National Yang Ming Chiao Tung University

Winbond Electronics Corporation

Sheng-Jyh Wang

I-Hsien Chen

Ching-Chun Huang

Yi-Chung Chen

Chen-Yu Tseng

A method and an electronic apparatus for locating the center of a pupil are provided. The method includes: obtaining a glint-free image including a pupil; generating a gradient magnitude graph and a gradient orientation graph based on the glint-free image; sifting a plurality of first pixels from the gradient magnitude graph, and finding a plurality of second pixels corresponding to the plurality of first pixels in the gradient orientation graph; taking a negative gradient orientation to each of the second pixels to generate a gradient emission graph including a plurality of emissions; defining a center of the pupil in the glint-free image based on the plurality of emissions in the gradient emission graph.

Method for locating center of pupil and electronic apparatus

In accordance with an embodiment, a method for image matting based on a hierarchical graph model, comprises receiving an input image including a plurality of image elements; generating a plurality of matting cells based on the input image, each cell including a group of image elements; calculating affinity values for the plurality of matting cells based on the input image; forming a graph based on the plurality of matting cells and the affinity values, the graph including a plurality of nodes representing the matting cells and a plurality of edges associated with the affinity values of the matting cells; and generating a plurality of matting components for the input image based on the graph.

Methods and systems for image matting and foreground estimation based on hierarchical graphs

A method for automatically detecting and tracking multiple targets in a multi-camera surveillance zone and system thereof. In each camera view of the system only a simple object detection algorithm is needed. The detection results from multiple cameras are fused into a posterior distribution, named TDP, based on the Bayesian rule. This TDP distribution represents a likelihood of presence of some moving targets on the ground plane. To properly handle the tracking of multiple moving targets with time, a sample-based framework which combines Markov Chain Monte Carlo (MCMC), Sequential Monte Carlo (SMC), and Mean-Shift Clustering, is provided. The detection and tracking accuracy is evaluated by both synthesized videos and real videos. The experimental results show that this method and system can accurately track a varying number of targets.

Method for automatic detection and tracking of multiple targets with multiple cameras and system therefor

The present invention discloses a dynamic calibration method for a single and multiple video capture devices. The present invention can acquire the variations of the pan angle and tilt angle of a single video capture device according to the displacement of the feature points between successive images. For a plurality of video capture devices, the present invention includes the epipolar-plane constraint between a plurality of video capture devices to achieve the goal of dynamical calibration. The calibration method in the present invention does not require specific calibration patterns or complicated correspondence of feature points, and can be applied to surveillance systems with wide-range coverage.

Growing community of inventors

Hsinchu, Taiwan

Sheng-Jyh Wang