Refine
Has Fulltext
- yes (26)
Is part of the Bibliography
- yes (26)
Year of publication
Document Type
- Master Thesis (26) (remove)
Language
- English (26) (remove)
Keywords
- China (6)
- Diamiktit (2)
- Dwyka (2)
- Fernerkundung (2)
- Fossil (2)
- Geochemie (2)
- Karbon (2)
- Ming dynasty (2)
- Namibia (2)
- Oranje (2)
Institute
- Institut für Kulturwissenschaften Ost- und Südasiens (9)
- Institut für Geographie (2)
- Institut für Geologie (2)
- Institut für Informatik (2)
- Institut für Mathematik (2)
- Physikalisches Institut (2)
- Theodor-Boveri-Institut für Biowissenschaften (2)
- Fakultät für Biologie (1)
- Fakultät für Physik und Astronomie (1)
- Institut für Anglistik und Amerikanistik (1)
- Institut für Geographie und Geologie (1)
- Institut für Psychologie (bis Sept. 2007) (1)
- Philosophische Fakultät (Histor., philolog., Kultur- und geograph. Wissensch.) (1)
- Rudolf-Virchow-Zentrum (1)
Sonstige beteiligte Institutionen
Streaming of videos has become the major traffic generator in today's Internet and the video traffic share is still increasing. According to Cisco's annual Visual Networking Index report, in 2012, 60% of the global Internet IP traffic was generated by video streaming services. Furthermore, the study predicts further increase to 73% by 2017. At the same time, advances in the fields of mobile communications and embedded devices lead to a widespread adoption of Internet video enabled mobile and wireless devices (e.g. Smartphones). The report predicts that by 2017, the traffic originating from mobile and wireless devices will exceed the traffic from wired devices and states that mobile video traffic was the source of roughly half of the mobile IP traffic at the end of 2012.
With the increasing importance of Internet video streaming in today's world, video content provider find themselves in a highly competitive market where user expectations are high and customer loyalty depends strongly on the user's satisfaction with the provided service. In particular paying customers expect their viewing experience to be the same across all their viewing devices and independently of their currently utilized Internet access technology. However, providing video streaming services is costly in terms of storage space, required bandwidth and generated traffic. Therefore, content providers face a trade-off between the user perceived Quality of Experience (QoE) and the costs for providing the service.
Today, a variety of transport and application protocols exist for providing video streaming services, but the one utilized depends on the scenario in mind. Video streaming services can be divided up in three categories: Video conferencing, IPTV and Video-on-Demand services. IPTV and video-conferencing have severe real-time constraints and thus utilize mostly datagram-based protocols like the RTP/UDP protocol for the video transmission. Video-on-Demand services in contrast can profit from pre-encoded content, buffers at the end user's device, and mostly utilize TCP-based protocols in combination with progressive streaming for the media delivery.
In recent years, the HTTP protocol on top of the TCP protocol gained widespread popularity as a cost-efficient way to distribute pre-encoded video content to customers via progressive streaming. This is due to the fact that HTTP-based video streaming profits from a well-established infrastructure which was originally implemented to efficiently satisfy the increasing demand for web browsing and file downloads. Large Content Delivery Networks (CDN) are the key components of that distribution infrastructure. CDNs prevent expensive long-haul data traffic and delays by distributing HTTP content to world-wide locations close to the customers. As of 2012, already 53% of the global video traffic in the Internet originates from Content Delivery Networks and that percentage is expected to increase to 65% by the year 2017. Furthermore, HTTP media streaming profits from existing HTTP caching infrastructure, ease of NAT and proxy traversal and firewall friendliness.
Video delivery through heterogeneous wired and wireless communications networks is prone to distortions due to insufficient network resources. This is especially true in wireless scenarios, where user mobility and insufficient signal strength can result in a very poor transport service performance (e.g. high packet loss, delays and low and varying bandwidth). A poor performance of the transport in turn may degrade the Quality of Experience as perceived by the user, either due to buffer underruns (i.e. playback interruptions) for TCP-based delivery or image distortions for datagram-based real-time video delivery.
