Demonstration venue · read-only. Every page can be browsed; the buttons that would change it are switched off. Create an account to run TaxoReview on your own data. A simulation from public publication records: the people listed were not invited and did not agree to review.
? Tutorial for this page

Michael Hahn 0003

dblp:44/9903-3 · DBLP profile ↗
← Back
5ranked-venue papers
0as first author
2since 2021 · last 2024
0000-0002-7809-1796ORCID · verified

Domains — the database's venue-derived domains; a paper can count in several

Applied, interdisciplinary, general and emerging computing · 3 · 2 since 2021Artificial intelligence and machine learning · 1Databases, data management, data science and information retrieval · 1

Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.

Computer graphics and multimedia
1 paper
Image and video processing · 67% Geometric modeling and processing · 33%

Topics — the 3 heaviest of 3, each with the papers that count most for it

TopicWeightPapersLastEvidence papers
Image and video processing
image segmentation
0.412019
Segmentation of Laser Point Clouds in Urban Areas by a Modified Normalized Cut Method · IEEE Trans. Pattern Anal. Mach. Intell. 2019
Geometric modeling and processing › point cloud processing
point cloud segmentation
0.412019
Segmentation of Laser Point Clouds in Urban Areas by a Modified Normalized Cut Method · IEEE Trans. Pattern Anal. Mach. Intell. 2019
Image and video processing › image segmentation › graph-based segmentation
spectral segmentation
0.412019
Segmentation of Laser Point Clouds in Urban Areas by a Modified Normalized Cut Method · IEEE Trans. Pattern Anal. Mach. Intell. 2019

