Prediction apparatus, encoding apparatus, inverse prediction apparatus, decoding apparatus, and computing apparatus
Abstract:
A candidate prediction block storage unit stores candidate prediction pixel data {“a” to “p”} (a four-by-four area) and neighboring pixel data {“A” to “D”}. A neighboring pixel series-generating unit in receipt of the neighboring pixel data {“A” to “D”} from the candidate prediction block storage unit feeds a first neighboring pixel series {“A”, “B”, “C”, and so on} and a second neighboring pixel series {“B”, “C”, “D”, and so on} into first and second lines, respectively. A predictive processing unit in receipt of the candidate prediction pixel data {“a” to “p”} from the candidate prediction block storage unit multiplies the candidate prediction pixel data by two, thereby providing multiplied data. The predictive processing unit subtracts the first neighboring pixel series from the multiplied data, thereby providing subtracted data. The predictive processing unit subtracts the second neighboring pixel series from the subtracted data, thereby providing further subtracted data. The predictive processing unit divides the further subtracted data by two, thereby providing divided data. A predicted block storage unit stores the divided data as predicted pixel data {“a#”, “b#”, and so on}. As a result, steps of storing and generating predictive pixels are eliminated.
Information query
Patent Agency Ranking
0/0