Reducing the block-effect in transform image coding

The authors describe an image decoding scheme which results in reduced block effects. A brief background on image compression is given, followed by a description of the origin of the blocking effect. It is shown that the block effect can be reduced by more accurately quantizing the transform coefficients. Filtering techniques for reducing the block effect tend to blur features crossing boundaries. Overlap addition methods work well but increase the bit rate. A novel transform called the LOT (lapped orthogonal transform), which overlaps adjacent blocks but does not increase the bit rate, is described