RESEARCH

Image & Vision

Investigates Image data fusion techniques that combine image and track data from multiple sensors to achieve improved accuracies and more specific inferences than could be achieved by the use of a single sensor alone. Our aim is to explore the state-of-the-art image processing algorithms for achieving effective data fusion as in:

Gesture Recognition

작성자 관리자 날짜 2021-04-08 14:52:55 조회수 328

1. Introduction                                                                             [Go Top]

    - Gesture Recognition is a classification algorithm what input hand coordinates means. This process based on statistical modeling by Hidden-Markov Model (HMM).
   

2. Main algorithm and principle                                                    [Go Top]

    - General gesture recognition algorithm structure

1.    Get hand coordinates from webcam or kinect sensor.

2.    Extract feature vector from sequential coordinates; Using 2D vector(angle, length)

3.    Training each gesture coordinates using HMM

4.    About Input hand coordinates, calculate likelihood for every HMM.

5.    Choose recognition result that has maximum likelihood.

 

3. Demo System                                                                       [Go Top]

    - Gesture Recognition Using Kinect Sensor

댓글 (0)

등록된 댓글이 없습니다.
작성 권한이 없습니다.