A television signal for a television transmission system conveys a motion information signal in addition to the vision signal. At the transmission side each field of a picture is subdivided into a plurality of regions and up to a given number of the most dominant motion vectors appropriate to a region are generated. These motion vectors are transmitted as the motion information signal and in a receiver are separated from the television signal in a decoder (12). The spatially offset pixels are identified in a first spatial offset circuit (14) from a vision signal delayed by a picture memory (16) and compared in a picture information comparison circuit (17) with pixels in the corresponding field of the current picture to identify the motion vector indicating the pixel of corresponding magnitude. This is applied to an input of a second spatial offset circuit (19) a second input of which receives the delayed vision signal to produce motion compensated pixels which together with pixels from the current vision signal are up-converted and interpolated in a unit (18) to produce an up-converted vision signal.