基于光学图像的舰船航迹检测

舰船航迹检测在海上交通管制和军事领域中有着广泛的应用前景．现有的航迹检测方法大都基于SAR图像，信息不直观而且算法复杂．提出了一种新颖的基于光学图像的航迹检测方法．首先利用纹理分析方法判断待检测区域中是否含有航迹，然后在通过判断的区域中使用条纹增强算法，凸现航迹特征，最后使用Radon变换检测航迹，并对结果进行了优化．与现有的检测方法相比，该方法针对性强，复杂度低．使用该方法对实际航拍图片进行了检测实验，取得了很好的效果．     

Determination of shallow water depth using optical satellite images

The Penghu archipelago comprises 64 basaltic volcanic isles lying on the Taiwan Strait between mainland China and Taiwan. The water around and within these isles is shallow and poses considerable difficulty in echo sounding detection for bathymetry. Most existing bathymetry data around such areas are in water depths of greater than 5 m. Therefore, when the water depth is less than 5 m the data tend to be over-extrapolated. In this study, a remote sensing method provides a more effective approach to recording shallow water depths compared to traditional soundings using multitemporal images collected by optical/near-infrared sensors from SPOT satellites. This method employs optical energy reflections to obtain the water depth. In this study, we made several improvements wherein a relative atmosphere correction technique was used to calibrate two images within a similar atmospheric condition. We then compared the satellite images acquired from different dates to obtain the local water attenuation coefficient of sunlight. Finally, we developed a means to estimate the water attenuation coefficient and bottom reflectance which will satisfy the two parameters across the study area. Our results show a high-resolution map of shallow bathymetry for the Penghu archipelago and revealed a maximum depth of about 20 m. This study provides an efficient approach for shallow bathymetry retrieval. Many detailed features revealed by this approach may contribute to further geological research and developments in harbour and coastal engineering.     

Cloud detection in satellite images using multi-objective social spider optimization

Cloud detection algorithms have emerged to automate image data analysis because of its prime influential factor in remote sensing image quality. Cloud detection algorithm still needs domain-expert intervention and large number of training examples to ensure good performance whose acquirement becomes difficult due to unavailability of labeled data as well as the time and process heads involved. The paper puts forward multi-objective social spider optimization (MOSSO) based efficient clustering technique to detect clouds in the visible range. This paper explains the proposed MOSSO algorithm along-with the analysis carried on 14 benchmark two-objective test problems against MOEA/D, MODE, MOPSO and SPEA2 multi-objective algorithms. Further, the strengths and weaknesses of the proposed algorithm are analyzed and have been used for the implementation of an efficient clustering technique named as MOSSO-C. Optimal centroid matrix for clustering is attained in MOSSO-C through environmental selection whose performance evaluation has been done on six synthetic databases and are compared with above mentioned conventional multi-objective algorithms. The obtained results encourage the use of MOSSO-C technique to get labeled data for training process of neural network classifier. This approach efficiently classifies the cloudy pixels against various Earth’s surfaces (water, vegetation and land). The paper also discusses the performance evaluation of proposed technique on four Landsat 8 data which shows on an average 96.37% performance accuracy in detecting cloudy pixels.     

深度卷积网络卫星图像水泥厂目标检测

目的 水泥厂作为重要的污染源企业需要对其进行统计和监管,近几年随着卫星遥感技术的发展和遥感影像分辨率的提高,使得基于卫星影像进行水泥厂目标检测成为可能。但是由于遥感图像中建筑目标的环境复杂多变,同时各个水泥厂在生产规模、设备构成、厂区结构、坐落方位上存在较大差异,图像表观上的形态各异和复杂环境干扰使得传统图像识别方法难以设计和提取有效特征。鉴于深度学习在视觉目标检测领域的成功应用,本文将研究应用深度卷积神经网络方法,实现在卫星图像上识别与定位水泥厂目标,为环保部门提供一种高效便捷的水泥厂目标检测和统计方法。方法 基于面向目标检测与定位的Faster R-CNN深度学习框架,以准确检测与定位水泥厂区域为目的,以京津冀地区的水泥厂位置作为训练和测试数据集,选用3种结构不同的提取特征卷积神经网络模型进行了对比实验。并针对小样本训练容易出现的过拟合和误检问题,采用图像去雾预处理、数据扩充、引入负样本等技术进一步提升模型能力。结果 测试集实验结果表明ResNet特征提取网络效果最好,准确率达到74%。为了进一步提高检出率并降低误检率,引入3种模型能力提升方法,在扩充检测数据集中的检出率达到94%,误检率降低到14%;在全球水泥厂数据集中的图像检出率达到96%,万幅随机图像的误检数量为30幅（0.3%）。对上海地区的卫星图像进行扫描检测,结果检测出11个已登记的水泥厂（共登记16个）,另外还检测出17个未登记的水泥厂。结论 对于卫星地图上水泥厂这种具有特殊建筑构造但也存在厂区几何形状各异、所处地理环境复杂、随季节性变化等特点,本文提出的基于深度卷积网络的卫星图像水泥厂检测方法,能够自动学习提取有效的图像特征并对目标进行准确检测。针对小样本训练问题,引入3种方法显著提高了模型的检测精度。在模型泛化能力测试中,经过优化后的模型在水泥厂建筑目标检测任务中表现良好,具有重要的应用价值。     

