Integrating intensity and texture differences for robust change detection
IEEE Transactions on Image Processing
| Hi-index | 0.01 |
Most distance learning or teaching systems use just regular television equipment, so the receiver is limited to passively watching the incoming low-resolution video stream. We have created a high-resolution image transfer system that creates a virtual copy of the sender's environment, captured by multiple high-resolution still cameras, within the receiver's computer. The receiver can peruse the virtual copy and zoom in to read what the lecturer has written on the board whenever desired. This paper shows how to reduce the transmission bandwidth required; the system periodically updates the images held in the receiver's computer when necessary by sending only the images in which a significant change is detected and that show the lecturer's most recent position. A simple method of detecting the lecturer's position accurately is introduced and we discuss the impact of sound localization on service acceptance.