Input video frame is obtained from Smartphone camera. It is subjected to various video processing
operations like Canny Edge Detection, binarization, successive frame subtraction for observing the
differences between the two successive frames by masking the difference, dilation, etc. All these algorithms
are implemented using OpenCV libraries.
In this section, we present the experimental results to demonstrate the practical implementation of the
system. An AAKASH tablet with Android version 4.2.2 was used as a test bench. AAKASH tablet is
designed by IIT-Bombay as a part of their research. This tablet is in the process of development, and various
android applications are implemented on it, to check the efficiency of the tablet in response to the application.
The experiments were also implemented on Begleboard-xM. The watermark used in our experiment is
32pixels x 32pixels in dimensions as shown in Fig 4(c). The frame obtained on the tablet for video processing,
and invisible watermarking would occupy the whole display with dimensions 640 x 480. The images of
various video processing and watermarking operations which were implemented on AAKASH and
Beagleboard-xM are shown. Fig. 2(a) and Fig. 2(b) show the results for canny edge detection and successive
frame subtraction on real time video frames captured from AAKASH tablet. Fig. 2(c) shows invisibly
watermarked frame. Fig. 2(d) shows dilation and frame differencing. Fig. 3(a), Fig. 3(b), Fig. 3(c) and Fig.
3(d) show visible watermarking process on AAKASH and various image processing techniques applied to
watermarked image. Finally, Fig. 4(a) and Fig. 4(b) show the overview of the visible watermarking
algorithms implemented by us on Beagleboard-xM.