To provide an image encoding method high in efficiency and simple in configuration by converting an image extending one part of image data compressed by a 1st compression-encoding means to the resolution of 2nd image having a different resolution, and outputting a 3rd digital image.

A switch 19 is provided at a high-resolution compression-encoding means 1. When encoding a high-resolution intra-frame encoding frame (I frame), the switch 19 is changed over, a difference between the result, for which the resolution of low-resolution image is converted by a resolution converting circuit 3, and a source image is calculated and the frame is encoded later. Concerning the other frame, the low-resolution image is not used for prediction. Since pixel density is considerably different between high-resolution and low-resolution images in general, efficiency is improved in the prediction from the image of timewisely different frames with the same resolution rather than in the prediction between the images of different resolutions. Therefore, since I frame is predicted from the low-resolution image because of no execution of timewise prediction, efficiency is improved rather than the encoding of I frame itself.