| 摘要 |
<p>Disclosed is an image encoding device that aims to increase encoding efficiency and to increase mode estimation accuracy using the characteristics of transformation coefficients. In response to an assessment value calculated on the basis of the characteristics of transformation coefficients, the range of estimated modes and selectable modes is limited, a mode that minimizes encoding cost is selected from among predetermined modes within the limited range of modes, when the number of the aforementioned limited modes is one, a flag for mode identification is not encoded, and when there are a plurality of modes, the aforementioned selected mode is encoded. As a result, the amount of encoding cost of the mode-identifying flag is decreased and encoding can be performed with a high encoding efficiency, mode estimation accuracy is increased, and furthermore, the amount of flag encoding is decreased and encoding efficiency is increased.</p> |
