scholarly journals Semi-Automatic Generation of Training Samples for Detecting Renewable Energy Plants in High-Resolution Aerial Images

2021 ◽  
Vol 13 (23) ◽  
pp. 4793
Author(s):  
Maximilian Kleebauer ◽  
Daniel Horst ◽  
Christoph Reudenbach
Keyword(s):  
High Resolution ◽  
Automatic Generation ◽  
Aerial Images ◽  
Large Area ◽  
Automated Generation ◽  
Pv Systems ◽  
Large Numbers ◽  
Training Samples ◽  
Address Data ◽  
Selection Of

Deep learning (DL)—in particular convolutional neural networks (CNN)—methods are widely spread in object detection and recognition of remote sensing images. In the domain of DL, there is a need for large numbers of training samples. These samples are mostly generated based on manual identification. Identifying and labelling these objects is very time-consuming. The developed approach proposes a partially automated procedure for the sample creation and avoids manual labelling of rooftop photovoltaic (PV) systems. By combining address data of existing rooftop PV systems from the German Plant Register, the Georeferenced Address Data and the Official House Surroundings Germany, a partially automated generation of training samples is achieved. Using a selection of 100,000 automatically generated samples, a network using a RetinaNet-based architecture combining ResNet101, a feature pyramid network, a classification and a regression network is trained, applied on a large area and post-filtered by intersection with additional automatically identified locations of existing rooftop PV systems. Based on a proof-of-concept application, a second network is trained with the filtered selection of approximately 51,000 training samples. In two independent test applications using high-resolution aerial images of Saarland in Germany, buildings with PV systems are detected with a precision of at least 92.77 and a recall of 84.47.

The Problem of Using and Measurement of Identifiable Ground Control Points on High Resolution Aerial Images

2019 ◽  
Vol 31 (1) ◽  
pp. 135-144
Author(s):  
Zdzisław Kurczyński ◽  
Krzysztof Bakuła ◽  
Magdalena Pilarska ◽  
Wojciech Ostrowski
Keyword(s):  
High Resolution ◽  
Urban Areas ◽  
Aerial Images ◽  
Control Points ◽  
Linear Type ◽  
Ground Control Points ◽  
The Cost ◽  
Image Orientation ◽  
Point Type ◽  
Selection Of

Abstract This paper shows the influence of the selection of photogrammetric control points as natural, identifiable points instead of signalized, premarked control points on the results of aerial triangulation of high-resolution aerial images with GSD below 10 cm. In the experiment, different selections of controls were tested using point-type and linear-type points with measurement of their centre or corner. In the experiment, 2 blocks with GSD of 5 and 10 cm were selected using the same measurements in 4 tested approaches with sets of natural identifiable points used by comparing the result with the reference variant. The experiment proves the possibility of using natural controls instead of premarked controls for images of urban areas. This can significantly reduce the cost of photogrammetric missions in urban areas where it is easy to find uniquely identifiable control points that can be used for image orientation.

scholarly journals Performance Validation of High Resolution Digital Surface Models Generated by Dense Image Matching with the Aerial Images

2014 ◽  
Vol XL-1 ◽  
pp. 429-433
Author(s):  
N. Yastikli ◽  
H. Bayraktar ◽  
Z. Erisir
Keyword(s):  
High Resolution ◽  
Image Matching ◽  
Automatic Generation ◽  
Aerial Images ◽  
Aerial Image ◽  
Ground Sample ◽  
Dense Image ◽  
Surface Models ◽  
Performance Validation ◽  
Sample Distance

The digital surface models (DSM) are the most popular products to determine visible surface of Earth which includes all non-terrain objects such as vegetation, forest, and man-made constructions. The airborne light detection and ranging (LiDAR) is the preferred technique for high resolution DSM generation in local coverage. The automatic generation of the high resolution DSM is also possible with stereo image matching using the aerial images. The image matching algorithms usually rely on the feature based matching for DSM generation. First, feature points are extracted and then corresponding features are searched in the overlapping images. These image matching algorithms face with the problems in the areas which have repetitive pattern such as urban structure and forest. <br><br> The recent innovation in camera technology and image matching algorithm enabled the automatic dense DSM generation for large scale city and environment modelling. The new pixel-wise matching approaches are generates very high resolution DSMs which corresponds to the ground sample distance (GSD) of the original images. The numbers of the research institutes and photogrammetric software vendors are currently developed software tools for dense DSM generation using the aerial images. This new approach can be used high resolution DSM generation for the larger cities, rural areas and forest even Nation-wide applications. In this study, the performance validation of high resolution DSM generated by pixel-wise dense image matching in part of Istanbul was aimed. The study area in Istanbul is including different land classes such as open areas, forest and built-up areas to test performance of dense image matching in different land classes. The obtained result from this performance validation in Istanbul test area showed that, high resolution DSM which corresponds to the ground sample distance (GSD) of original aerial image can be generated successfully by pixel-wise dense image matching using commercial and research institution’s software.

