Intelligent Video Conferencing


Overall, the Intelligent Video Conferencing project has progressed well, though certainly not as well as hoped. The video conferencing portion of the project was initially held up waiting for documentation from Connectix as to how to control the QuickCam, and then by problems introducing by the protective features of the original test platform, Microsoft's Windows. These problems, however, were not the only nor the most significant causes of delay in the project.

The main obstacle soon became the language chosen for the implementation of the video conferencing system--Sun's Java. First, because the language itself is still very new and under constant development, there are few good resources, be they printed, online or simply people. Therefore, for a large part, the design team was on its own to find the solution to each language-related problem that came up. In addition, while the language is very powerful in many ways, specifically it's graphical tools, Internetworking capabilities and cross-platform nature, it is currently used almost exclusively for adding multimedia effects to otherwise static pages on the World Wide Web. As a result, a majority of the books, magazines and online resources about the language focus on the use of applets on the Internet. Therefore, they were of little or no help in the development of an application.

However, despite the setbacks caused by the use of the Java language, the design team believes it has a tremendous potential to be used for platform-independent, graphically based networking software. Once you understand a given aspect of the language, it is an easy-to-use, powerful and flexible language, which has many features to make object-oriented programming easy.

In the Image Following group, testing had been done with the transducers as transmitters and receivers. As documented earlier, in this one-on-one setup, they performed as expected. However, when arranged in the array as originally designed and connected to the MC68HC11 EVB, the response time of the overall system was not sufficient to allow for the use of the reception time delay to be the basis for the motor controls. This prompted a redesign of the system to base movements on the information as to which of the two transducers (on a given axis) heard from the transmitter first.

After redesigning the system to use this methodology, the group also had the tasks of making engineering changes to the design to account for unforeseen obstacles. Examples include insufficient impedance and triggering which resulted in the wrong wave form being outputted. Other engineering refinements made to the system included changes to the motor stepping algorithm, using full steps instead of half steps, and implementation of software stops to prevent moving the camera to far in any direction. Finally, during the last phase of the testing procedure, a problem with one the motors led to another discovery--in real-world use, horizontal movement is nearly always sufficient, and vertical movement not needed, for a person located a reasonable distance from the camera and not constantly sitting then standing then sitting, etc. While the vertical movement is still a good feature for maximum flexibility, it can be removed from the system for a less expensive version of the product.

The next step for the project as a whole would be to implement the audio subsystem and enhance the transmitter pack from one that requires four 9V batteries to one that is more portable. Other possibilities for future versions include ports to other operating systems, integration with a different video system to allow for faster refresh rates, better image quality, etc., selection of different transmitters and receivers to allow for reimplementation of the time-delay version of the tracking software, and upgrading the tracking system to support tracking the user by alternative means, such as the user's voice. It is noted that some of these possibilities would, at least currently, price the product out of the range affordable to individuals, families and small businesses, one of the original goals of the project. However, as with all technologies, what starts as expensive and only for the richest or largest soon becomes second-nature for everybody regardless of social, economic or political status.