Object-based digital image predictive coding transfer method and apparatus

An image predictive coding apparatus for coding a target image by obtaining a predictive image from a reference image including an object image comprising a luminance signal indicating a pixel value and a shape signal indicating whether a pixel is located inside or outside the object image, said apparatus comprising:

• a device operable to divide the target image into a plurality of target subareas;
• a device operable to obtain a substitute pixel value for a pixel located outside the object image, which is determined according to the shape signal, in a subarea of the reference image by using a pixel value of a pixel located inside the object image, which is determined according to the shape signal, in the subarea of the reference image;
• a device operable to generate a padded predictive subarea by padding the pixel located outside the object image with the substitute pixel value;
• and
• a device operable to obtain and code a pixel value difference between one of the target subareas and the padded predictive subarea.

An image predictive coding method for coding a target image by obtaining a predictive image from a reference image including an object image comprising a luminance signal indicating a pixel value and a shape signal indicating whether a pixel is located inside or outside the object image, said method comprising:

• dividing the target image into a plurality of target subareas;
• determining, according to the shape signal, whether each pixel included in a subarea of the reference image is an inside pixel located inside the object image or an outside pixel located outside the object image, the subarea of the reference image including both the outside pixel and the inside pixel;
• obtaining a substitute pixel value for the outside pixel included in the subarea of the reference image by using a pixel value of the inside pixel included in the subarea of the reference image;
• generating a padded predictive subarea by padding the outside pixel located outside the object image with the substitute pixel value;
• and
• obtaining and coding a pixel value difference between the target subarea and the padded predictive subarea.