In forestry, treecanopy cover (CC) is an important biophysical indicator for characterizing terrestrial ecosystemsand modeling global biogeochemical cycles, e.g., woody biomass estimation, carbon balance analysis (sink/emission). However, currently available CC product cannot fully meet what we need while conducting woody biomass estimation in tropical savannas.It is thus necessary to develop an approach to estimate more reliable CC. Based on the acquisition of multisensor and multiresolution dataset, this study introduces an innovative multiscalemethod for this purpose taking the multiple savannas country Sudan as an example. The procedure includes: (1)Measurement of CC using Google Earth Pro in which very highresolutionimages such as QuickBirdand GeoEye images are available, and then the measured CC was coupled with atmospherically corrected and reflectance-based 16 frames of Landsat ETM+ vegetation indices (EVI, SARVI and NDVI)dated Nov 1999-2002 to establish the CC-VIs models; it was noted that among these indices NDVI indicates the best correlation with CC (CC = 153.09NDVI– 10.12, R 2 = 0.91);(2) The NDVI of Landsat ETM+ was calibrated against MODIS NDVI of the same time period (Nov 2002)to make sure that model developed from Landsat ETM+ data can be applied to MODIS data for upscalingto regional scale study; (3)Time-series MODIS NDVI data of the period Jan 2002–Dec 2009 (MODIS13Q1, 250m, 186 acquisitions) were acquired and used to decompose the woody component(NDVI) from seasonal changeand herbaceous component by time-series analysis;(4) The equation obtained in step 1 was applied to the decomposed MODIS woody NDVI images to derive country scale CC data. The produced CC was checked against the 287 ground measured CC obtained in step 1 and a good agreement (R 2 = 0.53-0.71) was found.It is hence concluded that the proposed multiscale approach is effective, operational and can be applied for reliable estimation of regional and even continental scales CC data.
Overview of the system is shown in Fig. 2. Given a low resolution input scene image, response map is computed from CNN, and bounding boxes are formed to indicate the detected text lines. One can see the noise caused by low resolution from the zoomed image in Fig. 2b. These for example, the characters are not smooth as in highresolutionimages, touching between characters may be clas- sified as non-text, missing characters due to imperfect geometrical properties, and non-text may be classified as text due to noise.
The second reason lies in the resolution of VHR images itself. In the proposed algorithm roads – more precisely: central lines on the roads – are used as a reference source for matching. Roads' widths are usually 5 to 15 m; this amounts to around one to three pixels in RapidEye resolution, but much more pixels in WorldView-2 resolution. Together with the previously described issues caused by imperfect detection – e.g. false positives such as buildings or urban elements along the roads – which is much more pronounced for the VHR images than for the highresolutionimages, it is clear that the central line of the road itself is not an appropriate reference source for fine positioning. Last but not least, VHR images have smaller swath, therefore the number of crossroads per image is usually smaller than on a RapidEye image. This substantially smaller redundancy causes problems if some crossroads are not detected well.
(Zhukov et al., 1999). A framework to sharpen low resolution im- ages with co-registered highresolutionimages is presented. The use of highresolutionimages is suggested to analyze the compo- sition of mixed pixels of the low resolution sensor. The unmix- ing can be constrained or unconstrained depending on whether the initial observations can be retrieved. The method proposed by (Zhukov et al., 1999) is a multi-sensor multi-resolution tech- nique. Vocabulary has been introduced : higher resolution im- ages are considered as the classifying instrument and lower ones as the measuring instrument. The analysis of mixed pixels of the measuring instrument is done by classifying the higher resolu- tion images. A synthetic response is then assigned to each class. (Park and Kang, 2004) mention unmixing-based methods defined in (Zhukov et al., 1999) as a possible means for the sharpening. (Zurita-Milla et al., 2007) used this spatial unmixing on optical images from different sensors : MERIS and Landsat TM. The images of greater resolution (Landsat TM) are classified, which makes the composition analysis of lower resolution pixels possi- ble. An estimation of a synthetic MERIS reflectance is carried out on each class of the Landsat TM Classification.
