| 摘要 |
A motion image data compression coding apparatus and an image data compression coding method capable of smoothing compressed data among frames are disclosed. A data amount of compressed data is obtained by a hybrid coding of a combination of an orthogonal transform such as a discrete cosine transform, a quantization and a variable length coding of motion image data of one frame. A quantization step size is expressed by a product of a scaling factor f and quantization matrix factors q. The scaling factor f is determined from a code amount produced up to the preceding frame, a target code amount and an AC level of the transform factors of the present frame. Prior to the quantization and the variable length coding, a code amount actually resulted by the same coding is detected, and by considering the actual code amount, the scaling factor f for the last quantization is corrected. <IMAGE> |
