vector machine (SVM) classifier. Proposed method produced slightly better results (1–5%) than SVM for extraction of pistachio trees and obtained accuracy for QLA and SVM were 96 and 91% for region 1, while 91 and 90% for region 2 respectively.     

遥感与化探数据融合技术在金矿预测中的应用

作为一种先进的多元数据融合技术,遥感与化探数据融合技术的目的在于能从原始的遥感图像中获得较高质量的图像和提取实际需要的蚀变信息.文章首先研究了几个典型的遥感与化探融合的算法(如PCA、ISODATA、MLC等),然后提出了一个新的使用了专家系统的基于3个层次(像素级、特征级和决策级)的遥感与化探数据融合方法.该方法通过在招远金矿区应用示范,结果表明这种新的方法较之传统的遥感与化探融合方法是一种行之有效的遥感找矿技术方法.     

An improved fast level set method initialized with a combination of k-means clustering and Otsu thresholding for unsupervised change detection from SAR images

Detection of changes in synthetic aperture radar (SAR) images is an important challenge due to the effects of speckle noise on these images. In recent years, appropriate methods for SAR-based-change detection have been developed based on the level set methods (LSM). These methods need to set parameters for defining a proper initial contour. Moreover, the gradient information is only employed in the total energy of these methods for segmentation of the difference image. In this study, a novel method has been proposed for unsupervised change detection of multitemporal SAR images based on the improved fast level set method (IFLSM) initialized with a combination of k-means and Otsu techniques. The proposed method utilizes the discrete wavelet transform (DWT) fusion strategy and edge enhancement to achieve a noise-resistant difference image from the mean-ratio and log-ratio images. Afterward, the generated binary change map (CM) by applying a combination of k-means and Otsu techniques on the difference image is used as the initial contour to achieve a final CM on difference image using the IFLSM. To check advantages of the proposed method, experiments are applied on two sets of multitemporal SAR images corresponding to artificial Chitgar Lake (under reconstruction) in Tehran (Iran) taken by TerraSAR-X satellite in 2011 and 2012, and corresponding to San Pablo and Briones reservoirs in California (USA) acquired by ERS-2 satellite in 2003 and 2004. Results of proposed method were compared with results of some well-known unsupervised change detection methods. Experimental results prove the sufficiency of the proposed method in unsupervised change detection in terms of accuracy, implementation time, and computational complexity.     

基于各类特征对齐迁移网络的多时相遥感图像分类

为了在目标域遥感图像不存在标记数据的情况下实现自动分类,论文提出一种基于特征对齐的迁移网络.网络以各类类心对齐和协方差对齐作为迁移策略,全面描述域间各类别之间的对应关系,实现知识迁移.另外,网络采用线性修正单元作为激活函数,能够产生稀疏特征,提高分类效果.该迁移网络能够同时获得对齐的特征和自适应分类器,不需要目标域的标记数据,实现无监督迁移学习.在多时相的Hyperion高光谱遥感图像和WorldView-2多光谱遥感图像上的实验结果证明了该迁移网络的有效性.      

砂样图像岩屑自动分割提取方法

通过将砂样图像进行单颗粒分割,识别砂样成分,可显著提高砂样岩性分析的准确性和效率。现有的砂样图像分割方法主要以传统分水岭算法和卷积神经网络为主,但由于对单颗粒岩屑轮廓细节提取不足,误分割率高。本文提出一种以图像融合算法为桥梁,将卷积神经网络和分水岭算法相结合的单颗粒图像分割提取方法。首先利用改进的Mask R-CNN网络快速分割砂样原图,获得其初分割图像;然后,将初分割图像与砂样原图进行融合,再使用改进的分水岭算法对融合结果进行分割;最后,利用砂样原图坐标点匹配方法,将分水岭分割得到的结果图像进行修正,完成单颗粒岩屑图像提取。实验结果表明,本文的单颗粒自动分割提取方法准确率高达96.77%,且模型更轻量和精准,为岩屑图像分割提供了一种可行且有效的方法,可满足有效测算油藏层构造变化、查找潜在沉积物源及储层动态变化的需求。     

