A characterising feature of the present invention is that, when a depth inter-view prediction flag is acquired in accordance with the current block type and an indication is given of the use of depth inter-view prediction in the coding of the current block in accordance with the depth inter-view prediction flag, colour information about a reference block corresponding to the current block is acquired, and colour information about the current block is predicted using the colour information about the reference block. By carrying out inter-view prediction using depth information, the present invention makes it possible to use depth information to guide a macro-block motion vector in a decoder. Consequently, the volume of transmitted bits can be reduced since the macro-block motion vector does not need to be coded separately. Also, the present invention makes it possible to effectively indicate whether to perform inter-view prediction or depth inter-view prediction, due to the use of a flag which indicates whether to perform inter-view prediction using depth information.