Spelling suggestions: "subject:"endmember"" "subject:"andmember""
1 |
Démixage d’images hyperspectrales en présence d’objets de petite taille / Spectral unmixing of hyperspectral images in the presence of small targetsRavel, Sylvain 08 December 2017 (has links)
Cette thèse est consacrée au démixage en imagerie hyperspectrale en particulier dans le cas où des objets de petite taille sont présents dans la scène. Les images hyperspectrales contiennent une grande quantité d’information à la fois spectrale et spatiale, et chaque pixel peut être vu comme le spectre de réflexion de la zone imagée. Du fait de la faible résolution spatiale des capteurs le spectre de réflexion observé au niveau de chaque pixel est un mélange des spectres de réflexion de l’ensemble des composants imagés dans le pixel. Une problématique de ces images hyperspectrales est le démixage, qui consiste à décomposer l’image en une liste de spectres sources, appelés endmembers, correspondants aux spectres de réflexions des composants de la scène d’une part, et d’autre part la proportion de chacun de ces spectres source dans chaque pixel de l’image. De nombreuses méthodes de démixage existent mais leur efficacité reste amoindrie en présence de spectres sources dits rares (c’est-à-dire des spectres présents dans très peu de pixels, et souvent à un niveau subpixelique). Ces spectres rares correspondent à des composants présents en faibles quantités dans la scène et peuvent être vus comme des anomalies dont la détection est souvent cruciale pour certaines applications.Nous présentons dans un premier temps deux méthodes de détection des pixels rares dans une image, la première basée sur un seuillage de l’erreur de reconstruction après estimation des endmembers abondants, la seconde basée sur les coefficients de détails obtenus par la décomposition en ondelettes. Nous proposons ensuite une méthode de démixage adaptée au cas où une partie des endmembers sont connus a priori et montrons que cette méthode utilisée avec les méthodes de détection proposées permet le démixage des endmembers des pixels rares. Enfin nous étudions une méthode de rééchantillonnage basée sur la méthode du bootstrap pour amplifier le rôle de ces pixels rares et proposer des méthodes de démixage en présence d’objets de petite taille. / This thesis is devoted to the unmixing issue in hyperspectral images, especiallyin presence of small sized objects. Hyperspectral images contains an importantamount of both spectral and spatial information. Each pixel of the image canbe assimilated to the reflection spectra of the imaged scene. Due to sensors’ lowspatial resolution, the observed spectra are a mixture of the reflection spectraof the different materials present in the pixel. The unmixing issue consists inestimating those materials’ spectra, called endmembers, and their correspondingabundances in each pixel. Numerous unmixing methods have been proposed butthey fail when an endmembers is rare (that is to say an endmember present inonly a few of the pixels). We call rare pixels, pixels containing those endmembers.The presence of those rare endmembers can be seen as anomalies that we want todetect and unmix. In a first time, we present two detection methods to retrievethis anomalies. The first one use a thresholding criterion on the reconstructionerror from estimated dominant endmembers. The second one, is based on wavelettransform. Then we propose an unmixing method adapted when some endmembersare known a priori. This method is then used with the presented detectionmethod to propose an algorithm to unmix the rare pixels’ endmembers. Finally,we study the application of bootstrap resampling method to artificially upsamplerare pixels and propose unmixing methods in presence of small sized targets.