Structural damage detection using imperialist competitive algorithm and damage function

In practical damage detection problems, experimental modal data is only available for a limited number of modes and in each mode, only a limited number of nodal points are recorded. In using modal data, the majority of the available damage detection solution techniques either require data for all the modes, or all the nodal data for a number of modes; neither of which may be practically available through experiments. In the present study, damage identification is carried out using only a limited number of nodal data of a limited number of modes. The proposed method uses the imperialist competitive optimization algorithm and damage functions. To decrease the number of design variables, several bilinear damage functions are defined to model the damage distribution. Damage functions with both variable widths and variable weights are proposed for increased accurately. Four different types of objective functions which use modal responses of damaged structure are investigated with the aim of finding the most suitable function. The efficiency of the proposed method is investigated using three benchmark numerical examples using both clean and noisy modal data. It is shown that by only using a limited number of modal data, the proposed method is capable of accurately detecting damage locations and reasonably accurately evaluate their extents. The proposed algorithm is most effective with noisy modal data, compared to other available solutions.     

遥感图像变化检测算法研究

提出了一种基于非下采样Contourlet变换和脉冲耦合神经网络的无监督的不同时相的卫星影像的变化检测新算法。该算法将非下采样Contourlet变换和脉冲耦合神经网络这两种方法结合,将它应用于不同时相的卫星影像的变化检测。实验结果表明,与传统方法相比,对于高斯和斑点噪声,该算法具有更高的抗噪性能和检测精确度。     

Semi-automatic recognition and mapping of rainfall induced shallow landslides using optical satellite images

We present a method for the semi-automatic recognition and mapping of recent rainfall induced shallow landslides. The method exploits VHR panchromatic and HR multispectral satellite images, and was tested in a 9.4 km² area in Sicily, Italy, where on 1 October 2009 a high intensity rainfall event caused shallow landslides, soil erosion, and inundation. Pre-event and post-event images of the study area taken by the QuickBird satellite, and information on the location and type of landslides obtained in the field and through the interpretation of post-event aerial photographs, were used to construct and validate a set of terrain classification models. The models classify each image element (pixel) based on the probability that the pixel contains (or does not contain) a new rainfall induced landslide. To construct and validate the models, a procedure in five steps was adopted. First, the pre-event and the post-event images were pan-sharpened, ortho-rectified, co-registered, and corrected for atmospheric disturbance. Next, variables describing changes between the pre-event and the post-event images attributed to landslide occurrence were selected. Next, three classification models were calibrated in a training area using different multivariate statistical techniques. The calibrated models were then applied in a validation area using the same set of independent variables, and the same statistical techniques. Lastly, combined terrain classification models were prepared for the training and the validation areas. The performances of the models were evaluated using four-fold plots and receiver operating characteristic curves. The method proved capable of detecting and mapping the new rainfall induced landslides in the study area. We expect the method to be capable of detecting analogous shallow landslides caused by similar (rainfall) or different (e.g. earthquake) triggers, provided that the event slope failures leave discernable features captured by the post-event satellite images, and that the terrain information and satellite images are of adequate quality. The proposed method can facilitate the rapid production of accurate landslide event-inventory maps, and we expect that it will improve our ability to map landslides consistently over large areas. Application of the method will advance our ability to evaluate landslide hazards, and will foster our understanding of the evolution of landscapes shaped by mass-wasting processes.     

