Subtitle detection device and subtitle detection method

Abstract

A subtitle detection device includes an image processing circuit, a language recognition circuit, an audio detection circuit, a language conversion circuit, and a similarity comparison circuit. The image processing circuit is configured to perform a text detection to an image to generate a first text region of the image. The language recognition circuit is configured to perform a language recognition to the first text region to generate a first image text. The audio detection circuit is configured to perform a speech recognition to an audio to generate an audio text. If the first image text and the audio text are in different language types, the language conversion circuit is configured to convert the first image text and the audio text into the same language type. The similarity comparison circuit is configured to compare the first image text with the audio text to generate a first similarity score. If the first similarity score is larger than a threshold, the image is determined to contain a subtitle.

Images