Based on the use of the layers information of high-resolutionimages, we propose a method of multi-scale segmentation of highresolution remote sensing images by integrating multiple features. The method extracts edge information by canny operator, and constructs from the function to obtain a weighted band edge weight value, according to certain segmentation criteria, then obtains an image of the original objects using Kruskal minimum spanning tree algorithm. Last simplified Mumford -Shah model is used to merge regions by spectral information and texture information to get the segmentation results. The method is not only adopting the edge, texture and other information of highresolutionimages to improve the accuracy of multi-scale segmentation, but also having on
the same topic, but uses the IKONOS images. The hierarchical image segmentation approach of multitemporal Landsat TM and ETM images was adopted by (Gamanya et al., 2009). The aim of this study is to perform CD by using fuzzy logic classification implemented in Definiens Professional software. The analysed area is approximately 900 km 2 . The classification method used in this research was developed and applied to their earlier work by the same authors (Gamanya et al., 2007). Automatic change detection of buildings in urban environment from VHSR satellite images using existing geodatabase was presented in (Bouziani et al., 2010). The proposed solution is based on the segmentation of Ikonos and QuickBird images. Generated segments are subsequently analyzed using a knowledge base which was created by using the GIS database and prior knowledge. The test area of unknown size is composed from two sub-areas covering parts of different cities (Canada, Sherbrooke and Morocco, Rabat). The determination of segmentation parameters, classification thresholds and object relations were obtained by use of geodatabases. This study builds on previous preliminary work of the same authors (Bouziani et al., 2007). The other group of authors (Huang et al., 2014) presented the urban building change detection from multitemporal QuickBird images based on the morphological building index (MBI). MBI describes a relationship between the spectral-spatial characteristics of buildings and the morphological operators. MBI is therefore able to automatically detect buildings from high-resolutionimages. This work builds on the previous works (Huang and Zhang, 2011; Huang and Zhang, 2012) which were dealing with the usage of MBI as well.
During the image capturing process of very highresolution imagery, numerous influential factors hinder the quality of these images, such as the shadows caused from the different angle of the sun, terrain features, and surface object occlusion (Dare, 2005; Tsai, 2006). Furthermore, the shadows in remote sensing images are regarded as image nuisances in numerous applications, specifically, change detection and image classification (Dare, 2005; Zhou et al., 2009), frequently affecting the accuracy of analytical results (Wilson, 1997). Especially in the mountainous environment, shadows frequently occur in terrain areas with steep slopes (Giles, 2001; Dare, 2005). As a result, the accuracy of land cover/ use mapping procedure over steep mountainous terrain is often low (Dorren et al., 2003; Shahtahmassebi et al., 2013). In Taiwan, the landscape is often characterized as alpine terrain, thus using very highresolutionimages for land cover/ use mapping will be severely affected by the shadow problem.
Multi-temporal remote sensing images classification is very useful for monitoring the land cover changes. Traditional approaches in this field mainly face to limited labelled samples and spectral drift of image information. With spatial resolution improvement, “pepper and salt” appears and classification results will be effected when the pixelwise classification algorithms are applied to high- resolution satellite images, in which the spatial relationship among the pixels is ignored. For classifying the multi-temporal highresolutionimages with limited labelled samples, spectral drift and “pepper and salt” problem, an object-based manifold alignment method is proposed. Firstly, multi-temporal multispectral images are cut to superpixels by simple linear iterative clustering (SLIC) respectively. Secondly, some features obtained from superpixels are formed as vector. Thirdly, a majority voting manifold alignment method aiming at solving highresolution problem is proposed and mapping the vector data to alignment space. At last, all the data in the alignment space are classified by using KNN method. Multi-temporal images from different areas or the same area are both considered in this paper. In the experiments, 2 groups of multi-temporal HR images collected by China GF1 and GF2 satellites are used for performance evaluation. Experimental results indicate that the proposed method not only has significantly outperforms than traditional domain adaptation methods in classification accuracy, but also effectively overcome the problem of “pepper and salt”.