An adaptive fuzzy-genetic algorithm approach for building detection using high-resolution satellite images

We propose a new approach for building detection using high-resolution satellite imagery based on an adaptive fuzzy-genetic algorithm. This novel approach improves object detection accuracy by reducing the premature convergence problem encountered when using genetic algorithms. We integrate the fundamental image processing operators with genetic algorithm concepts such as population, chromosome, gene, crossover and mutation. To initiate the approach, training samples are selected that represent the specified two feature classes, in this case “building” and “non-building”. The image processing operations are carried out on a chromosome-by-chromosome basis to reveal the attribute planes. These planes are then reduced to one hyperplane that is optimal for discriminating between the specified feature classes. For each chromosome, the fitness values are calculated through the analysis of detection and mis-detection rates. This analysis is followed by genetic algorithm operations such as selection, crossover and mutation. At the end of each generation cycle, the adaptive-fuzzy module determines the new (adjusted) probabilities of crossover and mutation. This evolutionary process repeats until a specified number of generations has been reached. To enhance the detected building patches, morphological image processing operations are applied. The approach was tested on ten different test scenes of the Batikent district of the city of Ankara, Turkey using 1 m resolution pan-sharpened IKONOS imagery. The kappa statistics computed for the proposed adaptive fuzzy-genetic algorithm approach were between 0.55 and 0.88. The extraction performance of the algorithm was better for urban and suburban buildings than for buildings in rural test scenes.     

Urban road extraction from high-resolution optical satellite images

In recent years, many approaches have been exploited for automatic urban road extraction. Most of these approaches are based on edge and line detecting algorithms. In this paper, a new integrated system for automatic extraction of main roads in high-resolution optical satellite images is present. Firstly, a multi-scale greylevel morphological cleaning algorithm is proposed to reduce the grey deviation of the road regions. Secondly, based on the greylevel difference between road surfaces and environmental objects, a colour high-resolution satellite image is segmented into a simplified imagemap by using the mean shift algorithm, which consists of three stages. The first stage deals with image filtering, the second stage deals with colour segmentation, and the third stage is proposed to fuse small regions in the segmented image. The mean shift filter algorithm not only smoothes the image, but also preserves abrupt changes (i.e. edges) in the local structure. The mean shift segmentation algorithm is a straightforward extension of the smoothing algorithm, which preserves discontinuity. From the histogram of the simplified imagemap, we can find the potential road surfaces, and use greylevel threshold to convert the segmented image into a binary one. The binary image is processed by using binary mathematical morphological closing and opening to remove small objects and to open the connected street blocks. We use a contour tracing algorithm to remove holes in street-block regions and to detect the street blocks' contours. In this research we found that many street blocks' contours were preserved perfectly, except for some of them which were depressed. Finally, we utilize the convex hull algorithm to smooth the street blocks' zigzag edges and to close the gaps in some street blocks, and then, we get the road edges. The integrated system for road network extraction is tested on the red band of an IKONOS multispectral image; all algorithms in this study are developed in C++ under Windows XP operating system. Results of the road network extraction are presented to illustrate the validation of the extracting strategy and the corresponding algorithms in this research, and future prospects are exposed.     

Retrieving multispectral satellite images using physics-basedinvariant representations

We present a set of algorithms and a search strategy for the robust content-based retrieval of multispectral satellite images. Since the property of interest in these images is usually the physical characteristics of ground cover, we use representations and methods that are invariant to illumination and atmospheric conditions. The representations and algorithms are derived for this application from a physical model for the formation of multispectral satellite images. The use of several representations and algorithms is necessary to interpret the diversity of physical and geometric structure in these images. Algorithms are used that exploit multispectral distributions, multispectral spatial structure, and labeled classes. The performance of the system is demonstrated on a large set of multispectral satellite images taken over different areas of the United States under different illumination and atmospheric conditions     

