Multi-Pointer Detection via Pattern Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Interactive input systems face challenges in accurately detecting multiple pointers due to ambiguity and occlusion issues when multiple pointers are within the fields of view of multiple imaging devices, leading to difficulties in determining correct pointer positions.
Innovation Solution
The system captures image frames using multiple imaging devices with overlapping fields of view, processes the images to identify targets, and analyzes these targets to determine if they represent projections of an input object, thereby identifying the object by recognizing patterns in the projections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple imaging devices with overlapping fields of view are used to detect multiple pointers, then the coverage area and detection capability are improved, but ambiguity and occlusion issues arise making it difficult to determine correct pointer positions
Solution Approach 1:
The patent introduces a pattern recognition system as an intermediary between the imaging devices and pointer detection. By analyzing patterns of reflected light from the touch surface, the system can disambiguate which detected points correspond to actual pointers, resolving the measurement precision issue while maintaining broad coverage
Solution Approach 2:
The system creates a digital copy or representation of the touch surface patterns and uses this model to interpret the imaging data. By comparing detected light reflections against the expected pattern model, the system can accurately identify pointer positions even when multiple pointers are present and overlapping views are available
2Productivity
If multiple pointers are detected within the fields of view of multiple imaging devices, then the system can detect more input objects, but ambiguity arises when multiple pointers cannot be differentiated
Solution Approach 1:
The patent applies local quality analysis by examining the specific pattern characteristics at each detected point on the touch surface. By analyzing the local pattern features (such as spacing, orientation, and shape) at each detection location, the system can differentiate between multiple pointers even when they are close together or overlapping in the imaging fields
Solution Approach 2:
The system segments the detection process into distinct stages: first detecting all light reflections, then analyzing pattern characteristics at each location, and finally grouping these into distinct pointer identities. This segmentation allows the system to handle multiple pointers systematically without information loss
3Reliability
If one pointer occludes another in the field of view of an imaging device, then the captured image includes fewer pointers than actually present, but the system should still be able to detect all pointers
Solution Approach 1:
The patent transitions from two-dimensional image analysis to three-dimensional pattern analysis by considering the tactile or spatial pattern information that extends beyond the simple 2D projection. By analyzing the pattern structure in an additional dimension, the system can infer the presence and position of occluded pointers based on the overall pattern configuration
Data Source
AI summary
A method comprises capturing image frames of an input area using a plurality of imaging devices, each having a field of view encompassing at least a portion of the input area; processing captured image frames to identify a plurality of targets therein; analyzing the identified plurality of targets to determine if the targets represent a plurality of projections of an input object; and if so, identifying a pattern of the projections thereby to identify the input object.