Trees outside forests (TOF) are widely recognized to be pivotal resources, for sustainable natural resource management, due to their role in offering variety of goods, such as timber, fruits, and fodder(Arnold, 1996, Auclair et al., 2000, Boffa 2000b, Kumar and Sastry, 1999) as well as services like water, carbon(Castaneda et al, 2011) and biodiversity. Forest Conservation efforts involving reduction of deforestation and degradation may have to increasingly rely on alternatives provided by TOF (Saxena,1997, Namwata, et al, 2012) in catering to economic demands in forest edges. Spatial information systems involving imaging, analysis and monitoring to achieve objectives under protocols like REDD+, require incorporating content from areas under forest as well as trees outside forests, to aid holistic decisions. In this perspective, automation in retrieving information on area under trees, growing outside forests, using highresolution imaging is essential so that measuring and verification of extant carbon pools, are strengthened. Due to characteristic nature of highresolutionimages, where in conventional pixel based approaches cease to provide meaningful classification results, object based image analysis (OBIA) approach alone enables content extraction with varying degrees of automation. OBIA as implemented in commercial suite of e-Cognition (Ver 8.9) consists of different types of segmentations at user defined scale, to derive objects that possibly match expected in situ elements. Scale refers coarsely to the level of aggregation of pixels. Segmentation is followed by potential to apply wide range of spectral, textural and object geometry based parameters for classification. Software offers innovative blend of raster and vector features that can be juxtaposed flexibly, across scales horizontally or vertically. Study was carried out with an objective of retrieving image content consisting of trees outside forests using OBIA uponhigh resolution panchromatic and multispectral sensor onboard IRS satellites. Study features as one among the earliest efforts in this category of approach. Comparative assessment of different segmentation algorithms in highresolution multispectral datasets in terms of over- segmentation and under-segmentation was carried out involving
optical data from satellites is easier to access, such as WorldView- 2, compared to LiDAR data acquired from airborne sensors. Ac- quisition of optical data from satellites requires less planning, less man-power and is considerable cheaper than the acquisition of LiDAR data from airborne sensors. Therefore, extending the applicability of GeoRaySAR to use DSMs from high-resolution optical data is a crucial step toward realistic applications. Studies on the generation of DSMs with the usage of optical stereo images acquired with space borne sensors are presented in e.g. (Zhang and Gr¨un, 2006) and (Hoja et al., 2005). These studies indicate that DSMs generated from space borne sensor can have a height accuracy between 1 to 3 m, depending on the structure of the study area. Based on image matching with em- phasis on tri-stereo data (Carl et al., 2013), DSMs have been gen- erated from highresolution WorldView-2 stereo images for ur- ban scenes, which resulted in few height outliers. (Carl et al., 2013) confirms that the geometric quality of DSMs generated from WorldView-2 data can be used for automated SAR image simulation and, based on that, the interpretation of urban scenes. This paper presents an extended approach for interpretation of ur- ban areas with the usage of GeoRaySAR, using an automatic pre- processing chain. The main objective of the paper is to preprocess DSMs generated from WorldView-2, a highresolution optical satellite, for identification of buildings in SAR images (acquired with the satellite TerraSAR-X). These DSMs, in comparison to the previously used manually filtered DSMs, need to be prepro- cessed in terms of noise reduction and removal of vegetation. For evaluation of the preprocessing and determining the limitations of GeoRaySAR, study areas consisting of diverse types of buildings and densities are chosen.
The objective of this research is to improve the efficiency of the work in creating the high-definition road maps. In this paper, a novel approach of utilizing both of images and point cloud acquired from MMS is proposed for road sign detection and recognition. The approach is evaluated using experimental data which is acquired by MMS traveling more than 180km on an express way and a high way in Japan. The results show that the proposed approach has performed well even under the challenging conditions such as halation, shadows, partial occlusions and large deformations etc. The average recall of sign detection is 98.4%, and it is slightly better than that of manual extraction 98%. Hence, it is possible to improve the efficiency through replacing the manual task with automatic sign recognition.