局部显著特征下的光学遥感图像舷靠舰船检测

目的 在光学遥感图像中,针对舷靠舰船灰度和纹理特征与港口相近,传统方法检测效果不理想的问题,提出一种基于局部显著特征的舷靠舰船检测方法。方法 首先,对原始图像预处理得到海陆分割后的二值图像;然后,提取二值图像中的直线段作为局部显著特征检测舰船目标;再将直线段提取结果与舰首检测相结合,建立舷靠舰船检测模型;最后,通过计算舰船几何尺寸及环境信息分析确定舰船目标。结果 在两幅不同场景的光学遥感图像中验证本文方法并与其他算法进行对比,本文方法识别率可达100%,且不存在误检和漏检情况,相比于其他算法具有一定优势。在舰船背景复杂或停泊朝向不定时,文中方法可有效判别舰船停靠方向并对舰船目标进行正确标记。结论 在复杂背景环境及其他干扰下,应用本文方法检测舷靠舰船目标准确率高,鲁棒性强,具有较高适应性。     

RTDNet：面向高分辨率卫星影像的赤潮探测网络

目的 赤潮是一种常见的海洋生态灾害,严重威胁海洋生态系统安全。及时准确获取赤潮的发生和分布信息可以为赤潮的预警和防治提供有力支撑。然而,受混合像元和水环境要素影响,赤潮分布精细探测仍是挑战。针对赤潮边缘探测的难点,结合赤潮边缘高频特征学习与位置语义,提出了一种计算量小、精度高的网络模型RTD-Net （red tide detection network）。方法 针对赤潮边缘探测不准确的问题,设计了基于RIR （residual-in-residual）结构的网络,以提取赤潮边缘水体的高频特征;利用多感受野结构和坐标注意力机制捕获赤潮水体的位置语义信息,增强赤潮边缘水体的细节信息并抑制无用的特征。结果 在GF1-WFV （Gaofen1 wide field of view）赤潮数据集上的实验结果表明,所提出的RTDNet模型赤潮探测效果不仅优于支持向量机（support vector machine,SVM）、U-Net、Deep-Labv3+及HRNet （high-resolution network）等通用机器学习和深度学习模型,而且也优于赤潮指数法GF1_RI （Gaofen1 red tide index ）以及赤潮探测专用深度学习模型RDU-Net （red tide detection U-Net）,赤潮误提取、漏提取现象明显减少,F1分数在两幅测试图像上分别达到了0.905和0.898,相较于性能第2的模型DeepLabv3+提升了2%以上。而且,所提出的模型参数量小,仅有2.65 MB,约为DeepLabv3+的13%。结论 面向赤潮探测提出一种基于RIR结构的赤潮深度学习探测模型,通过融合多感受野结构和注意力机制提升了赤潮边缘探测的精度和稳定性,同时有效降低了计算量。本文方法展现了较好的应用效果,可适用于不同高分辨率卫星影像的赤潮探测。     

Classified road detection from satellite images based on perceptual organization

Extracting roads from satellite images is an important task in both research and practice. This work presents an improved model for road detection based on the principles of perceptual organization and classification fusion in human vision system (HVS). The model consists of four levels: pixels, primitives, structures and objects, and two additional sub‐processes: automatic classification of road scenes and global integration of multiform roads. Based on the model, a novel algorithm for detecting roads from satellite images is also proposed, in which two types of road primitives, namely blob‐like primitive and line‐like primitive are defined, measured, extracted and linked using different methods for dissimilar road scenes. A hierarchical search strategy driven by saliency measurement is adopted in both linking processes. The blob primitives are linked using heuristic grouping and the line primitives are connected through genetic algorithm (GA) evolution. Finally, all of the linked road segments are normalized with centre‐main lines and integrated into global smooth road curves through tensor voting. Experimental results show that the algorithm is capable of detecting multiform roads from real satellite images with high adaptability and reliability.     

Multi-channel higher-order local autocorrelation for object detection on satellite images

ABSTRACT

An automatic object-detection method is necessary to facilitate the efficient analysis of satellite images consisting of multispectral images. Considering that the relationship between spectrums is important for discriminating objects in multispectral images. This paper proposes a feature extraction method that can capture both spatial and spectral relationships of the multispectral images. Moreover, image preprocessing and dimensionality reduction procedures are introduced for stable feature extraction. In this study, we conducted experiments for detecting two types of objects by using Landsat 8 images. The proposed method improved the detection performance relative to other image features.     

