Feature extraction method and device based on video recognition and storage medium

The invention relates to the technical field of video processing, and discloses a feature extraction method and device based on video recognition and a storage medium, and the method comprises the steps: obtaining a target video and an expected number; extracting voice data, converting the voice data into text fragments, and cleaning the text fragments: removing non-standard data to obtain an intermediate text; obtaining all simple sentence final results, simple sentence starting time and simple sentence ending time in the intermediate text to form a standard text; dividing text paragraphs from the standard text; marking time intervals for the text paragraphs; according to the duration value a of the target video, the expected number b and the number value c of the text paragraphs, calculating interval time x = a/b and the number y = b/c of the corresponding features; if yes; 1, merging the text paragraphs to obtain a final paragraph Distributing a feature number for the final paragraph; if clt; b, distributi