In order to overcome QoE degradations due to insufficient network resources, content provider have to consider adaptive video streaming. One possibility to implement this for HTTP/TCP streaming is by partitioning the content into small segments, encode the segments into different quality levels and provide access to the segments and the quality level details (e.g. resolution, average bitrate). During the streaming session, a client-centric adaptation algorithm can use the supplied details to adapt the playback to the current environment. However, a lack of a common HTTP adaptive streaming standard led to multiple proprietary solutions developed by major Internet companies like Microsoft (Smooth Streaming), Apple (HTTP Live Streaming) and Adobe (HTTP Dynamic Streaming) loosely based on the aforementioned principle. In 2012, the ISO/IEC published the Dynamic Adaptive Streaming over HTTP (MPEG-DASH) standard. As of today, DASH is becoming widely accepted with major companies announcing their support or having already implemented the standard into their products. MPEG-DASH is typically used with single layer codecs like H.264/AVC, but recent publications show that scalable video coding can use the existing HTTP infrastructure more efficiently. Furthermore, the layered approach of scalable video coding extends the adaptation options for the client, since already downloaded segments can be enhanced at a later time.
The influence of distortions on the perceived QoE for non-adaptive video streaming are well reviewed and published. For HTTP streaming, the QoE of the user is influenced by the initial delay (i.e. the time the client pre-buffers video data) and the length and frequency of playback interruptions due to a depleted video playback buffer. Studies highlight that even low stalling times and frequencies have a negative impact on the QoE of the user and should therefore be avoided. The first contribution of this thesis is the identification of QoE influence factors of adaptive video streaming by the means of crowd-sourcing and a laboratory study.
MPEG-DASH does not specify how to adapt the playback to the available bandwidth and therefore the design of a download/adaptation algorithm is left to the developer of the client logic. The second contribution of this thesis is the design of a novel user-centric adaption logic for DASH with SVC. Other download algorithms for segmented HTTP streaming with single layer and scalable video coding have been published lately. However, there is little information about the behavior of these algorithms regarding the identified QoE-influence factors. The third contribution is a user-centric performance evaluation of three existing adaptation algorithms and a comparison to the proposed algorithm. In the performance evaluation we also evaluate the fairness of the algorithms. In one fairness scenario, two clients deploy the same adaptation algorithm and share one Internet connection. For a fair adaptation algorithm, we expect the behavior of the two clients to be identical. In a second fairness scenario, one client shares the Internet connection with a large HTTP file download and we expect an even bandwidth distribution between the video streaming and the file download. The forth contribution of this thesis is an evaluation of the behavior of the algorithms in a two-client and HTTP cross traffic scenario.
The remainder of this thesis is structured as follows. Chapter II gives a brief introduction to video coding with H.264, the HTTP adaptive streaming standard MPEG-DASH, the investigated adaptation algorithms and metrics of Quality of Experience (QoE) for video streaming. Chapter III presents the methodology and results of the subjective studies conducted in the course of this thesis to identify the QoE influence factors of adaptive video streaming. In Chapter IV, we introduce the proposed adaptation algorithm and the methodology of the performance evaluation. Chapter V highlights the results of the performance evaluation and compares the investigated adaptation algorithms. Section VI summarizes the main findings and gives an outlook towards QoE-centric management of DASH with SVC.
