Eye Gaze Tracking Using Custom Template Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye gaze tracking methods suffer from inaccuracy, lack of real-time capability, difficulty in administration, high cost, instability, and inconsistency.
Innovation Solution
A system comprising an eye camera and processing devices captures frame-by-frame images of the eye, creates a custom template, and uses a tracking algorithm to determine the gaze angle, with calibration points and multiple correlation calculations at low and high resolutions to achieve precise and real-time eye movement tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If frame by frame eye images are captured and processed in real-time using tracking algorithms, then measurement precision of eye gaze is improved, but device complexity and processing requirements increase
Solution Approach 1:
The patent divides the eye image processing into multiple independent components: detecting the glint position (reflection point), detecting the pupil center position, calculating the vector between them, and determining gaze direction. This segmentation allows each component to be processed independently and efficiently, reducing overall computational complexity while maintaining measurement precision
Solution Approach 2:
The patent performs preliminary actions by capturing frame-by-frame images at a high frame rate (e.g., 60 fps) before full processing is needed. The glint and pupil positions are pre-detected in each frame, and these preliminary detections are stored and processed sequentially to determine gaze direction, enabling real-time performance without requiring all processing to occur simultaneously
2Measurement precision
If multiple correlation calculations are performed at high resolution by splitting images into multiple pieces with parallel processing, then measurement precision is improved, but use of energy and computational resources increases
Solution Approach 1:
The patent divides the high-resolution eye image into multiple smaller pieces or regions of interest for parallel correlation calculations. Each processing device handles a specific region, performing correlation calculations independently. This segmentation reduces the computational burden on each individual processor, enabling parallel execution that improves precision while distributing energy consumption across multiple devices
Solution Approach 2:
The patent performs correlation calculations on only the relevant portions of the eye image (such as the iris and pupil regions) rather than processing the entire high-resolution image. By focusing computational resources on the critical areas that contain gaze information, the system achieves high measurement precision with reduced overall energy consumption
3Measurement precision
If custom templates of the eye are created and used in tracking algorithms, then measurement precision and consistency are improved, but difficulty in detecting and measuring increases
Solution Approach 1:
The patent implements automatic template creation that uses the operator's own eye image to generate the custom template. The system detects key features (glint, pupil center, iris boundaries) from a sample eye image and automatically constructs the template without requiring manual annotation or complex user input. This self-service approach maintains high measurement precision while significantly reducing the difficulty of template creation
4Adaptability or versatility
If head tracking device is integrated with eye camera system, then adaptability of the system is improved, but device complexity increases
Solution Approach 1:
The patent combines the head tracking device with the eye camera system into an integrated platform. The head tracking data and eye image data are merged and processed together to determine overall gaze direction. This merging allows the system to adapt to different tracking needs (head position, eye position, or combined) while sharing common hardware and software resources, thereby improving adaptability without proportionally increasing complexity
Data Source
AI summary
The invention relates to systems and methods for tracking the movement of an operator's eyes in order to determine the real-time gaze angle of the eyes of the operator. In one embodiment, the invention may utilize an eye camera and a processing device. A frame by frame image of the eye may be captured utilizing the camera and the processing device, and a custom template resembling the operator's eye may be created utilizing the image produced by the eye camera. The custom template may be substantially the shape and size of the pupil of the eye of the operator, or may also represent additional features of the eye of the operator, such as the iris, sclera, cornea, or eye-lids. A tracking algorithm may determine a substantially real-time gaze angle of the eye.