2D Cursor Positioning via Facial Feature Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
During meetings, participants often struggle to indicate specific points on presentation slides as others may not see where they are pointing, and existing pointing detection solutions are complex and require 3D representations, additional sensors, or cameras.
Innovation Solution
A 2D cursor positioning system that identifies a meeting attendee's face and hand in a 2D video image, defines a rectangle based on face dimensions, searches for a pointing indicator within this rectangle, and calculates a cursor position on the display screen using object recognition and tracking techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3D representation and additional sensors are used for pointing detection, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts and eliminates the unnecessary 3D representation requirement from the pointing detection system. By using 2D video images and image processing techniques, the system removes the complex 3D sensing components while maintaining pointing detection accuracy through facial feature analysis and hand position tracking in the 2D plane.
Solution Approach 2:
The patent uses 2D video images as a copy of the real scene instead of requiring direct 3D sensing. By processing 2D images that capture the presenter and screen, the system derives pointing information through image analysis, avoiding the need for complex 3D sensors while maintaining detection capability.
2Measurement precision
If 3D representation and additional cameras are used for pointing detection, then measurement precision is improved, but ease of manufacture deteriorates
Solution Approach 1:
The patent makes the existing video camera multi-functional by using it for both video conferencing and pointing detection. The same 2D video feed that captures the meeting is also processed to detect hand positions and facial features, eliminating the need for separate detection hardware and simplifying system implementation.
Solution Approach 2:
The system uses its existing video capture capability to serve the additional function of pointing detection. By analyzing the 2D video images already being captured for the meeting, the system performs pointing detection without requiring external sensors or cameras, making the system easier to manufacture and deploy.
3Adaptability or versatility
If generic gesture control schemes are used for pointing detection, then adaptability is improved, but device complexity increases
Solution Approach 1:
The patent segments the gesture recognition task into specific facial feature detection and hand position tracking within 2D images. By focusing on detecting the presenter's face, hands, and pointing gestures in the video feed, the system achieves adaptability for pointing detection without implementing a complex generic gesture control scheme.
Data Source
AI summary
In one embodiment, a method includes identifying a face, of a meeting attendee pointing to a display screen, in a first two-dimensional image from a two-dimensional video, determining at least one dimension of the face in the first two-dimensional image, defining a rectangle in the first two-dimensional image, at least one first dimension of the rectangle being a function of the at least one dimension of the face, searching for an image of a pointing indicator in the rectangle resulting in finding the pointing indicator at a first position in the rectangle, and calculating a cursor position of a cursor on the display screen based on the first position. Related apparatus and methods are also described.