Superpixel-based active contour model for unsupervised change detection from satellite images

This study proposes a superpixel-based active contour model (SACM) for unsupervised change detection from satellite images. The accuracy of change detection produced by the traditional active contour model suffers from the trade-off parameter. The SACM is designed to address this limitation through the incorporation of the spatial and statistical information of superpixels. The proposed method mainly consists of three steps. First, the difference image is created with change vector analysis method from two temporal satellite images. Second, statistical region merging method is applied on the difference image to produce a superpixel map. Finally, SACM is designed based on the superpixel map to detect changes from the difference image. The SACM incorporates spatial and statistical information and retains the accurate shapes and outlines of superpixels. Experiments were conducted on two data sets, namely Landsat-7 Enhanced Thematic Mapper Plus and SPOT 5, to validate the proposed method. Experimental results show that SACM reduces the effects of the trade-off parameter. The proposed method also increases the robustness of the traditional active contour model for input parameters and improves its effectiveness. In summary, SACM often outperforms some existing methods and provides an effective unsupervised change detection method.     

Geometrical processing of multi-sensoral multi-temporal satellite images for change detection studies

A technique for geometrical processing of multi-sensoral and multispectral satellite images for the purposes of change detection studies is presented here. The technique involves geometrical rectification (geocoding), and image registration with two-dimensional image correlation. The application of the technique has been demonstrated in an area within the Niger Sahel in West Africa. The study was conducted with MSS and TM image data. The procedure results in image registration accuracy of 0.28pixel, which in this instance is good for change detection purposes.     

基于直线检测算法的卫星图片中建筑物轮廓提取

提出一种方法,可以从卫星图像中自动检测建筑物.介绍了直线提取和直线合并的算法,分别讨论算法的实现结果和对结果的评价.建筑物检测的结果为矢量的二维候选数据,缩短了原始图像数据和最后对图像理解之间的差距.     

A parallel algorithm for detection of linear structures in satellite images

To enhance linear structures in a gray level image, local operations with an additive score are normally used. Here a multiplicative score is used instead which gives better results than the additive one. The problem of segmenting the image of the multiplicative score is then dealt with where the threshold value can be automatically selected. The experimental results on some satellite images are reported.     

Reconnaissance mapping and management of salt-affected soils using satellite images

The Indo-Gangetic alluvial plain is subjected to large scale soil alkalization. In order to map and characterize salt-affected soils, with the aim of applying management techniques, Etah district in Uttar Pradesh, located between 26o^o45' to 28^o02' N and 78^o15' to 79^o20' E was selected. Multidate, high resolution, IRS-LISS II, geocoded FCC images on 1550 000 scale were used. Integrating visual image interpretation, physiographic analysis, ground data and laboratory analysis of soil samples, a legend for mapping salt-affected soils (SAS) was formulated. Based on variations in physicochemical properties: nature, intensity and depth wise distribution of salts, five categories of SAS requiring specific reclamation measures were identified. Soil categories S2, S3 and S4 have a gypsum requirement (GR) of 20, 12 and 4t/ha^-1 respectively. Reclamation of medium to heavy textured highly alkali soils requires the addition of amendments and a rice-wheat rotation for the initial 3-4 years. Under resource constraints, pit planting of Prosopis juliflora can bring about slow but effective reclamation. The soil category S5 is slightly alkali in the substratum, needing only biological reclamation by growing salt tolerant varieties of rice and wheat crops. Soils of category S1 are saline and need management by hydrological treatments. Incorporation of village boundaries on a map showing SAS would facilitate decision taking in planning reclamation projects and accelarate management operations directly at village level.     

Estimating soybean ground cover from satellite images using neural-networks models

The ground cover is a necessary parameter for agronomic and environmental applications. In Argentina, soybean (Glycine max (L.) Merill) is the most important crop; therefore it is necessary to determine its amount and configuration. In this work, neural-network (NN) models were developed to calculate soybean percentage ground cover (fractional vegetation cover, fCover) and to compare the accuracy of the estimate from Moderate-Resolution Imaging Spectroradiometer (MODIS) and Landsat satellites data. The NN design included spectral values of the red and near-infrared (NIR) bands as input variables and one neuron output, which expressed the estimated coverage. Data of fCover were acquired throughout the growing season in the central plains of Córdoba (Argentina); they were used for training and validating the networks. The results show that the NNs are an appropriate methodology for estimating the temporal evolution of soybean coverage fraction from MODIS and Landsat images, with coefficients of determination (R ²) equal to 0.90 and 0.91, respectively.