METHOD FOR OUTPUT CREATION BASED ON VIDEO CONTENT CHARACTERISTICS

摘要

The present application is directed to new methods for automatically determining several characteristics of frames in a video sequence and automatically recommending or preparing image output products based on those frame characteristics. In some embodiments, motion characteristics of particular image frames are calculated, and those motion characteristics are automatically used to prepare or recommend image output products suitable for the motion characteristics of the frames. In other embodiments, facial, audio, and overall image quality are assessed and used to automatically recommend or prepare image output products. In still other embodiments, image frames in a video sequence are analyzed for various user-specified characteristics, which characteristics are then used to automatically recommend or prepare image output products.

主权项

1. A method of selecting a group of frames from a video sequence, comprising:
accessing a video sequence from a collection of media assets stored in a memory; using a data processor to automatically analyze the video sequence to compute motion characteristics of the video frames; classifying the video frames according to the computed motion characteristics; for each group of classified video frames, computing a quality measure for the video frames and an overall quality value for the video frames; storing the group of classified video frames for each motion class in a memory.