This OGC Engineering Report describes the high-resolution flood information scenario carried out under the Urban Climate Resilience Thread of the Testbed 11 Initiative. The scenario was developed for two areas of interest: the San Francisco Bay Area and in Mozambique. The scenarios for these two locations demonstrate the interoperation and capabilities of open geospatial standards in supporting data and processing services. The prototype HighResolution Flood Information System addresses access and control of simulation models and high-resolution data in an open, worldwide, collaborative Web environment. The scenarios were designed to help testbed participants examine the feasibility and capability of using existing OGC geospatial Web Service standards in supporting the on-demand, dynamic serving of flood information from models with forecasting capacity. Change requests to OGC standards have also been identified through the Testbed activity.
Unmanned aerial vehicles (UAVs) became popular platforms for the collection of remotely sensed geodata in the last years (Hardin & Jensen 2011). Various applications in numerous fields of research like archaeology (Hendrickx et al., 2011), forestry or geomorphology evolved (Martinsanz, 2012). This contribution deals with the generation of multi-temporal crop surface models (CSMs) with very highresolution by means of low-cost equipment. The concept of the generation of multi-temporal CSMs using Terrestrial Laserscanning (TLS) has already been introduced by Hoffmeister et al. (2010). For this study, data acquisition was performed with a low-cost and low-weight Mini-UAV (< 5kg). UAVs in general and especially smaller ones, like the system presented here, close a gap in small scale remote sensing (Berni et al., 2009; Watts et al., 2012). In precision agriculture frequent remote sensing on such scales during the vegetation period provides important spatial information on the crop status. Crop growth variability can be detected by comparison of the CSMs in different phenological stages. Here, the focus is on the detection of this variability and its dependency on cultivar and plant treatment. The method has been tested for data acquired on a barley experiment field in Germany. In this contribution, it is applied to a different crop in a different environment. The study area is an experiment field for rice in Northeast China (Sanjiang Plain). Three replications of the cultivars Kongyu131 and Longjing21 were planted in plots that were treated with different amounts of N-fertilizer. In July 2012 three UAV-campaigns were carried out. Establishment of ground control points (GCPs) allowed for ground truth. Additionally, further destructive and non-destructive field data were collected. The UAV-system is
Deep learning algorithms require very large datasets for the train- ing phase. To overcome the lack of images (only 605 thumbnails in the training dataset) and improve performance, a number of data augmentation techniques can be used to enlarge the size of the dataset. In this work, we employ the Keras library (Chollet, 2015) to increase the training database size by combining several data augmentation methods such as rotation, translation, horizon- tal flip, vertical and horizontal shift. From the combination of all transformations, we can obtain fifty images for the dataset from a single ground-truth image. Concerning the ”other objects” cate- gory, we first choose, ten times more randomly extracted thumb- nails from the dataset image and make sure that the thumbnails do not contain manholes. Context of each manhole cover is also added in this second category. As we do several boosting itera- tions, all false positives will be added to the ”other objects” cat- egory. Finally, number of ”manhole covers” thumbnails in the dataset is 18 405 and the ”other objects” dataset size is 458 915. 3.6 Object Detection evaluation
Against high spectral resolution of hyperspectral images, these images have low spectral resolution. This is because of technological restrictions. To overcome this problem one approach is using PAN and hyperspectral sensor together. For example PRISMA observation system (PRISMA, 2013) consists of both hyperspectral and panchromatic sensors. In hyperspectral images there are pixels consisting of more than one distinct material called mixed pixels. Spectral unmixing methods extract spectral signatures of these distinct materials (endmembers) and abundance fractions of them. To solve spectral unmixing problem two models can be used: linear and nonlinear. Linea mixing model (LMM) is widely used for spectral unmixing of hyperspectral images since it is simpler and in most cases there is no interaction between materials in the scene (Keshava and Mustard, 2002). Some linear unmixing methods include: VCA (Nascimento and Dias, 2005), NMF (Pauca et al., 2006), etc.
