| 摘要 |
A coding rate upper limit value storing unit and a coding rate lower limit value storing unit store a pair of a preset quantization accuracy and an upper limit value and a lower limit value on a coding rate corresponding to the preset quantization accuracy respectively. A coding rate calculating unit calculates a coding rate based on a code amount generated for a frame which was coded the last time. A first comparator and a second comparator compare a coding rate calculated by the coding rate calculating unit with an upper limit value and a lower limit value on a coding rate corresponding to a quantization accuracy with which a frame was coded the last time respectively. If the coding rate from the coding rate calculating unit is greater than the upper limit value or is smaller than the lower limit value, a quantization accuracy which is more accurate or less accurate than the quantization accuracy with which a frame was coded the last time is selected from quantization accuracies respectively. If a quantization accuracy is in a range between the lower limit value and the upper limit value, the range including the lower limit value and the upper limit value, the quantization accuracy with which a frame was coded the last time is selected. A quantization accuracy directing unit outputs the thus selected quantization accuracy to a quantizing unit, and the quantizing unit performs coding with the received quantization accuracy. <IMAGE> |
