摘要 |
In order to obtain a reduction of the bit rate when digitizing a color video signal using a picture transform, the color video signal is sampled with a frequency which is twice the color auxiliary carrier frequency. The video signal samples thus obtained are assembled into two-dimensional subpictures each comprising P consecutive video signal samples of each time Q consecutive line signals. Each subpicture can be considered as the sum of a series of superposed basic pictures, each having its own coefficient (which is a weighting factor representing the contribution of the basic picture in the subpicture). Each coefficient is encoded with a number of bits assigned thereto. The basic pictures are assembled from picture squares which are either of a first type (completely white) or a second type (completely black). The Hadamard matrix transform determines the basic picture patterns. Each of the two color information signals u(t) and v(t) in the color video signal now contributes to a subpicture in such a way that this contribution can be fully described by one basic picture. Only the coefficients associated with these basic pictures need not be encoded accurately.
|