Video encoding apparatus and method

A video encoding apparatus comprising:

• a feature amount computation unit configured to divide an input video signal into a plurality of scenes each comprising at least one temporally-continuous frame, and compute a statistical feature amount for each scene;
• an encoding parameter generator to generate an encoding parameter for each scene based on the statistical feature amount computed by the feature amount computation unit;
• a number-of-generated-bits prediction unit configured to predict the number of bits to be generated when the input video signal is encoded using the encoding parameter generated by the encoding parameter generator;
• an encoding parameter correcting unit configured to correct the encoding parameter based on a result of the prediction of the number of generated bits which is obtained by the number-of-generated-bits prediction unit;
• an encoder to encode the input video signal using the corrected encoding parameter and generate an encoded bit stream;
• and
• an output unit configured to output the encoded bit stream generated by the encoder as an encoded output, wherein the feature amount computation unit includes a classification unit configured to classify the plurality of scenes into a plurality of scene types, based on the statistical feature amount relating to a motion vector, and the encoding parameter generator includes a setting unit configured to correct a frame rate as well as a quantization step by setting weight parameters to the frame rate and the quantization step size, respectively, for each scene according to the scene types.

A video encoding method comprising:

• dividing an input video signal into a plurality of scenes each comprising at least one temporally-continuous frame;
• computing a statistical feature amount for each scene;
• generating an encoding parameter for each scene based on the statistical feature amount computed by the feature amount computing step;
• predicting the number of bits to be generated when the input video signal is encoded using the encoding parameter generated by the encoding parameter generating step;
• correcting the encoding parameter based on a result of the prediction of the number of generated bits which is obtained by the number-of-generated-bits predicting step;
• and
• encoding the input video signal using the corrected encoding parameter to generate an encoded bit stream,
• wherein the encoding parameter generating step includes correcting a frame rate as well as a quantization step size by setting weight parameters to the frame rate and the quantization step size, respectively, for macro blocks of frames to be encoded for each scene on the bases of the statistical feature amount relating to a distribution of luminance for each macro block.

A video encoding method comprising:

• dividing an input video signal into a plurality of scenes each comprising at least one temporally-continuous frame;
• computing a statistical feature amount for each scene;
• generating an encoding parameter for each scene based on the statistical feature amount computed by the feature amount computing step;
• predicting the number of bits to be generated when the input video signal is encoded using the encoding parameter generated by the encoding parameter generating step;
• correcting the encoding parameter based on a result of the prediction of the number of generated bits which is obtained by the number-of-generated-bits predicting step;
• and
• encoding the input video signal using the corrected encoding parameter to generate an encoded bit stream, wherein the feature amount computing step includes classifying the plurality of scenes into a plurality of scene types, based on the statistical feature amount relating to a motion vector, and the encoding parameter generating step includes correcting a frame rate as well as a quantization step size by setting weight parameters to the frame rate and the quantization step size, respectively, for each scene according to the scene types.

A recording medium having a computer program recorded therein for encoding an input video signal, the computer program comprising:

• instruction means for instructing the computer to divide an input video signal into a plurality of scenes each comprising at least one temporally-continuous frame and compute a statistical feature amount for each scene;
• instruction means for instructing the computer to generate an encoding parameter for each scene based on the statistical feature amount;
• instruction means for instructing the computer to predict the number of bits generated when the input video signal is encoded using the encoding parameter;
• instruction means for instructing the computer to correct the encoding parameter based on a result of the prediction of the number of generated bits;
• and
• instruction means for instructing the computer to encode the input video signal using the corrected encoding parameter and generate an encoded bit stream,
• wherein the means for instructing the computer to generate the encoding parameter includes means for instructing the computer to correct a frame rate as well as a quantization step size by setting weight parameters to the frame rate and the quantization step size, respectively, for macro blocks of frames to be encoded for each scene on the bases of the statistical feature amount relating to a distribution of luminance for each macro block.