The present invention provides an encoder and a decoder of digital picture data, and the encoder/decoder can realize high precision transform with less quantity of transferred data, when a parameter of the digital picture data is not an integer but has numbers of digits, to which the Affine transformation can be applicable. The encoder/decoder comprises the following elements: (a) picture compression means for encoding an input picture and compressing the data, (b) coordinates transform means for outputting coordinate data which is obtained by decoding the compressed data and transforming the decoded data into a coordinate system, (c) transformation parameter producing means for producing transformation parameters from the coordinates data, (d) predicted picture producing means for producing predicted picture from the input picture by the transformation parameter, and (e) transmission means for transmitting the compressed picture and the coordinates data.
A video indexing method and device for selecting keyframes from each detected scene in the video. The method and device detects fast motion scenes by counting the number of consecutive scene changes detected.
Efficient methods for performing motion compensation are described. The methods are particularly useful during motion compensation based decoding and recoding. In one aspect, the look ahead motion compensation methods obtain motion vector information for a macroblock being reconstructed in advance of motion compensation of the macroblock. The information is then converted into memory transfer instruction; that can be used to perform transferring of reference image data in advance of the motion compensation operations. A reference sub-region needed for motion compensation, as identified by the motion vector, is then obtained in advance of motion compensation of the macroblock.
The present invention provides efficient methods for performing motion compensation. The methods are particularly useful during motion compensation based decoding and recoding. In one aspect, motion compensation in accordance with the present invention improves on-chip memory usage. This is done by creating and storing a reference window in the on-chip memory. The reference window is constructed such that the reference frame portions required for motion compensation of a current frame are contained within the reference window and in the on-chip memory with a high frequency. In another aspect, selective methods are provided to recode compressed video data. The methods selectively apply re-quantization and motion compensation to the residual error between frames of the bitstream based on the amount of motion in the video data.
