Scaling interlaced images for encoding and decoding of digital video signals of multiple scanning standards

摘要

In a method of scaling interlaced video signals suitable for layered coding, an interpolation process is restricted within a frame where the interpolation of one field in a frame is derived from the same field and its opposite field in the same frame. The first field is converted to a first progressive image such that an interpolated line in the first field is formed by a weighted average of two or more adjacent lines in the first field plus a weighted average of two or more adjacent lines in the second field where one of these lines is at the same position as the interpolated line in the first field. Similarly, the second field is converted to a second progressive image such that an interpolated line in the second field is formed by a weighted average of two or more adjacent lines in the second field plus a weighted average of two or more adjacent lines in the first field where one of the lines in the first field is at the same position as the interpolated line in the second field. The first and second progressive images are then resampled vertically according to a predetermined resampling ratio m/n, where m and n are integers with n indicating the number of output samples out of m input samples. The sum of the weights for filtering the lines in the opposite field is zero.