Diploma work: Monthly report 3/97

Review March 1997

In the last month I focussed my work on the generation of the delta images and the developement of a GUI with X11 and Motif. In detail I achieved the following results:

• I could eliminate nearly all the shadows around the pen tip during the capturing with the camera with a different illumination of the room. I replaced the spotlights I used to illuminate the sheet of paper with a general illumination of the whole room and got rid of the small and dark shadows.
• On the other hand I implemented the proposal of Mr. Ha to eliminate those shadows during the generation of the delta images. Because those shadows are moving from frame to frame to a different position there are parts on the images that are suddenly vanishing. Because ink doesn’t show this behavior I can delete those vanishing spots from the preceding delta frame.
• I was able to tune the speed of the filters. The processing of a single frame lasts now about 5 seconds on a fast machine (SUN Ultra or Intel Pentium).
• I implemented a new GUI using Motif. Inside of this tool I’m able to quickly test my filters and the delta images. The speed of the displaying is quick and a real-time zoom feature is available.

Preview April 1997

Unfortunately I couldn’t finish my prototype because the implementation of the GUI took much more time than I expected in the planning. But I’m sure to finish the prototyp during April and to get results in a special file format (still to be defined) by end of April. But there are already enough informations and results to present in my talk at the 22.4.97.

My goals for April 1997 are:

• Finishing the basics of the GUI. I have to concentrate mostly on speed and some memory leaks that appered in the X-Server.
• Tuning and perhaps some adoptions to the generation of delta images.
• At the moment I process color images but I discovered that I only use the greyscale information of the images inside of my filters. I plan to analyse if greyscale infos are sufficient for the image processings. If this is really the case I will switch so greyscale images to speed up the whole processing.
• Writing output to my own file format, which still has to be specified.
• Prepare and give a talk on the occasion of the FKI Seminar at the 22.4.97.

zurück zur Uebersicht der Diplomarbeit

Last modified 23.12.1999 00:18:51 by Thomas von Siebenthal