Optical-Flow Face Tracking in Interpolated Video Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face tracking technologies are sensitive to changes in illumination, leading to deteriorated performance and increased power consumption, and require extensive processing time.
Innovation Solution
An electronic device equipped with an optical flow circuit, face detection module, and face tracking module generates facial position information in interpolated frames using optical flow maps and detection data, reducing the need for additional image processing and optimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional face tracking methods are used, then face tracking can be performed, but performance deteriorates and power consumption increases under rapid brightness changes
Solution Approach 1:
The system performs face detection and optical flow calculation in advance on key frames (first and second frames) before generating interpolated frames. By preparing detection data and optical flow maps beforehand, the system avoids repeated heavy processing during brightness changes, maintaining reliable tracking while reducing real-time power consumption
Solution Approach 2:
The system creates interpolated frames by copying and transforming detection data and optical flow information from original frames. Instead of re-processing the entire image sequence during brightness changes, the system generates new frame data by interpolating from existing processed frames, maintaining tracking accuracy while significantly reducing computational load and power consumption
2Measurement precision
If extensive image processing is performed, then face tracking accuracy is maintained, but processing time increases
Solution Approach 1:
The system segments the video stream into key frames and interpolated frames. Face detection and optical flow calculation are performed only on key frames, while interpolated frames are generated by combining detection data and optical flow information from key frames. This segmentation reduces processing time while maintaining accuracy by avoiding redundant processing of all frames
Solution Approach 2:
The system performs computationally intensive face detection and optical flow calculation in advance on key frames before generating interpolated frames. By completing the heavy processing work beforehand, the system can quickly generate interpolated frames without repeated extensive processing, reducing overall processing time while maintaining tracking accuracy
Data Source
AI summary
A method of generating facial position information in interpolated frames of an electronic device including a face detection module, an optical flow circuit, and a face tracking module, the method including receiving, by the face tracking module, a frame and detection data from the face detection module and receiving an optical flow map from the optical flow circuit, determining, by the face tracking module, at least one patch of the optical flow map based on the frame and the detection data, calculating an estimated position of a face in the frame based on the patch, and generating facial position information in interpolated frames based on the calculated estimated position, wherein the detection data includes bounding box information of the face included in the frame, and landmark information of the face.