Dynamic Presentation Area Normalization for Obstructed Camera Views
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing videoconferencing systems struggle with capturing and presenting user-supplied presentation material on non-perpendicular surfaces, such as whiteboards, due to camera angles that cause distortion and obstruction by surrounding objects, requiring manual boundary definition or fixed size limits.
Innovation Solution
A dynamic presentation area captured image normalizing system that automatically detects and adjusts the presentation area on planar surfaces, removing obstructions and reorienting images to provide an orthogonal view, using image differencing and machine-learning object detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a camera is oriented at a non-perpendicular angle to the presentation surface, then the camera can capture a wider viewing area, but the captured image becomes distorted and obstructed by surrounding objects
Solution Approach 1:
The system dynamically adjusts the presentation area boundaries and normalization parameters based on detected objects and their positions. The boundaries are not fixed but adapt in real-time as objects move or change position, allowing the system to maintain accurate orthogonal representation while capturing a wide viewing area.
Solution Approach 2:
The system introduces an intermediary normalization process that transforms the distorted captured image into an orthogonal representation. This intermediary step (image differencing, object detection, boundary definition, and normalization) mediates between the wide-angle capture and the requirement for distortion-free presentation material display.
2Measurement precision
If manual boundary definition is used, then the presentation area can be precisely defined, but the system complexity and operational burden increase
Solution Approach 1:
The system performs self-service by automatically detecting objects, determining presentation area boundaries, and defining the normalization region without requiring manual user input. The system uses image differencing and machine learning object detection to autonomously identify what constitutes the presentation area and how to normalize it.
Solution Approach 2:
The system changes parameters dynamically based on detected object positions and characteristics. Instead of using fixed boundaries, the system adjusts boundary coordinates, normalization scales, and transformation parameters in real-time based on the detected presentation material and surrounding objects.
3Device complexity
If fixed size limits are imposed on the presentation area, then the system is simpler to implement, but the adaptability to changing presentation content is reduced
Solution Approach 1:
The presentation area boundaries are made dynamic rather than fixed. The system continuously monitors and adapts the boundaries based on detected objects and their positions, allowing the presentation area to expand or contract as needed to accommodate changing presentation content while maintaining orthogonal representation.
Solution Approach 2:
The system performs preliminary object detection and boundary determination before normalization occurs. By pre-identifying the presentation area and surrounding objects, the system can prepare appropriate normalization parameters and boundaries in advance, enabling flexible adaptation to various presentation scenarios.
Data Source
AI summary
An information handling system executing code instructions of a dynamic presentation area captured image normalizing system and operating a web camera hardware system to transceive data for a videoconferencing application during a videoconferencing session, the camera to capture an image of user-supplied presentation material on a planar presentation surface in a viewing area of the image, a hardware processor to execute code instructions to identify the user-supplied presentation material within the image, determine boundaries of a polygon enclosing the user-supplied presentation material, generate an orthogonal view image of an orthogonally-adjusted user-supplied presentation material by proportionally reorienting pixels of the user-supplied presentation material across a rectangular polygon to appear as orthogonally oriented to the user and a network interface device to transmit the orthogonal view image of the orthogonally-adjusted user-supplied presentation material to a remote participant in the videoconferencing session.