Browsing by Author "Guo, Xiaolei"

Now showing 1 - 2 of 2

Open Access
Droplet feature extraction from images obtained from spray drying process
(2012) Guo, Xiaolei
The Studienarbeit "Droplet feature extraction from images obtained from spray drying process" described a method that can extracting a droplet from an image. Features like axes-lengths and area of the droplet could also be calculated. At first several image pre-processing methods such as sobel filter and morphological operations have been discussed. After that different ways for compositing these methods have been introduced and the effect of these ways has been discussed. The function for calculating the droplet's properties has also been introduced. The result has been evaluated, then the algorithm was set. Edge detection, morphological operation and features extraction were the necessary steps in the algorithm. The algorithm was tested on several regular shapes, and it worked well. Ways to improve this algorithm has also been discussed. The amount of memory usage and processing time have been analyzed. Two functions in the first algorithm required reading whole image beforehand, thus two alternative methods have been introduced. The result was analyzed and evaluated. The advantage and disadvantage of this modified algorithm were discussed.
Open Access
Evaluation of a methodology for migration of the database layer to the cloud based on an eScience case study
(2013) Guo, Xiaolei
In the last years Cloud computing has become popular among IT organizations. Applications and data can be designed to be run on the Cloud, and can also be partially or completely migrated to the Cloud. The application can be separated into 3 layers: the presentation layer, the business logic layer and the data layer. The data layer can further be separated into 2 sub-layers: the data access layer and the database layer. Migration tasks are applied on these layers. In order to benefit from the Cloud technology, migrating the data layer or the business logic layer or both of them into the Cloud is required. For the migration there are different migrating scenarios and patterns. The migrating methodologies are different cross different providers. During the migration procedure different problems may occur. The types, causes, and solutions of these problems may also be different. Identifying the properties of these problems and comparing the migrating methodologies are important to the Cloud users. In this thesis we migrate both the data layer and business logic layer of a Scientific Workflow System to the Cloud. The migration scenario "Cloud Bursting" is considered. Several quantitative and qualitative metrics chosen from an ISO/ICE standard are introduced and are used to define the properties of the migrating methodologies. Two migrating methodologies are applied, one is the methodology of Bachmann, the other one is the methodology of Amazon. The required refactoring of the application architecture is investigated and the needed modifications are recorded and explained. Data of the metrics of the two migrating methodologies are collected, compared, and discussed.