Owing to recent progress in photogrammetry and surveying technologies (e.g., LiDAR), it is now possible to produce centimeter-level ultra-highresolution digital elevation models (DEMs) that enable more accurate observations of geomorphologic changes. Many researchers have investigated methods for detecting large-scale landslides and debris flow areas with these technologies for the purpose of monitoring geomorphic changes, calculating the extent of changes and assessing debris flow behavior (McKean and Roering, 2004; Du and Teng, 2007; Tsutsui et al., 2007; Schelidl et al., 2008; Bull et al., 2010; Kim et al., 2014). However, because these studies examined LiDAR DEMs with relatively long intervals between acquisitions (≥5 years), some natural topographic changes were The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLI-B8, 2016
We present an adaptive system for the automatic assessment of both physical and anthropic fire impact factors on periurban forestries. The aim is to provide an integrated methodology exploiting a complex data structure built upon a multi resolution grid gathering historical land exploitation and meteorological data, records of human habits together with suitably segmented and interpreted highresolution X-SAR images, and several other information sources. The contribution of the model and its novelty rely mainly on the definition of a learning schema lifting different factors and aspects of fire causes, including physical, social and behavioural ones, to the design of a fire susceptibility map, of a specific urban forestry. The outcome is an integrated geospatial database providing an infrastructure that merges cartography, heterogeneous data and complex analysis, in so establishing a digital environment where users and tools are interactively connected in an efficient and flexible way.
As the important unit of social economy, residential areas are the carriers of population, buildings and other exposures in earthquake risk assessment. The exact and rapid determination of spatial distribution of residential area is of significance in reducing uncertainty of seismic risk assessment. The residential areas are usually extracted and updated using remote sensing images, however, two major problems have to be concerned. Firstly, the residential area extraction and updating is a critical step for the processing of massive the remote sensing images with highresolution. Secondly, the identification ability of remote sensing image have limited the accuracy of residential areas because of the spectral reflectance feature of the mixed land cover type, such as houses, bare land, and vegetation, etc. Zhao P., 2003 caused by not only own spectral characteristics of internal composition (Yang C.J., 2001) but also the significant influence of terrain factors, especially in the mountains with the complex terrain and sparse buildings (Hu W.Y., 2005). In order to improve the interpretation accuracy of the residents, the methods using the elevation and slope has been well studied by researchers (Wilkinson, G.G.,1990; Cao W.F.,1998; Chen Y.H, 2006; Zhang Y.,2006).
Image fusion is one way to use remote sensing to improve our understanding of the Earth’s surface through the synthesis of huge volumes of satellite and geographical data. Pansharpening is an image-fusion technique that generates higher spatial resolution multispectral images by combining lower resolution multispectral images with higher resolution panchromatic images. Most pansharpening methods alter the spectral properties of multispectral images while improving the spatial resolution (Ehlers et al., 2010, Alparone et al., 2007). Therefore, the preservation of spectral characteristics is essential when evaluating the performance of the methods. As spectral deformation can result from the difference in the spectral wavelengths of panchromatic and multispectral images, it is necessary to investigate the influence of the spectral characteristics on the quality of pansharpened images to improve pansharpening methods.
The Illegal Building (IB) detection and monitoring has been in the spotlight in the recent decades due to the increase of construction in mega cities. The currently used method of IB monitoring is mainly based on the presence of human operator and searching all over the city randomly (Saremizadeh, 2012). Obviously, this method is cost, time, man power and petrol consuming and may be ended in untimely detection of IBs or collusion between municipality inspectors and constructors in order not to report the building infractions. This is despite the fact that the IB detection should be accurate and timely to prevent the continuation of these kinds of building constructions in the shortest possible time (Saremizadeh, 2012). Another method of IB detection, which has been taken into consideration in the recent years, is based on using high- resolution satellite images, automatic changed detection methods, and image processing techniques (Khalili Moghadam et al., 2015). Less intervention of the human operator (municipality inspectors), time saving, and cost effectiveness are the main advantages of this method. This semi-automatic method of IB detection has been developed to cover the problems of previous methods of IB monitoring. However, using the type of satellite images in IB detection process in order to achieve higher accuracies has long been an important