Methods — techniques the papers use, named apart from their topics

normalized cut · 0.4krylov subspace method · 0.4eigenvalue problem · 0.4
YearPublicationVenuePosition
2024 ICESat-2 Data Application for DEM Bias Compensation Based on Point-to-Surface Matching
abstract
Creating high-quality digital elevation models (DEMs) from high-resolution satellite images requires an accurate sensor model. Rational polynomial coefficients (RPCs) have been widely employed instead of rigorous sensor models. However, the generated DEMs can be distorted because of the inherent RPC biases. Conventionally, the bias is compensated by incorporating ground control points (GCPs) into the standard processing workflow, which requires additional effort to measure the ground and image coordinates of the GCPs. Furthermore, in an inaccessible area, such as polar regions, it is difficult to perform field surveying of ground coordinates. In this study we investigate whether and how the data recorded by NASA’s Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) mission can be used as GCPs without requiring an operator to measure the image coordinates. The first step is generating DEMs by image matching from high-resolution satellite images with the given RPCs. A point-to-surface matching method that matches the ICESat-2 data to the DEM surface was developed to correct the DEM and improve its precision. For the experimental investigations, we use KOMPSAT-3 imagery and ICESat-2 ATL08 data for a Korea test site and KOMPSAT-3A imagery along with ICESat-2 ATL06 and ATL08 data for a test site at the Thwaites Glacier in Antarctica. For the Korea test site, we additionally perform a land cover classification to select ATL08 segments that are uniquely assignable to the terrain surface. After point-to-surface matching, the DEMs are corrected in both test areas. The accuracy analyses show that the proposed method improves the accuracy (root mean square error) of the corrected DEM from 10 m to 2 m for the Korea test site and from 14 m to 1 m for the Antarctic test site. Furthermore, we placed particular emphasis on the terrain slope. To investigate the impact of terrain slope on the bias compensation process, we separate the ATL08 data of the Korea test site into groups of different terrain slope. It turns out that the deviation of the ATL08 data is lowest at about 1 m when the DEM slope is within 10°. Beyond a 30° slope, the RMSE increases by a factor of 7 compared to the 0-10° range.
HyoSeong Lee, Michael Hahn 0003
IEEE Trans. Geosci. Remote. Sens.2
2021 High-Resolution Satellite Stereo Matching by Object-Based Semiglobal Matching and Iterative Guided Edge-Preserving Filter
abstract
This letter presents an object-based stereo matching and content-based guided disparity map refinement for high-resolution satellite stereo images. The proposed method uses the superpixels for object-based cost filtering and then object-based semiglobal matching. In the proposed object-based stereo matching strategy, two improvements including homogeneity weight for cost filtering, and weighted cost aggregation by image objects are considered. After cost aggregation, the generated disparity map is computed using the winner takes all and then refined with a new iterative guided edge-preserving filter. The proposed method with Census cost function has been implemented on high-resolution satellite stereo images and then is compared with the LiDAR ground truth. Moreover, the proposed method is compared with five state-of-the-art stereo matching methods. The experimental results on along-track satellite stereo images from Pleiades and IKONOS images and cross-track multidate images from WV-III demonstrate that the proposed method significantly improves the result of stereo matching.
Nurollah Tatar, Hossein Arefi, Michael Hahn 0003
IEEE Geosci. Remote. Sens. Lett.3
2019 Segmentation of Laser Point Clouds in Urban Areas by a Modified Normalized Cut Method
abstract
Normalized Cut is a well-established divisive image segmentation method, which we adapt in this paper for the segmentation of laser point clouds in urban areas. Our focus is on polyhedral objects with planar surfaces. Due to its target function, Normalized Cut favours cuts with "short cut lines" or "small cut surfaces", which is a drawback for our application. We therefore modify the target function, weighting the similarity measures with distance-dependent weights. We call the induced minimization problem "Distance-weighted Cut" (DWCut). The new target function leads to a generalized eigenvalue problem, which is slightly more complicated than the corresponding problem for the Normalized Cut; on the other hand, the new target function is easier to interpret and avoids some drawbacks of the Normalized Cut. We point out an efficient method for the numerical solution of the eigenvalue problem which is based on a Krylov subspace method. DWCut can be beneficially combined with an aggregation in order to reduce the computational effort and to avoid shortcomings due to insufficient plane parameters. We present examples for the successful application of the Distance-weighted Cut principle and evaluate its results by comparison with the results of corresponding manual segmentations.
Avishek Dutta, Johannes Engels, Michael Hahn 0003
IEEE Trans. Pattern Anal. Mach. Intell.3
2016 Wavelet-morphology based detection of incipient linear cracks in asphalt pavements from RGB camera imagery and classification using circular Radon transform
Yashon O. Ouma, Michael Hahn 0003
Adv. Eng. Informatics2
2007 Semi-automatic true orthophoto production by using LIDAR data
abstract
Uzun yıllardan beri orman arazilerinin kontrolü ve yönetimi, su baskınlarına karşı risk analizi gibi çeşitli uygulamalarda kullanılan LIDAR (LIght Detection and Ranging) teknolojisi, son yıllarda temelinde Gerçek Ortofoto üretimine dayalı üç boyutlu şehir modeli oluşturulmasında popüler olarak kullanılmaya başlanmıştır. Geleneksel Ortofoto üretiminde Sayısal Arazi Modeli’nin (SAM) kullanılması sonucu, binalar ve arazi üzerinde kalan diğer objeler, perspektif iz düşümden kaynaklanan görüntü ötelemesinden dolayı Ortofoto üzerinde doğru olarak konumlandırılamamaktadır. Ortofoto üretiminin, SYM kullanılarak gerçekleştirilmesiyle bu sorunlar büyük ölçüde aşılmakta ve elde edilen Ortofoto “Gerçek Ortofoto” olarak adlandırılmaktadır. Buna rağmen, gerçek Ortofoto üretimi farklı sorunları beraberinde getirmektedir. Binaların gerçek konumlarına getirilmesi sonucunda, Ortofoto üzerinde resim bilgisi olmayan alanlar oluşmaktadır. Bunların yanında aşılması gereken diğer bir sorun, gerçek Ortofotoda; yolların, köprülerin ve bina çatı kenarlarının sınırlarının keskin bir şekilde görünmemesidir ve şimdiye kadar yapılan araştırmalarda bu sorunun aşılmasında SYM’nin kalitesinin önemli bir rol oynadığı ve Lazer noktaları ile üretilen SYM’nin tek başına yeterli olmadığı gözlenmiştir. Bu tez çalışmasında, Ortofotonun geliştirilmesi üzerine uygun methodlar araştırılmış ve iki farklı method test edilmiştir. (1) İlk metod olarak, gerçek Ortofoto üretimi için fotogrametride yaygın bir şekilde kullanılan SYM gerçek Ortofoto üretiminde kullanılmış ve (2) ikinci method olarak, hava fotoğraflarının üç boyutlu değerlendirilmesi ile sayısallaştırılan binaların çatı kenarları, vektör veri olarak SAM’ın üzerine eklenerek gerçek Ortofoto üretilmiştir. Bu bildiride, üretim işlem adımları ve karşılaşılan sorunlara detaylı bir şekilde değinilerek, iki farklı yöntemle üretilen gerçek Ortofotonun karşılaştırılması anlatılmaktadır. Deneysel araştırma için Stuttgart şehir merkezinin Lidar veri seti kullanılmıştır.
Arif Günay, Hossein Arefi, Michael Hahn 0003
IGARSS3