A Hybrid Approach for Building Extraction From Spaceborne Multi-Angular Optical Imagery

Turlapaty, Anish C.; Gokaraju, Balakrishna; Du, Qian; Younan, Nicolas H.; Aanstoos, James V.

doi:10.1109/jstars.2011.2179792

Cited by 47 publications

(25 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [28,38], buildings and their shadows are detected using MBI and MSI, and then used to detect built-up areas from general remotely sensed imagery. Stereo imagery consists of two or three images having different viewing angles, and thus can be used to produce height information (e.g., DSMs, and disparity images) by image matching.…”

Section: Built-up Area Detection Methods Using Height Informationmentioning

confidence: 99%

A New Stereo Pair Disparity Index (SPDI) for Detecting Built-Up Areas from High-Resolution Stereo Imagery

et al. 2017

View full text Add to dashboard Cite

Within-class spectral variation and between-class spectral confusion in remotely sensed imagery degrades the performance of built-up area detection when using planar texture, shape, and spectral features. Terrain slopes and building heights extracted from auxiliary data, such as Digital Surface Models (DSMs) however, can improve the results. Stereo imagery incorporates height information unlike single remotely sensed images. In this study, a new Stereo Pair Disparity Index (SPDI) for indicating built-up areas is calculated from stereo-extracted disparity information. Further, a new method of detecting built-up areas from stereo pairs is proposed based on the SPDI, using disparity information to establish the relationship between two images of a stereo pair. As shown in the experimental results for two stereo pairs covering different scenes with diverse urban settings, the SPDI effectively differentiates between built-up and non-built-up areas. Our proposed method achieves higher accuracy built-up area results from stereo images than the traditional method for single images, and two other widely-applied DSM-based methods for stereo images. Our approach is suitable for spaceborne and airborne stereo pairs and triplets. Our research introduces a new effective height feature (SPDI) for detecting built-up areas from stereo imagery with no need for DSMs.

show abstract

Section: Built-up Area Detection Methods Using Height Informationmentioning

confidence: 99%

A New Stereo Pair Disparity Index (SPDI) for Detecting Built-Up Areas from High-Resolution Stereo Imagery

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Different techniques take advantage of specific acquisition techniques such as Light Detection and Ranging (LiDAR), or optical stereo images, to build a Digital Terrain Model (DTM) and a Digital Surface Model (DSM), and to then derive the heights of buildings [5,6]. Another possible set of techniques make use of multiple optical views and template matching algorithms [7,8]. Oblique airborne imagery has also been considered in other efforts [9].…”

Section: Existing Approaches To Floor Number Determinationmentioning

confidence: 99%

Extensive Exposure Mapping in Urban Areas through Deep Analysis of Street-Level Pictures for Floor Count Determination

Iannelli¹,

Dell’Acqua

2017

Urban Science

View full text Add to dashboard Cite

Abstract:In order for a risk assessment to deliver sensible results, exposure in the concerned area must be known or at least estimated in a reliable manner. Exposure estimation, though, may be tricky, especially in urban areas, where large-scale surveying is generally expensive and impractical; yet, it is in urban areas that most assets are at stake when a disaster strikes. Authoritative sources such as cadastral data and business records may not be readily accessible to private stakeholders such as insurance companies; airborne and especially satellite-based Earth-Observation data obviously cannot retrieve all relevant pieces of information. Recently, a growing interest is recorded in the exploitation of street-level pictures, procured either through crowdsourcing or through specialized services like Google Street View. Pictures of building facades convey a great amount of information, but their interpretation is complex. Recently, however, smarter image analysis methods based on deep learning started appearing in literature, made possible by the increasing availability of computational power. In this paper, we leverage such methods to design a system for large-scale, systematic scanning of street-level pictures intended to map floor numbers in urban buildings. Although quite simple, this piece of information is a relevant exposure proxy in risk assessment. In the proposed system, a series of georeferenced images are automatically retrieved from the repository where they sit. A tailored deep learning net is first trained on sample images tagged through visual interpretation, and then systematically applied to the entire retrieved dataset. A specific algorithm allows attaching "number of floors" tags to the correct building in a dedicated GIS (Geographic Information System) layer, which is finally output by the system as an "exposure proxy" layer.

