| Summary: | 碩士 === 崑山科技大學 === 資訊管理研究所 === 93 === Without the limitation of time and space, distance learning has become a popular learning method due to the growth of the information technology and Internet. At present, the contents in most distance learning systems are expressed with characters and formulas which are not suitable for the courses that have high relevance with images, such as anatomy, histology, image diagnostics, geography and astronomy. In order to provide a distance learning platform for those courses, an image and marking based e-learning system has been implemented in this research. An anatomy and a histology courses are used to be examples of this platform.
This research uses Java language to build user interface. We provide PHP program and Java modules based on JDBC-ODBC, and JDBC to connect different databases. The client program of this e-learning system can be run as Java applet or application mode. Each image in this system is constructed by multi-images with different resolutions. The use of multi-resolution images to construct images in this system not only reduces times of transmitting image through network, but also displays very high-resolution images on memory-limited computers. By adopting the contain method in polygon class of Java language to replace the geometric retrieval functions of the PostgreSQL database, our e-learning system can achieve database independence? And by incorporating minimum bounding rectangle (MBR) of R-tree to limit the query results from database, the delay time of geometric operations can be reduced.
As the result of the experiment, the response time of our e-learning is 8 times faster than those of without MBR, but slightly slower than geometric retrieval functions performed by PostgreSQL database. However, the delay time caused by Java geometric operation is acceptable in that the restriction of the selectiveness of the database can be removed.
|
|---|