Burn severity was measured within the Mediterranean sclerophyll forests of south-west Western Australia (WA) using remote sensing data from the Moderate Resolution Imaging Spectroradiometer (MODIS). The region of south-west WA is considered as a high fire prone landscape and is managed by the state government’s Department of Conservation and Land Management (CALM). Prescribed fuel reduction burning is used as a management tool in this region. The measurement of burn severity with remote sensing data focused on monitoring the success and impact of prescribed burning and wildfire in this environment. The high temporal resolution of MODIS with twice daily overpasses in this area was considered highly favourable, as opportunities for prescribed burning are temporally limited by climatic conditions. The Normalised Burn Ratio (NBR) was investigated to measure burn severity in the forested area of south-west WA. This index has its heritage based on data from the Landsat TM/ETM+ sensors (Key and Benson, 1999 [1],[2]) and was transferred from Landsat to MODIS data. The measurement principally addresses the biomass consumption due to fire, whereas the change detected between the pre-fire image and the post-fire image is quantified by the ÄNBR. The NBR and the Normalised Difference Vegetation Index (NDVI) have been applied to MODIS and Landsat TM/ETM+ data. The spectral properties and the index values of the remote sensing data have been analysed within different burnt areas. The influence of atmospheric and BRDF effects on MODIS data has been investigated by comparing uncorrected top of atmosphere reflectance and atmospheric and BRDF corrected reflectance. The definition of burn severity classes has been established in a field trip to the study area. However, heterogeneous fire behaviour and patchy distribution of different vegetation structure made field classification difficult. Ground truth data has been collected in two different types of vegetation structure present in the burnt area. The burn severity measurement of high resolution Landsat data was assessed based on ground truth data. However, field data was not sufficient for rigorous validation of remote sensing data. The NBR index images of both sensors have been calibrated based on training areas in the high resolution Landsat image. The burn severity classifications of both sensors are comparable, which demonstrates the feasibility of a burn severity measurement using moderate spatial resolution 250m MODIS data. The normalisation through index calculation reduced atmospheric and BRDF effects, and thus MODIS top of at-mosphere data has been considered suitable for the burn severity measurement. The NBR could not be uniformly applied, as different structures of vegetation influenced the range of index values. Furthermore, the index was sensitive to variability in moisture content. However, the study concluded that the NBR on MODIS data is a useful measure of burn severity in the forested area of south-west WA.
Bees are subject to permanent threat from predators such as ants. Their nests with large quantities of brood, pollen and honey represent lucrative targets for attacks whereas foragers have to face rivalry at food sources. This thesis focused on the role of stingless bees as third party interactor on ant-aphid-associations as well as on the predatory potential represented by ants and defense mechanisms against this threat. Regular observations of an aphid infested Podocarpus for approaching stingless bees yielded no results. Another aim of this thesis was the observation of foraging habits of four native and one introduced ant species for assessment of their predatory potential to stingless bees. All species turned out to be dietary balanced generalists with one mostly carnivorous species and four species predominantly collecting nectar roughly according to optimal foraging theory. Two of the species monitored, Rhytidoponera metallica and Iridomyrmex rufoniger were considered potential nest robbers. As the name implies, stingless bees lack the powerful weapon of their distant relatives; hence they specialized on other defense strategies. Resin is an important, multipurpose resource for stingless bees that is used as material for nest construction, antibiotic and for defensive means. For the latter purpose highly viscous resin is either directly used to stick down aggressors or its terpenic compounds are included in the bees cuticular surface. In a feeding choice experiment, three ant species were confronted with the choice between two native bee species - Tetragonula carbonaria and Austroplebeia australis - with different cuticular profiles and resin collection habits. Two of the ant species, especially the introduced Tetramorium bicarinatum did not show any preferences. The carnivorous R. metallica predominantly took the less resinous A. australis as prey. The reluctance towards T. carbonaria disappeared when the resinous compounds on its cuticle had been washed off with hexane. To test whether the repulsive reactions were related to the stickiness of the resinous surface or to chemical substances, hexane extracts of bees’ cuticles, propolis and three natural tree resins were prepared. In the following assay responses of ants towards extract treated surfaces were observed. Except for one of the resin extracts, all tested substances had repellent effects to the ants. Efficacy varied with the type of extract and species. Especially to the introduced T. bicarinatum the cuticular extract had no effect. GCMS-analyses showed that some of the resinous compounds were also found in the cuticular profile of T. carbonaria which featured reasonable analogies to the resin of Corymbia torelliana that is highly attractive for stingless bees. The results showed that repellent effects were only partially related to the sticky quality of resin but were rather caused by chemical substances, presumably sesqui- and diterpenes. Despite its efficacy this defense strategy only provides short time repellent effects sufficient for escape and warning of nest mates to initiate further preventive measures.
