Quantization error in ternary delta modulation schemes [image bandwidth compression]

Image bandwidth compression using differential encoding is reviewed. The ternary delta modulation scheme, being simple to implement and efficient in bit-rate reduction, has been used. A method of pixel differences has been used to overcome the sensitivity problems and two prediction algorithms are shown. The quantizer used is based on the constant area quantization (CAQ) proposed by J.J. Pearson (1975), originally used for imaging by remote pilotless vehicles. The CAQ scheme used one-bit prediction and is included for comparison purposes. The relative positions of the basis pixels are shown. The quantization error of the scheme which arises from the use of the ternary quantizer imposes fundamental limitations on the quality of the reconstruction of images. Theoretical modeling as well as actual measurements of the mean square error are included