Image encoding apparatus, image encoding method, image decoding apparatus, and image decoding method

摘要

The present invention provides an image encoding/decoding technique that is capable of achieving the higher compression efficiency. An image encoding method comprises: an intra prediction step which performs intra prediction on a block basis to generate a predicted image; a subtraction step which calculates the difference in prediction between the predicted image generated by the intra prediction step and an original image; a frequency conversion step which performs frequency conversion processing for the difference in prediction; a quantization step which subjects the output of the frequency conversion step to quantization processing; and a variable-length encoding step which subjects the output of the quantization step to variable-length encoding processing; wherein the intra prediction encoding step predicts a target pixel to be encoded by use of pixel values of two reference pixels between which the target pixel to be encoded is located.

主权项

1. An image decoding method using intra prediction, the method comprising:
a first step for performing inverse variable length coding processing; a second step for inverse quantization processing the data that has been processed in the first step; a third step for decoding prediction residuals by inverse frequency conversion processing the data obtained by inverse quantization in the second step; and a fourth step for obtaining a decoded image as a cluster of decode target pixels by performing a decoding process using the prediction residuals and a decoded pixel, wherein, processes in the fourth step include:
a first process for generating a first reference pixel using a pixel in a decoded block adjacently located on left side or upper side of a decode target block, and for generating a prediction pixel in a prediction mode, in which a calculation process is performed using the first reference pixel and a decoded second reference pixel, the second reference pixel being adjacently located on left side or upper side of the decode target block and located at a different side from the side of the first reference pixel; anda second process for generating a prediction pixel using a decoded pixel adjacently located to the decode target block in a different prediction mode from the prediction mode in the first process, and wherein a decode target pixel is obtained by using the prediction pixel generated by the first process or the second process and the prediction residuals.