基于图像描述的煤岩裂隙CT图像多尺度特征

基于煤岩孔隙系统多尺度结构特征对深入认识多尺度流体运移机制的重要性,提出了基于图像描述的煤岩CT图像孔隙结构的多尺度精细描述方法。采用了图像的多策略分割技术提取目标,利用Freeman链码对目标的边界进行表达,研究了由形态学、统计矩、链码、计盒维数构造目标之间的关系、目标占有区域与边界的图像描绘子、以及分形描绘子;综合运用上述方法对煤岩CT图像中的大尺度宏观裂纹目标、小尺度细观裂隙目标进行了识别。结果表明,宏观裂纹可由灰度阈值法实现目标提取;小尺度细观裂隙需采用较复杂的分割策略,如基于索贝尔梯度算子的分水岭变换;进一步应用链码表达、图像描绘子和分形描绘子,实现了煤岩孔隙结构在欧氏空间与分形空间的多尺度精确描述。     

Fast wavelet-based stochastic simulation using training images

Spatial uncertainty modelling is a complex and challenging job for orebody modelling in mining, reservoir characterization in petroleum, and contamination modelling in air and water. Stochastic simulation algorithms are popular methods for such modelling. In this paper, discrete wavelet transformation (DWT)-based multiple point simulation algorithm for continuous variable is proposed that handles multi-scale spatial characteristics in datasets and training images. The DWT of a training image provides multi-scale high-frequency wavelet images and one low-frequency scaling image at the coarsest scale. The simulation of the proposed approach is performed on the frequency (wavelet) domain where the scaling image and wavelet images across the scale are simulated jointly. The inverse DWT reconstructs simulated realizations of an attribute of interest in the space domain. An automatic scale-selection algorithm using dominant mode difference is applied for the selection of the optimal scale of wavelet decomposition. The proposed algorithm reduces the computational time required for simulating large domain as compared to spatial domain multi-point simulation algorithm. The algorithm is tested with an exhaustive dataset using conditional and unconditional simulation in two- and three-dimensional fluvial reservoir and mining blasted rock data. The realizations generated by the proposed algorithm perform well and reproduce the statistics of the training image. The study conducted comparing the spatial domain filtersim multiple-point simulation algorithm suggests that the proposed algorithm generates equally good realizations at lower computational cost.     

Air pollution forecasting from sky images with shallow and deep classifiers

Air pollution is one of the most important problems in the new era. Detecting the level of air pollution from an image taken by a camera can be informative for the people who are not aware of exact air pollution level be declared daily by some organizations like municipalities. In this paper, we propose a method to predict the level of the air pollution of a location by taking an image by a camera of a smart phone then processing it. We collected an image dataset from city of Tehran. Afterward, we proposed two methods for estimation of level of air pollution. In the first method, the images are preprocessed and then Gabor transform is used to extract features from the images. At the end, two shallow classification methods are employed to model and predict the level of air pollution. In the second proposed method, a Convolutional Neural Network(CNN) is designed to receive a sky image as an input and result a level of air pollution. Some experiments have been done to evaluate the proposed method. The results show that the proposed 9 method has an acceptable accuracy in detection of the air pollution level. Our deep classifier achieved accuracy about 59.38% which is 10 about 6% higher than traditional combination of feature extraction and classification methods.     

