Video Digest Generation Using IMU Movement Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating video digests are inefficient and of poor quality due to the need for extensive computing power and the impact of video quality on feature extraction.
Innovation Solution
A method and apparatus for generating a video digest that utilizes real-time movement information from a video acquisition device to dynamically determine the frequency for key frame extraction, thereby improving the accuracy and efficiency of video digest generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If video signal is analyzed frame by frame to extract features and generate video digest, then video digest quality is improved, but computing power consumption increases and generation speed decreases
Solution Approach 1:
The patent extracts only the essential movement information from video frames using IMU sensors rather than analyzing all frame features. By taking out only the critical motion data (acceleration, orientation changes) and using it to select key frames, the system achieves both quality preservation and computational efficiency.
Solution Approach 2:
The patent performs preliminary action by recording movement information during video acquisition and pre-determining key frame positions before digest generation. This advance preparation of movement data allows rapid key frame selection without requiring comprehensive frame analysis during the digest creation process.
2Measurement precision
If video signal is analyzed frame by frame to extract features, then key frames can be identified, but computing complexity increases
Solution Approach 1:
The patent introduces movement information from IMU sensors as an intermediary to bridge video content and key frame selection. Instead of directly analyzing video frames for key frame identification, the system uses movement data as a mediator to determine which frames are worth capturing, significantly reducing computational complexity while maintaining accuracy.
Solution Approach 2:
The patent replaces the mechanical/computational system of frame-by-frame video analysis with a sensor-based measurement system. By substituting visual feature extraction with IMU movement detection, the system achieves key frame identification with much lower computing complexity.
3Measurement precision
If video quality is poor, then feature extraction accuracy decreases, but the existing method still requires extensive computing power
Solution Approach 1:
The patent uses movement information as an intermediary that is independent of video quality. Since IMU sensors measure physical movement rather than visual features, the key frame selection remains accurate even when video quality is poor, avoiding the trap of spending computing power on extracting features from low-quality frames.
Data Source
AI summary
A method and an apparatus for generating a video digest, and a readable storage medium are provided. The method is applied to a video acquisition device that includes a hardware unit configured to obtain movement information in real time, the movement information includes speed information and direction information of the video acquisition device. The method includes: starting acquisition of a video signal in response to an acquisition start instruction, and recording movement information corresponding to each frame during the acquisition of the video signal; determining a frequency for key frame extraction according to the movement information corresponding to each frame; determining a key frame from the acquired video signal according to the frequency for key frame extraction; and in response to an acquisition end instruction, generating a video digest according to the key frame.