Face Tracking Using Edge Feature Displacement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing face tracking algorithms, such as the Camshift algorithm based on skin color, suffer from low tracking accuracy due to environmental and background interference, and struggle with quickly moving targets and varying face expressions.

Innovation Solution

A method and device for face tracking that utilizes edge feature points to determine the movement trajectory of a target face by calculating relative displacement between edge feature points in successive frames, improving accuracy and robustness against noise and environmental changes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If skin color model is used for face tracking, then real-time tracking can be achieved, but tracking accuracy deteriorates due to environmental and background interference

Engineering Contradiction:

Improvereal-time tracking speedVSAvoidtracking accuracy

Core Design Contradiction:

SpeedVSMeasurement precision

Solution Approach 1:

The patent changes the feature parameter from skin color (which is sensitive to environmental factors) to edge features (which are more stable and invariant to lighting and background changes). This parameter transformation maintains real-time tracking capability while significantly improving tracking accuracy by focusing on structural boundaries rather than color information.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the skin color-based tracking mechanism with an edge feature-based tracking mechanism. Instead of relying on color segmentation which is vulnerable to environmental interference, the system uses edge detection and matching algorithms that are more robust to changes in lighting, background, and facial expressions.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Ease of manufacture

If skin color feature is used for tracking, then the algorithm is simple to implement, but anti-interference ability deteriorates

Engineering Contradiction:

Improvealgorithm implementation simplicityVSAvoidanti-interference ability

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The patent transforms the feature parameter from skin color to edge features, which fundamentally changes the nature of the tracking data. Edge features provide structural information that is inherently more resistant to environmental interference, thereby improving reliability while maintaining algorithmic simplicity through standard image processing techniques.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent uses disposable edge feature extraction from each frame rather than maintaining complex color models. By extracting edge features independently from each frame and matching them, the system achieves robust anti-interference capability without requiring complex pre-trained color models or extensive calibration.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

3Adaptability or versatility

If Camshift algorithm is used, then self-adaptive adjustment of searching window can be achieved, but tracking accuracy deteriorates under varying face expressions and movements

Engineering Contradiction:

Improveself-adaptive window adjustmentVSAvoidtracking accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent maintains the dynamic self-adaptive window adjustment capability of Camshift but enhances it by using edge features instead of skin color. The searching window dynamically adapts to face position and size changes while edge feature matching provides more accurate correspondence between frames, even under varying expressions and movements, thereby improving tracking accuracy.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent segments the face tracking problem into edge detection, edge feature extraction, and edge matching components. By focusing on edge segmentation rather than color segmentation, the system achieves better adaptability to expression changes while maintaining the self-adaptive window adjustment mechanism for accurate tracking.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS10747988B2Method and device for face tracking and smart terminal

Publication Date: 2020.08.18 GOERTEK INC

AI summary

A method and device for face tracking and a smart terminal. The method includes: collecting a frame of an image in a video stream and conducting face detection and identification, to identify a target face; determining a position of a window to be tracked that contains the target face according to the identified target face, and extracting an edge feature point in the window to be tracked of the reference frame; determining an initial position of a window to be tracked on the current frame in the video stream, and extracting an edge feature point in the window to be tracked of the current frame; calculating a relative displacement between the edge feature point of the reference frame and the edge feature point of the current frame; and adjusting the initial position of the window to be tracked of the current frame by using the relative displacement.