Skin-tone Filtering for Portable Face Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face detection algorithms for consumer portable devices are either computationally expensive and power-intensive due to their ability to handle various lighting conditions, or ineffective in outdoor and indoor settings due to their reliance on controlled lighting conditions.
Innovation Solution
The implementation of skin-tone filtering using an orange-distance filter, which analyzes digital image frames to identify human skin tones independently of lightness, reduces computational resources and power consumption by preprocessing images to emphasize facial features and neutralize lighting effects, allowing for effective face detection on portable devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If face detection algorithms that address various lighting conditions are used, then face detection accuracy is improved, but computational cost and power consumption increase
Solution Approach 1:
The patent segments the face detection process into multiple stages: initial skin-tone filtering to identify candidate regions, followed by more sophisticated analysis only on those regions. This segmentation allows the system to maintain high detection accuracy while reducing overall computational cost and power consumption by avoiding intensive processing on the entire image.
Solution Approach 2:
The patent applies preliminary skin-tone filtering and lighting normalization before performing main face detection algorithms. This preliminary action preprocesses the image to emphasize skin regions and reduce lighting variations, making subsequent detection more efficient and accurate while reducing the computational burden on the main detection algorithm.
2Use of energy by moving object
If face detection algorithms specialized for controlled lighting conditions are used, then computational resources are reduced, but effectiveness decreases in uncontrolled environments
Solution Approach 1:
The patent changes the color space parameters by converting images to HSV (Hue, Saturation, Value) color space and specifically analyzing the Hue channel. This parameter transformation allows the system to identify skin tones independently of lighting conditions, as skin-tone hues remain relatively stable across different illuminations, thereby improving adaptability without requiring intensive computational resources.
Solution Approach 2:
The patent introduces an intermediary skin-tone filtering step that acts as a bridge between the image input and the main face detection algorithm. This intermediary process uses simple hue-based filtering to identify potential skin regions, making the subsequent detection algorithm more robust to lighting variations while maintaining computational efficiency.
3Measurement precision
If comprehensive face detection algorithms are used, then detection accuracy is improved, but device power drainage increases
Solution Approach 1:
The patent segments the image processing into hierarchical stages: first applying lightweight skin-tone detection to identify candidate regions, then applying more computationally intensive face detection only to those segmented regions. This reduces overall power drainage while maintaining detection accuracy by avoiding unnecessary processing of non-face areas.
Solution Approach 2:
The patent performs preliminary skin-tone filtering and hue-based candidate identification before running the main face detection algorithm. This preliminary action reduces the amount of data requiring intensive processing, thereby reducing power drainage while preserving detection accuracy through the use of optimized preprocessing steps.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, are disclosed relating to skin-tone filtering for reducing the impact of lighting conditions, while providing a low-computation solution for effective face detection. In one aspect, methods include sampling a digital image frame from among a series of digital image frames. The methods further include analyzing pixels within the sampled digital image frame to determine whether pixels in the sampled digital image frame have a hue, independent of lightness, that is within a range of hues corresponding to human skin tone. Further, the methods include deciding whether the sampled digital image frame includes a depiction of human skin based on a result of the analyzing.