scholarly journals VERY HIGH RESOLUTION LAND COVER MAPPING OF URBAN AREAS AT GLOBAL SCALE WITH CONVOLUTIONAL NEURAL NETWORKS

2020 ◽  
Vol XLIII-B3-2020 ◽  
pp. 201-208
Author(s):  
T. Tilak ◽  
A. Braun ◽  
D. Chandler ◽  
N. David ◽  
S. Galopin ◽  
...  
Keyword(s):  
High Resolution ◽  
Land Cover ◽  
Urban Areas ◽  
Ground Truth ◽  
Bare Soil ◽  
Aerial Images ◽  
Training Dataset ◽  
Large Area ◽  
Land Cover Map ◽  
Very High

Abstract. This paper describes a methodology to produce a 7-classes land cover map of urban areas from very high resolution images and limited noisy labeled data. The objective is to make a segmentation map of a large area (a french department) with the following classes: asphalt, bare soil, building, grassland, mineral material (permeable artificialized areas), forest and water from 20cm aerial images and Digital Height Model.We created a training dataset on a few areas of interest aggregating databases, semi-automatic classification, and manual annotation to get a complete ground truth in each class.A comparative study of different encoder-decoder architectures (U-Net, U-Net with Resnet encoders, Deeplab v3+) is presented with different loss functions.The final product is a highly valuable land cover map computed from model predictions stitched together, binarized, and refined before vectorization.

Study of Tip-Surface Interactions in Scanning Tunneling Microscopy (STM) by Reflection Electron Microscopy (REM)

1990 ◽  
Vol 48 (1) ◽  
pp. 320-321
Author(s):  
W. Lo ◽  
J.C.H. Spence ◽  
M. Kuwabara
Keyword(s):  
High Resolution ◽  
Elastic Strain ◽  
Tunneling Current ◽  
Surface Damage ◽  
Surface Interactions ◽  
Graphite Surface ◽  
Scanning Tunneling ◽  
Large Area ◽  
Tunneling Microscopy ◽  
High Resolution Tem

Work on the integration of STM with REM has demonstrated the usefulness of this combination. The STM has been designed to replace the side entry holder of a commercial Philips 400T TEM. It allows simultaneous REM imaging of the tip/sample region of the STM (see fig. 1). The REM technique offers nigh sensitivity to strain (<10−4) through diffraction contrast and high resolution (<lnm) along the unforeshortened direction. It is an ideal technique to use for studying tip/surface interactions in STM.The elastic strain associated with tunnelling was first imaged on cleaved, highly doped (S doped, 5 × 1018cm-3) InP(110). The tip and surface damage observed provided strong evidence that the strain was caused by tip/surface contact, most likely through an insulating adsorbate layer. This is consistent with the picture that tunnelling in air, liquid or ordinary vacuum (such as in a TEM) occurs through a layer of contamination. The tip, under servo control, must compress the insulating contamination layer in order to get close enough to the sample to tunnel. The contaminant thereby transmits the stress to the sample. Elastic strain while tunnelling from graphite has been detected by others, but never directly imaged before. Recent results using the STM/REM combination has yielded the first direct evidence of strain while tunnelling from graphite. Figure 2 shows a graphite surface elastically strained by the STM tip while tunnelling (It=3nA, Vtip=−20mV). Video images of other graphite surfaces show a reversible strain feature following the tip as it is scanned. The elastic strain field is sometimes seen to extend hundreds of nanometers from the tip. Also commonly observed while tunnelling from graphite is an increase in the RHEED intensity of the scanned region (see fig.3). Debris is seen on the tip and along the left edges of the brightened scan region of figure 4, suggesting that tip abrasion of the surface has occurred. High resolution TEM images of other tips show what appear to be attached graphite flakes. The removal of contamination, possibly along with the top few layers of graphite, seems a likely explanation for the observed increase in RHEED reflectivity. These results are not inconsistent with the “sliding planes” model of tunnelling on graphite“. Here, it was proposed that the force due to the tunnelling probe acts over a large area, causing shear of the graphite planes when the tip is scanned. The tunneling current is then modulated as the planes of graphite slide in and out of registry. The possiblity of true vacuum tunnelling from the cleaned graphite surface has not been ruled out. STM work function measurements are needed to test this.

