A method and apparatus for dynamic contrast improvement in video pictures is proposed. There is a strong demand for picture contrast and high quality video. The picture contrast can't be increased by simply increasing the video signal amplitude because exceeding the display dynamic range in some video scenes causes unwanted disturbing effects. A basic solution for dynamic contrast improvement includes that the video pictures are analysed framewise in real time and the parameters of a transfer function (11) for contrast improvement are adjusted depending on the analysis results for the best subjective picture quality. Within this basic solution it is proposed to perform a step of adaptive signal splitting in which a noise reduced luminance signal (Y_CORED) is generated which then is transformed by means of the transfer function. This improves the picture quality by reducing the noise influence