3D Shape Reconstruction for Occluded Ball Separation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately determine when a ball separates from a player's hand in sports games like basketball due to occlusion issues, making it difficult to enforce rules such as the 24-second rule.
Innovation Solution
A system that generates three-dimensional shape data from multiple images, estimates the size of objects, and specifies the position of the ball relative to the player using a comparison with a reference value, enabling precise determination of when the ball separates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple imaging apparatuses are used to capture images for determining ball separation, then the ability to detect occluded objects improves, but the system complexity and processing difficulty increase
Solution Approach 1:
The patent transitions from two-dimensional image analysis to three-dimensional shape data representation. By constructing 3D voxel models from multiple 2D images, the system creates a volumetric representation that inherently resolves occlusion issues, as objects are represented in three spatial dimensions rather than projected onto a flat plane.
Solution Approach 2:
The patent segments the captured scene into discrete voxel units that can be individually processed and assembled. Each voxel represents a small volumetric element, allowing the system to build up complete 3D representations of players and balls by integrating data from multiple imaging apparatuses, thereby managing complexity through modular processing.
2Measurement precision
If three-dimensional shape data is generated from multiple images to determine ball separation, then the accuracy of determining when the ball separates from the player's hand improves, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing images to extract silhouette information and pre-constructing 3D voxel models before the actual ball separation determination is needed. This allows the system to have ready-made 3D representations of players and balls, reducing the computational burden during real-time separation detection.
Solution Approach 2:
The patent replaces complex mechanical image analysis with a standardized 3D voxel construction process. By substituting the need for sophisticated 2D image processing algorithms with a more straightforward 3D voxel assembly method, the system achieves better accuracy while managing processing time through algorithmic simplification.
3Reliability
If the system uses silhouette information to construct three-dimensional models, then the ability to identify objects in occlusion situations improves, but the manufacturing complexity of the processing system increases
Solution Approach 1:
The patent creates simplified 3D voxel copies of actual objects based on their silhouette projections. Instead of directly processing complex raw image data, the system generates volumetric copies that retain the essential shape and position information needed for reliable object identification, while being much easier to manipulate and analyze computationally.
Data Source
AI summary
The information processing apparatus generates three-dimensional shape data representing a three-dimensional shape of a first object by using a plurality of images obtained by a plurality of imaging apparatuses capturing the first object. The information processing apparatus estimates a size of the first object by using the three-dimensional shape data. The information processing apparatus specifies a position of a second object that is captured by the plurality of imaging apparatuses based on results of a comparison between the estimated size of the first object and a predetermined reference value.