Abstract: The present invention relates to a method of image processing or coding comprising dividing image data into a plurality of blocks of image data wherein each block is defined by a first positional index and a second positional index. For each block of image data, firstly the block of image data is processed at step 402 to generate at least one parameter value required to process one or more adjacent blocks of image data, secondly a position of a memory element in a parameter memory is determined by combining the first positional index of the block of image data with the second positional index of the block of image data, and thirdly the generated at least one parameter value is stored at step 406 in a memory element in the position by overwriting at least one parameter value generated from a previously processed block of image data which has been previously stored at the determined position in the parameter memory.

Abstract: The present invention relates to a method of image processing or coding comprising dividing image data into a plurality of blocks of image data wherein each block is defined by a first positional index and a second positional index. For each block of image data, firstly the block of image data is processed at step 402 to generate at least one parameter value required to process one or more adjacent blocks of image data, secondly a position of a memory element in a parameter memory is determined by combining the first positional index of the block of image data with the second positional index of the block of image data, and thirdly the generated at least one parameter value is stored at step 406 in a memory element in the position by overwriting at least one parameter value generated from a previously processed block of image data which has been previously stored at the determined position in the parameter memory.

Abstract: Video signals are subjected to a filtering process in which the signals are multiplied by a first variable gain coefficient and a delayed sample of the signals is derived by a delay circuit and the delayed sample is multiplied by a second variable gain coefficient. The two multiplied signals are added together. The magnitudes of the first and second gain coefficients are varied in dependence upon a comparison between the input video signal relative to the delayed sample from the delay circuit and in an opposite sense to one another. By a suitable choice of the gain coefficients, the filtering process can be used to preserve the sharpness of large transitions in picture information and the impulse response of the filter can be readily modified by re-calculating the gain coefficient. The filtering process may be used in a cascade of stages to achieve filtering in vertical, horizontal and temporal domains.