show abstract

“…First, the pan-sharpening is performed on the 2 meter multispectral datasets of Worldview-2 and 50 centimeter panchromatic data. The shadows in the pan-sharpened image are masked using the product yellow and red bands form the MS dataset [3]. Successively from the above masked image, the relative height of the buildings are obtained by using the template matching algorithm.…”

Section: Experimentation Of 2d-fft Parallelizationmentioning

confidence: 99%

“…Implementation of the template matching algorithms was discussed elaborately in our earlier study given in the Ref. [3]. Template matching algorithm evaluates for the matching frame sizes by Fourier transforming both frames of two multi-view angle images and obtaining their correlations in the inverse joint Fourier distribution.…”

Section: Experimentation Of 2d-fft Parallelizationmentioning

confidence: 99%

Comprehensive review of evolution of satellite sensor specifications against speedup performance of pattern recognition algorithms in remote sensing

Gokaraju

Bhushan

Anantharaj³

et al. 2015

2015 IEEE Applied Imagery Pattern Recognition Workshop (AIPR)

Self Cite

View full text Add to dashboard Cite

The objectives of this study are as follows: (i) Discuss the necessity of HPC in remote sensing community towards contemporary scientific solution requirements; (ii) Investigate the speedup in performance of the template matching algorithm with FFT parallelization using hybrid Central Processing Units (CPUs)/Graphics Processing Units (GPUs); (iii) Apply the speedup algorithms for detection of real-time man-made structures such as buildings from remote sensing datasets, for constructing a 3-Dimensional city modelling.Index Terms-high performance computing, pattern recognition, remote sensing, speed-up, template matching. REMTOE SENSING as BIG DATA SCIENCE PROBLEMThe influx of earth observation data acquired by both airborne and space-borne imaging sensors has outgrown tremendously in the present decade since their inception. It is very noteworthy in terms of improvements in the sensor technology based on spatial, spectral, temporal and multiangular resolutions producing nearly continual stream of high dimensional data. All these developments lead to a rapid rise in diversity of remote sensing applications benefiting largely with the discounted prices and easy access to such data. Lately, on-demand ready and near real-time availability of information has become very key components in any decision support/emergency response system exploiting huge voluminous data. With this proliferation of data and need for fast processing, the commercial industry and academic community are relying more towards automatic/semi-automatic approaches rather than conventional visual interpretation and supervised image processing. It is highly efficient to integrate advanced pattern recognition or signal processing techniques for automation of remote sensing information processing and dissemination of decisions through a GIS database management system. However, these machine intelligent tasks are more complex to implement on a larger scale basis, and thus require high computational resources and high performance computing solvers.The peer-review literature in IEEE geoscience and remote sensing society (GRSS) shows a very less number of experimental studies incorporating high performance computing (HPC) strategies such as MPI based parallel programming, OpenMP based multi-threading, and graphic processor unit based accelerators towards remote sensing solutions [6], [9]. HPC for remote sensing applications is relatively very new compared to computational fluid dynamics (CFD) field, where the latter evolved over three decades. Ref.[6], published in Sep. 2011 is one of the first volumes to explore state-of-the-art HPC techniques in the context of remote sensing studies [19]. It focuses on the computational complexity of hyperspectral processing algorithms, which is in great demand of parallel computing. Moreover, the improvements in the high performance computational ability of remote sensing and GIS application solutions, especially filtering or classification techniques, have not shown the same speedup as that of the evolved computational r...

show abstract

A Hybrid Approach for Building Extraction From Spaceborne Multi-Angular Optical Imagery

Cited by 47 publications

References 32 publications

A New Stereo Pair Disparity Index (SPDI) for Detecting Built-Up Areas from High-Resolution Stereo Imagery

A New Stereo Pair Disparity Index (SPDI) for Detecting Built-Up Areas from High-Resolution Stereo Imagery

Extensive Exposure Mapping in Urban Areas through Deep Analysis of Street-Level Pictures for Floor Count Determination

Comprehensive review of evolution of satellite sensor specifications against speedup performance of pattern recognition algorithms in remote sensing

Contact Info

Product

Resources

About