煤矿掘进工作面低照度视频增强技术研究

针对煤矿掘进工作面视频光照较低、亮度不均、纹理模糊、噪声较多等问题,提出一种煤矿掘进工作面低照度视频增强算法。首先,利用卷积的可分离性将视频图像进行一维水平卷积与垂直卷积,再利用完美反射法实现视频图像自动白平衡,并使用图像混合增强技术提高视频图像整体亮度。然后,基于大气散射模型与暗通道先验方法,通过递归分割将图像分割为高光区、中间调和暗调区,并求取对应区间通道像素最大值,将其3者均值作为大气光照估计值,引入调节因子对透射率进行调整优化,并使用拉普拉斯锐化操作,增加图像高频成分、抑制图像低频成分,提高图像对比度。最后,基于改进的大气散射模型对掘进工作面低照度视频进行去雾处理。实验结果表明,视频增强算法能够对煤矿掘进工作面低照度视频进行实时增强、去雾处理,避免了视频图像暗淡、失真、模糊和突变等问题。相较于Retinex算法、ALTM算法和暗通道先验算法,视频增强算法大幅度提高了视频图像的信息熵、标准差和平均梯度,且具有较好的实时处理速度,能够为掘进工作面视频的目标识别、目标跟踪、目标监测和图像分割等后续处理提供优质、可靠的支撑。      

数字钻孔图像兴趣区域分割与岩体结构面特征识别

针对现有数字钻孔图像分析技术的不足,提出新分析方案以实现数字钻孔摄像技术（BCT）所采集钻孔内壁图像的自动化、定量化结构面检测、识别与分析.首先,对数字钻孔图像进行预处理,设计特征信号DH,检测并获取兴趣区域;然后,使用低精度Hough变换快速检测结构面兴趣区域内的结构面分布特征,利用聚类算法分离出单一结构面后,针对单一结构面进行亚像素级Hough变换,以获取结构面的正弦参数;最后,根据结构面正弦参数,计算出结构面的倾向、倾角等信息.通过对如美水电站左坝肩数字钻孔图像实例进行分析,利用本算法完成图像中结构面的自动分割与特征识别,成功获取其几何信息,并与传统人工辅助方案结果进行对比,验证了该算法的可靠性.      

A tomographic imagery segmentation methodology for three-phase geomaterials based on simultaneous region growing

X-ray computed tomography is a powerful non-destructive technique used in many domains to obtain the three-dimensional representation of objects, starting from the reconstitution of two-dimensional images of radiographic scanning. This technique is now able to analyze objects within a few micron resolutions. Consequently, X-ray microcomputed tomography opens perspectives for the analysis of the fabric of multiphase geomaterials such as soils, concretes, rocks and ceramics. To be able to characterize the spatial distribution of the different phases in such complex and disordered materials, automated phase recognition has to be implemented through image segmentation. A crucial difficulty in segmenting images lies in the presence of noise in the obtained tomographic representation, making it difficult to assign a specific phase to each voxel of the image. In the present study, simultaneous region growing is used to reconstitute the three-dimensional segmented image of granular materials. First, based on a set of expected phases in the image, regions where specific phases are sure to be present are identified, leaving uncertain regions of the image unidentified. Subsequently, the identified regions are grown until growing phases meet each other with vanishing unidentified regions. The method requires a limited number of manual parameters that are easily determined. The developed method is illustrated based on three applications on granular materials, comparing the phase volume fractions obtained by segmentation with macroscopic data. It is demonstrated that the algorithm rapidly converges and fills the image after a few iterations.     

一种面向对象的最优分割尺度计算模型

作为信息提取和分类的前提,面向对象的影像分割尺度参数的设置直接影响到提取和分类的精度。本文以GF-2影像数据为例,在已有分割理论和方法的基础上提出一种基于最优分割尺度的计算模型（OS模型）。该模型以主成分分析所得的主成分以及新建的归一化植被指数（normalized vegetation index,NDVI）特征层作为分割参考层,综合考虑均质因子的影响,构建加权尺度评价指数,插值拟合最优分割尺度。构建误差系数（E_c）对模型进行评价,结果表明：OS模型误差系数（E_c=1.15%）小于传统模型（E_c=3.28%）,且分割对象更均匀、与实际地物更接近。