摘要 |
An encoding apparatus for encoding each of target pixels composing one screen, comprising a predictor for generating a prediction value by predicting a target pixel value representing a target pixel on the basis of pixels around the target pixel, and a difference generator for generating a difference between the prediction value generated by the predictor and the target pixel value. A quantizer generates a quantized value by quantizing the difference generated by the difference generator, an entropy encoder generates encoded data by entropy-encoding the quantized value generated by the quantizer, and a quantization controller selects one of k quantization methods, which is to be used by the quantizer for every part image composing the one screen. Each of the k quantization methods is a quantization method which makes an error between the difference generated by the difference generator and a quantized value obtained by quantizing the difference using the quantizer fall within a range of -Nm to +Nm, wherein m=1 to k.
|