Face Detection in Spherical Images Using Stitch Line Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face detection techniques for spherical images are computationally expensive and struggle to accurately identify faces near stitch lines due to distortion and varying sizes of image objects, leading to increased resource usage and potential false negatives.
Innovation Solution
The system renders views of spherical images along stitch lines by determining a scaling factor based on face characteristics, adjusting the depiction of face portions across multiple views to ensure consistent size and orientation, thereby reducing computational load and improving detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional face detection techniques are applied to spherical images, then face detection can be performed, but computational expense increases and detection accuracy decreases for faces near stitch lines
Solution Approach 1:
The patent divides the spherical image into multiple overlapping 2D views or hemispheres. By segmenting the spherical image and processing each view separately with appropriate scaling, the system reduces computational expense while maintaining detection accuracy for faces near stitch lines.
Solution Approach 2:
The patent applies different scaling factors to different regions of the spherical image based on local distortion characteristics. Views closer to the center use different scaling than those near the edges or stitch lines, optimizing detection accuracy locally while reducing overall computational load.
2Measurement precision
If spherical images are processed without scaling adjustment, then processing is simpler, but face detection accuracy deteriorates due to varying object sizes across views
Solution Approach 1:
The patent changes the scaling parameter dynamically based on the view being processed. Each 2D view of the spherical image is scaled according to its specific distortion characteristics, allowing consistent face size representation across all views while maintaining manageable processing complexity through systematic parameter adjustment.
3Reliability
If multiple views of spherical image are rendered with consistent face scaling, then detection accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary scaling calculations and establishes scaling factors for each view before actual face detection begins. By pre-processing the spherical image to account for distortion and scaling requirements, the system reduces processing time during the actual detection phase while maintaining high detection accuracy.
Data Source
AI summary
A face located along a stitch line in a spherical image is detected by rendering views of regions of the spherical image along the stitch line. The spherical image may be produced by combining first and second images. A first view of a projection of the spherical image is rendered. A scaling factor for rendering a second view of the projection is determined based characteristics of the first portion of the face. The second view is then rendered according to the scaling factor. The use of the scaling factor to render the second view causes a change in the depiction of the second portion of the face. For example, the scaling factor can indicate to change the resolution or expected size of the second portion of the face when rendering the second view. A face is then detected within the spherical image based on the rendered first and second views.