To provide the image output device which has simple constitution and can accurately and easily determine an output image.

The device is equipped with a voice information input means 210 by which an operator inputs voice information and a read means 20 which reads moving picture information recorded on a recording medium and voice information recorded corresponding thereto. Consequently, a scene which is recorded corresponding to the voice information that match the voice information inputted from the voice information input means is extracted, determined as an output scene, and outputted in the form of a still picture.