**5. Additional improvement techniques**

Despite that it outperforms the conventional motion estimation methods, our proposed DWT based method still having some problems. As we have mentioned before, the DWT representation suffers from the problem of aliasing and the fact that it is a shift variant transformation. Moreover, the block based motion estimation causes the blocking effect which gives a discontinuity in the block boundaries of the predicted image. That is what drives us to develop some additional techniques to overcome these problems.

These techniques make the motion estimation process more precise and more rapid by detecting the moving zones and limiting the estimation operation to it, adding a sub-pixel precision to the motion vector computing, applying the motion estimation to a shifting variants of the original image aiming to make the estimation a shift invariant operation, overlapping the frame blocks to correct the motion vector by their neighbouring vectors and finally, refining the prediction by changing the block size and re-predicting the blocks which are falsely predicted. In this section we will describe these techniques as well as the causes that conduct us to implement them.