|
2 |
Detektion von Referenzspektren in multisensoralen BilddatenGreiwe, Ansgar 22 May 2006 (has links)
Liegen für ein Untersuchungsgebiet spektral und geometrisch hoch aufgelöste Daten unterschiedlicher Sensoren vor, liegt eine kombinierte Nutzung der Datenquellen zur Optimierung der Klassifikationsergebnisse nahe. In einem entscheidungsbasierten Fusionsansatz wird die Klassifikationsgenauigkeit von Bildsegmenten geometrisch hoch aufgelöster Bilddaten durch die Einbindung zusätzlicher Materialinformationen gesteigert. Diese werden aus den Ergebnissen eines Klassifikationsverfahrens zur Materialdetektion, dem Spectral Angle Mapper, abgeleitet. Die zur Auswertung der hyperspektralen Bilddaten notwendigen Referenzspektren wurden in den Untersuchungen bislang manuell definiert. Nachteil dieser manuellen Referenzspektrenselektion ist die subjektive Auswahl der hyperspektralen Bildpixel, deren Spektren als Referenz in die Analyse eingehen. In dieser Arbeit wird ein Verfahren zur automatischen segmentbasierten Referenzspektrenselektion vorgestellt, dessen Konzept auf der Berechnung der spektralen Ähnlichkeit so genannter Referenzkandidaten und die anschließende Gruppierung ähnlicher Bildpixel basiert. Ein Maß für die spektrale Ähnlichkeit der Referenzkandidaten wird durch die Berechnung der Korrelationskoeffizienten ihrer Reflektanzspektren ermittelt. Die Gruppierung spektral ähnlicher Kandidaten erfolgt durch ein dichte-basiertes Clustering. Die Referenzspektren werden abschließend durch die Mittelung der Einzelspektren eines Clusters erzeugt.Die Leistungsfähigkeit des vorgestellten Gesamtkonzeptes zur automatisierten segmentbasierten Referenzspektrenselektion wird in einer Konzeptstudie mit ausgewählten Referenzflächen für elf unterschiedliche Materialien demonstriert. In einer abschließenden Auswertung eines Testgebietes wird die Anwendbarkeit des Konzeptes nachgewiesen
|
3 |
Improved Endmember Mixing Analysis (EMMA): Application to a Nested Catchment, Provo River, Northern UtahThompson, Alyssa Nicole 15 August 2023 (has links) (PDF)
An endmember mixing analysis (EMMA) is a hydrograph separation technique used to identify and quantify stream source contributions, but the error within the results of the analysis itself can be difficult to quantify. Employing EMMA to accurately quantify these contributions is particularly important for critical watersheds that supply water to large populations, such as montane watersheds. We applied EMMA to the Provo River, a nested catchment with three monitoring locations in northern Utah, to understand the limitations and potential improvements that could be made to EMMA. Four main endmembers (quartzite groundwater, soil water, snow and carbonate groundwater) were identified for the watershed and differentiated using the conservative tracers δ18O, δ2H, Si, HCO3-, Mg2+, K+, and Ca2+. In a traditional EMMA approach, a principal components analysis (PCA) is used to identify endmembers for a single location in a watershed, and the principal component (PC) scores are used to calculate the fractional contributions of each endmember. However, we found that calculating the fractional contributions of the endmembers in tracer space resulted in less error in the calculations compared to performing the calculation in PC defined space (U-space). Performing the mixing in tracer space with four endmembers showed that during spring runoff, snow was the main endmember with inputs ranging from 23 – 66% for the highest part of the watershed and 14 – 60% for the lowest part of the watershed. During baseflow, the stream was dominated by groundwater with contributions ranging from 23 – 60% quartzite groundwater for the upper part of the watershed and 30 – 57% carbonate groundwater for the lower part of the watershed. The amount of error present in the results depended on the scale of the catchment and the number of endmembers included, with more error in downstream locations relative to upstream locations. The nested catchment approach is a further improvement on traditional EMMA because it allows for identification of missing endmembers and error analysis for characterizing stream chemistry in several locations in a complex watershed.
|
4 |
Accelerated Hyperspectral Unmixing with Endmember Variability via the Sum-Product AlgorithmPuladas, Charan 26 May 2016 (has links)
No description available.
|
5 |
Mapping and Assessing Urban Impervious Areas Using Multiple Endmember Spectral Mixture Analysis: A Case Study in the City of Tampa, FloridaWeng, Fenqing 01 January 2012 (has links)
The advance in remote sensing technology helps people more easily assess urban growth. In this study, the utility of multiple endmember spectral mixture analysis (MESMA) is examined in a sub-pixel analysis of Landsat Thematic Mapper (TM) imagery to map urban physical components in Tampa, FL. The three physical components of urban land cover (LC): impervious surface, vegetation and soil, were compared using the proposed MESMA with a traditional spectral mixture analysis (SMA). MESMA decomposes each pixel to address the heterogeneity of urban LC characteristic by allowing the number and types of endmembers to vary on a per pixel basis. This study generated 642 spectral mixture models of 2-, 3-, and 4-endmembers for each pixel to estimate the fractions of impervious surface, vegetation, soil, and shade in the study area with a constraint of lowest root mean square error (RMSE). A comparative analysis of the impervious surface areas (ISA) mapped with MESMA and SMA demonstrated that MESMA produced more accurate results of mapping urban physical components than those by SMA. With the multiyear Landsat TM data, we quantified sub-pixel %ISA and the %ISA changes to assess urban growth in the City of Tampa, Florida during the past twenty years. The experimental results demonstrate that the MESMA approach is effective in mapping and monitoring urban land use/land cover changes using moderate-resolution multispectral imagery at a sub-pixel level.
|
Page generated in 0.0372 seconds