Provided is an image decoding device including an acquisition unit configured to acquire clean random access (CRA) information used to identify one or more CRA pictures in an image sequence inserted into a data region from a header region of a file format including the header region and the data region, a control unit configured to specify one CRA picture in the image sequence as a decoding start picture using the CRA information when an instruction of random access is detected, and a decoding unit configured to decode the image sequence from the decoding start picture specified by the control unit.