Autonomous mobile robots operating in unknown terrain have to guide
their drive decisions through local perception. Local mapping and
traversability analysis is essential for safe rover operation and low level
locomotion. This thesis deals with the challenge of building a local,
robot centric map from ultra short baseline stereo imagery for height
and traversability estimation.
Several grid-based, incremental mapping algorithms are compared and
evaluated in a multi size, multi resolution framework. A new, covariance
based mapping update is introduced, which is capable of detecting sub-
cellsize obstacles and abstracts the terrain of one cell as a first order
surface.
The presented mapping setup is capable of producing reliable ter-
rain and traversability estimates under the conditions expected for the
Cooperative Autonomous Distributed Robotic Exploreration (CADRE)
mission.
Algorithmic- and software architecture design targets high reliability
and efficiency for meeting the tight constraints implied by CADRE’s
small on-board embedded CPU.
Extensive evaluations are conducted to find possible edge-case scenar-
ios in the operating envelope of the map and to confirm performance
parameters. The research in this thesis targets the CADRE mission, but
is applicable to any form of mobile robotics which require height- and
traversability mapping.
The present study focuses on Rulin waishi 儒林外史 (The Unofficial History of the Scholars), a well-known Qing novel, from the perspective of gender. It attempts to contribute to the discussion about Chinese masculinity by identifying the representation of the scholars' masculinity in Rulin waishi and offer a better understanding of the novel's position regarding femininity and women.
This project shows that the novel nevertheless reflects rather than challenges gender ideologies of its time. The ideal manhood showed in the novel comprises real virtues and authentic learning. It goes against the traditional, orthodox Confucian masculinity which advocates that officialdom is the glorious path to fulfill a learned man's masculinity. It is mainly due to Wu Jingzi's own failure in the civil service examinations and official careers. Regarding the relation of masculinity and sexuality, the novel reveals that a masculine man is not tempted by female charm but can enjoy a harmonious and companionate marriage. Besides, scholars show great anxiety about their masculinity since they are in a marginal position in society. Their manliness is challenged by officials, merchants, and even commoners, as well as their colleagues.
Through a careful examination of stories of Pinniang, Miss Lu, and Mrs. Wang, it reveals that the novel holds a conventional opinion on women although it criticizes widow suicide and shows an egalitarian husband-wife relationship. It praises Confucian womanly virtues, such as following and serving the husband, managing the household, and keeping chastity. Female sexuality is blamed as an evil temptation to lead men to go astray. Women’s learning gains legitimacy when serving to fulfill domestic responsibilities. It carries the Confucian message that men should take the lead and maintain order in the household and reinforces the rightful patriarchy.
In a word, rather than go ahead of its time, Rulin waishi holds a conservative attitude towards gender issues.
The present study discusses money and conflicts and anxiety over money in late Ming vernacular stories and contextualizes these stories in the contemporary society of economic prosperity and rapid changes. The high monetization and extensive use of silver and copper cash as currency brought both wealth and conflicts in various aspects of society. Eleven vernacular stories from several collections are adopted as source materials for the close examination, including Jingshi tongyan (Stories to Caution the World, 1624) and Xingshi hengyan (Stories to Awaken the World, 1627) by Feng Menglong (1574-1646) and the two Pai’an jingqi (Slapping the Table in Amazement, 1628 and 1632) collections by Ling Mengchu (1580-1644), etc. The analysis then focuses on the relationship between money and four topics, the late Ming context, social relations, gender ideals, and religion. Multiple voices and various viewpoints in these narratives show human beings’ struggles in taming and dominating money, the increasingly familiar and essential object in everyday life. Generally, when people cannot control money properly, there is a fear of its detrimental power to humans and social relations within and beyond families. On the contrary, characters, who are able to control money, are praised.