Multi-Perspective Human Body Positioning for Precise Light Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lighting systems for entertainment face inaccuracies in human body tracking due to manual control, interference from active and passive tags, and environmental obstructions, leading to unreliable positioning and tracking.
Innovation Solution
A human body positioning method using multiple camera perspectives to capture images from different angles, extract feature points, match and calculate 3D coordinates, and weight reprojection errors for accurate projection distribution, eliminating mis-matching interference without the need for active or passive tags.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual control of light beam projection is used, then the lighting system is simple to operate, but the positioning accuracy is poor due to anthropic factors
Solution Approach 1:
The patent replaces manual mechanical control with an automated computer vision system. Multiple cameras capture images, feature points are extracted and matched, 3D coordinates are calculated, and light fixtures are automatically controlled to track human bodies. This substitution eliminates anthropic factors while maintaining ease of operation through automated processing.
Solution Approach 2:
The patent introduces an intermediary computer vision system between the human bodies and the lighting control. The system uses camera images, feature point extraction, and 3D reconstruction as intermediate steps to accurately determine human body positions and control light beam projection without direct manual intervention.
2Measurement precision
If active tags are used for tracking, then the positioning can be achieved, but the data transmission may be interfered by human body or surrounding scene causing inaccurate positioning
Solution Approach 1:
The patent extracts feature points directly from camera images of human bodies without using any tags or markers. By taking out the dependency on active tags and their data transmission, the system eliminates interference from human bodies or surrounding scenes that would affect tag signal transmission. The positioning is achieved purely through visual feature extraction and 3D reconstruction.
3Reliability
If passive tags are used for tracking, then the positioning can be achieved, but the tag may be obscured causing failure of tracking
Solution Approach 1:
The patent removes passive tags from the system entirely and instead extracts feature points directly from visual images. This eliminates the problem of tag obscuration since the feature points are derived from the camera's field of view and can be detected as long as the human body is visible to the cameras, providing more reliable tracking without physical tags that can be blocked.
Solution Approach 2:
The patent transitions from 2D tag detection to 3D position reconstruction through multi-perspective imaging. By capturing images from multiple camera angles and performing 3D reconstruction through feature point matching, the system obtains reliable positioning information that is not susceptible to obscuration from any single viewpoint.
4Measurement precision
If multi-perspectives and weighting method are used, then the positioning accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent introduces a computer-based image processing and calculation system as an intermediary to handle the complexity of multi-perspective processing. The system automatically performs feature point extraction, matching, 3D coordinate calculation, and weighted averaging, transforming a complex multi-step process into an automated computational workflow that maintains high positioning accuracy while managing system complexity through software automation.
Data Source
AI summary
A human body positioning method based on multi-perspectives includes simultaneously capturing pictures from at least two perspectives angles and extracting human body feature points, matching the human body feature points of a same human body of a same position in different pictures to form matched point groups and calculating 3D coordinates of corresponding 3D point of each matched point group in a 3D space, further calculating and weighting spatial geometric error between the 3D point and each human body feature point of the corresponding matched point group in each original picture to obtain reprojection error of the 3D point, and subsequent to projecting each 3D point onto the ground to obtain single-point 2D coordinates thereof on the ground, weighting the projected single-point 2D coordinates according to the reprojection error of the corresponding 3D point to obtain a projection distribution of the corresponding human body on the ground.