Encoder for encoding data defining coded positions representing a trajectory of an object based on spatio-temporal data representing a time series of positions of the object. The encoder comprises a deriver, a selector, and a coder. The deriver is configured for deriving a plurality of components of a motion vector difference between a last coded motion coded vector pointing from a second last coded position to a last coded position and a current motion vector pointing from the last coded position to a current position of the object. The selector is configured for selecting a component of the plurality of components of the motion vector difference. The coder is configured for coding the current motion vector based on the selected component of the motion vector difference to obtain a current coded motion vector, thereby defining a current coded position.