Digitalisasi Lecturing Note

Abstract

The development of technology can make human being to do a lot of things more simple. Teaching method currently has many variations so that the learning process is more alive. Learning by explaining the process which can be seen is very beneficial. Sometimes in a process, hand gestures play an important role in understanding the whole process. The use of webcam can be another alternative to digitize what is written by users. This thesis aims to digitize lecturing notes on paper using a device webcam.

The system includes processes like image retrieval, paper detection, projection mapping, skin detection and removal, and processing results with threshold method. Paper edge detection process uses the method of canny edge detection. In paper edge detection, webcam image becomes image with white color for area of paper. Projection mapping process is used to map the image with trapezoid shape of paper into image with rectangular shape of paper. Skin detection process uses HSV and YCbCr color space. In skin detection process, hand which still appears in image becomes image without hand. The application was created using the C# programming language with Visual C# 2010 as its IDE.

The experiment results indicate that the color, ink thickness, threshold value, height of the webcam, size and distance between scratches, and lighting level will affect the final outcome of the application. In addition, the hand’s position during the writing may cause a shadow on the paper. The best result of all methods which have been tried is the result of the threshold method.