Video transmission through a shallow underwater acoustical channel

In order to achieve video transmission through a shallow underwater acoustical channel, an image compression algorithm is required which is both bandwidth efficient and error tolerant. Low signal-to-noise ratios and strong rapidly varying multi-path interference mean that relatively high error rates can be expected. Consequently, conventional image compression techniques are unsuitable as they are almost always highly sensitive to bit errors, so much so that a single error can completely destroy a whole image. The proposed image compression and protection algorithm combines several subsystems. A modified SPIHT algorithm compresses the image. A conventional rate compatible punctured convolutional code provides the bulk of the error protection whilst an adaptation of an error resilient entropy coding technique provides additional protection against occasional remaining errors.