摘要 |
Disclosed are a video detection method and device. The implementation of the method comprises: acquiring N full-screen images of consecutive frames, and dividing the N full-screen images into grids of a preset size in a same division mode; detecting whether images in the grids are changed in the N full-screen images by using a grid as a unit, and recording grids corresponding to changed images; and determining a smallest region in grid regions of overwritten records as a candidate video region. Full-screen images of consecutive frames are divided into grids of a preset size in the same division mode, then grids corresponding to changed images are determined by means of image detection, and finally a candidate video region is determined accordingly. The process of determining a candidate video region does not depend on a driver, and accordingly the corresponding technical problem caused when a video region is determined with the aid of the driver can be avoided, and a video detection scheme that is universal and that does not affect hardware performance is provided. |