High Resolution Displacement of Functional Groups onto Fluorocarbon Resin Surface by Using ArF Excimer Laser

1996 ◽  
Vol 451 ◽  
Author(s):  
T. Shimizu ◽  
M. Murahara
Keyword(s):  
Contact Angle ◽  
High Resolution ◽  
Laser Beam ◽  
Thin Layer ◽  
Functional Groups ◽  
Excimer Laser ◽  
Laser Fluence ◽  
Large Area ◽  
Three Solutions ◽  
Arf Excimer Laser

ABSTRACTA Fluorocarbon resin surface was selectively modified by irradiation with a ArF laser beam through a thin layer of NaAlO2, B(OH)3, or H2O solution to give a hydrophilic property. As a result, with low fluence, the surface was most effectively modified with the NaAlO2 solution among the three solutions. However, the contact angle in this case changed by 10 degrees as the fluence changed only 1mJ/cm2. When modifying a large area of the surface, high resolution displacement could not be achieved because the laser beam was not uniform in displacing functional groups. Thus, the laser fluence was successfully made uniform by homogenizing the laser beam; the functional groups were replaced on the fluorocarbon resin surface with high resolution, which was successfully modified to be hydrophilic by distributing the laser fluence uniformly.

Mapping and Monitoring of the Land Use/Cover Changes in the Wider Area of ltanos, Crete, Using Very High Resolution EO Imagery With Specific Interest in Archaeological Sites

2019 ◽  
Author(s):  
Sawyer Reid stippa ◽  
George Petropoulos ◽  
Leonidas Toulios ◽  
Prashant K. Srivastava
Keyword(s):  
High Resolution ◽  
Spatial Resolution ◽  
Satellite Images ◽  
Archaeological Site ◽  
Digital Photography ◽  
Archaeological Sites ◽  
Large Area ◽  
High Resolution Imagery ◽  
Site Mapping ◽  
Rule Sets

Archaeological site mapping is important for both understanding the history as well as protecting them from excavation during the developmental activities. As archaeological sites generally spread over a large area, use of high spatial resolution remote sensing imagery is becoming increasingly applicable in the world. The main objective of this study was to map the land cover of the Itanos area of Crete and of its changes, with specific focus on the detection of the landscape’s archaeological features. Six satellite images were acquired from the Pleiades and WorldView-2 satellites over a period of 3 years. In addition, digital photography of two known archaeological sites was used for validation. An Object Based Image Analysis (OBIA) classification was subsequently developed using the five acquired satellite images. Two rule-sets were created, one using the standard four bands which both satellites have and another for the two WorldView-2 images their four extra bands included. Validation of the thematic maps produced from the classification scenarios confirmed a difference in accuracy amongst the five images. Comparing the results of a 4-band rule-set versus the 8-band showed a slight increase in classification accuracy using extra bands. The resultant classifications showed a good level of accuracy exceeding 70%. Yet, separating the archaeological sites from the open spaces with little or no vegetation proved challenging. This was mainly due to the high spectral similarity between rocks and the archaeological ruins. The satellite data spatial resolution allowed for the accuracy in defining larger archaeological sites, but still was a difficulty in distinguishing smaller areas of interest. The digital photography data provided a very good 3D representation for the archaeological sites, assisting as well in validating the satellite-derived classification maps. All in all, our study provided further evidence that use of high resolution imagery may allow for archaeological sites to be located, but only where they are of a suitable size archaeological features.

Mapping of landscape-morphological structure of the Northern-Western part of Anapa bay bar by digital high resolution aerial images

2017 ◽  
Vol 927 (9) ◽  
pp. 22-29
Author(s):  
V.I. Kravtsovа ◽  
E.R. Chalova
Keyword(s):  
High Resolution ◽  
Morphological Structure ◽  
Middle Part ◽  
Aerial Images ◽  
Medical Resource ◽  
Sea Waves ◽  
Coast Line ◽  
Waves Interaction ◽  
Account Due ◽  
Digital Landscape

Anapa bay bar is a valuable recreational-medical resource. Digital landscape-morphological mapping of its the Northern-Western part was created by digital aero survey materials for monitoring of its statement. Compiled maps show that in the Western part of region dune belt is degradated, front dune hills destroyed due to spreading of settlement Veselovka buildings to beach, and due to mass enactments carrying out at bay bar of lake Solenoe. Here it is necessary to decide the problem of defense from waves flooding by construction of artificial hills. The middle part of region, around Bugaz lagoon, is using for unregulated recreation of extreme sportsmen – windsurfing and kiting – with seasonal recreation in camping from tent-city and campers. Many short roads to sea beach, orthogonal to coast line, have been transformed to corridors of blowing and sea waves interaction to lagoon lowland with dune belt destroying. In the Eastern part of region, at Bugaz bay bar, dune belt is conserve, it changes under natural sea and wind processes action. At some places sea waves are erode windward front dune slope. Just everywhere sand accumulative trains are forming at leeward slope of front dune. Showed peculiarities of landscape morphological structure mast be taken in account due treatment of measures for bay bar defense and keeping.

Sign in / Sign up

Export Citation Format

Share Document