Wavelet transform apparatus and method based on prediction-update lifting scheme, and adaptive prediction apparatus and method and recording medium for the same

A wavelet transform apparatus based on a prediction-update lifting scheme, comprising:

• a decomposer for decomposing an input signal into first even polyphase samples and first odd polyphase samples;
• an updater for calculating a first prediction value by predicting even polyphase samples from the first odd polyphase samples and calculating updated even polyphase samples based on the first prediction value and the first even polyphase samples;
• and
• a predictor for calculating a second prediction value by predicting odd polyphase samples from the updated even polyphase samples and calculating residual odd polyphase samples based on the second prediction value and the first odd polyphase samples,
• wherein the predictor decomposes an image into a predetermined number of decomposed areas, selects one or more odd polyphase samples in each of the decomposed areas, selects first prediction directions for minimizing energy in a high band from a plurality of prediction direction candidate groups configured around first updated even polyphase samples influenced when being updated from selected odd polyphase samples, determines a final prediction direction candidate group from selected first prediction directions based on a selection frequency, selects a second prediction direction for minimizing energy in the high band from the final prediction direction candidate group configured around the second updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted, with respect to each of all odd polyphase samples in each decomposed area, and calculates the second prediction value based on selected second prediction direction and updated even polyphase samples corresponding to the second prediction direction.

A wavelet transform method based on a prediction-update lifting scheme, comprising:

• decomposing an input signal into first even polyphase samples and first odd polyphase samples;
• calculating a first prediction value by predicting even polyphase samples from the first odd polyphase samples and calculating updated even polyphase samples based on the first prediction value and the first even polyphase samples;
• and
• calculating a second prediction value by predicting odd polyphase samples from the updated even polyphase samples and calculating residual odd polyphase samples based on the second prediction value and the first odd polyphase samples,
• wherein the step of calculating the second prediction value and the residual odd polyphase samples comprises:
• configuring a plurality of prediction direction candidate groups centered at first updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted;
• selecting a first prediction direction for minimizing energy in a high band from a plurality of configured prediction direction candidate groups;
• and
• calculating the second prediction value based on selected first prediction direction and updated even polyphase samples corresponding to the first prediction direction.

A wavelet transform method based on a prediction-update lifting scheme, comprising:

• decomposing an input signal into first even polyphase samples and first odd polyphase samples;
• calculating a first prediction value by predicting even polyphase samples from the first odd polyphase samples and calculating updated even polyphase samples based on the first prediction value and the first even polyphase samples;
• and
• calculating a second prediction value by predicting odd polyphase samples from the updated even polyphase samples and calculating residual odd polyphase samples based on the second prediction value and the first odd polyphase samples,
• wherein the step of calculating the second prediction value and the residual odd polyphase samples comprises:
• decomposing an image into a predetermined number of decomposed areas;
• selecting one or more odd polyphase samples in each of the decomposed areas, selecting first prediction directions for minimizing energy in a high band from a plurality of prediction direction candidate groups configured around first updated even polyphase samples influenced when being updated from selected odd polyphase samples, and determining a final prediction direction candidate group from selected first prediction directions based on a selection frequency;
• and
• selecting a second prediction direction for minimizing energy in the high band from the final prediction direction candidate group configured around the second updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted, with respect to each of all odd polyphase samples in each decomposed area, and calculating the second prediction value based on selected second prediction direction and updated even polyphase samples corresponding to the second prediction direction.

An adaptive prediction apparatus for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, comprising:

• a prediction sample influence determiner for configuring first updated even polyphase samples influenced when being updated from an odd polyphase sample to be predicted;
• a prediction direction selector for configuring plural prediction direction candidate groups centered on the first updated even polyphase samples and selecting a first prediction direction for minimizing energy in a high band from configured plural prediction direction candidate groups;
• and
• a predictor for calculating a first prediction value based on the first prediction direction and updated even polyphase samples corresponding to the first prediction direction and calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.

An adaptive prediction apparatus for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, comprising:

• a video area divider for decomposing an image into a plurality of decomposed areas;
• a prediction direction candidate group determiner for selecting first prediction directions for minimizing energy in the high band from a plurality of prediction direction candidate groups configured around first updated even polyphase samples influenced when being updated from one or more odd polyphase samples selected in each of the decomposed areas, and determining a final prediction direction candidate group from selected first prediction directions based on a selection frequency;
• and
• a predictor for selecting a second prediction direction for minimizing energy in the high band from the final prediction direction candidate group configured around the second updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted, with respect to each of all odd polyphase samples in each of the decomposed areas, calculating a first prediction value based on selected second prediction direction and updated even polyphase samples corresponding to the second prediction direction, and calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.

An adaptive prediction method for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, comprising:

• configuring first updated even polyphase samples influenced when being updated from an odd polyphase sample to be predicted;
• configuring plural prediction direction candidate groups centered on the first updated even polyphase samples;
• selecting a first prediction direction for minimizing energy in a high band from configured plural prediction direction candidate groups;
• and
• calculating a first prediction value based on the first prediction direction and updated even polyphase samples corresponding to the first prediction direction and calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.

An adaptive prediction method for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, comprising:

• decomposing an image into a plurality of decomposed areas;
• selecting one or more odd polyphase samples in each of the decomposed areas;
• selecting first prediction directions for minimizing energy in a high band from a plurality of prediction direction candidate groups configured around first updated even polyphase samples influenced when being updated from selected one or more odd polyphase samples;
• determining a final prediction direction candidate group from selected first prediction directions based on a selection frequency;
• selecting a second prediction direction for minimizing energy in the high band from the final prediction direction candidate group configured around the second updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted, with respect to each of all odd polyphase samples in each of the decomposed areas;
• calculating a first prediction value based on selected second prediction direction and updated even polyphase samples corresponding to the second prediction direction;
• and
• calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.

A non-transitory computer-readable recording medium recording a program for implementing adaptive prediction functions for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, the adaptive prediction functions comprising:

• configuring first updated even polyphase samples influenced when being updated from an odd polyphase sample to be predicted;
• configuring plural prediction direction candidate groups centered on the first updated even polyphase samples;
• selecting a first prediction direction for minimizing energy in a high band from configured plural prediction direction candidate groups;
• and
• calculating a first prediction value based on the first prediction direction and updated even polyphase samples corresponding to the first prediction direction and calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.

A non-transitory computer-readable recording medium recording a program for implementing adaptive prediction functions for predicting an odd polyphase sample from updated even polyphase samples by a wavelet transform based on a prediction-update lifting scheme, the adaptive prediction functions comprising:

• decomposing an image into a predetermined number of decomposed areas;
• selecting one or more odd polyphase samples in each of the decomposed areas, selecting first prediction directions for minimizing energy in a high band from a plurality of prediction direction candidate groups configured around first updated even polyphase samples influenced when being updated from selected one or more odd polyphase samples, and determining a final prediction direction candidate group from selected first prediction directions based on a selection frequency;
• and
• selecting a second prediction direction for minimizing energy in the high band from the final prediction direction candidate group configured around the second updated even polyphase samples influenced when being updated from a corresponding odd polyphase sample to be predicted, with respect to each of all odd polyphase samples in each of the decomposed areas, calculating a first prediction value based on selected second prediction direction and updated even polyphase samples corresponding to the second prediction direction, and calculating residual odd polyphase samples based on the first prediction value and corresponding odd polyphase samples.