Pose Recognition using Kinect for Home Care System

Slides:



Advertisements
Similar presentations
人機介面 Gesture Recognition
Advertisements

Robust Part-Based Hand Gesture Recognition Using Kinect Sensor
Image-based Clothes Animation for Virtual Fitting Zhenglong Zhou, Bo Shu, Shaojie Zhuo, Xiaoming Deng, Ping Tan, Stephen Lin * National University of.
Robust 3D Head Pose Classification using Wavelets by Mukesh C. Motwani Dr. Frederick C. Harris, Jr., Thesis Advisor December 5 th, 2002 A thesis submitted.
Copyright 2007 Healing Touch Program Introduction to.
Department of Electrical and Computer Engineering He Zhou Hui Zheng William Mai Xiang Guo Advisor: Professor Patrick Kelly ASLLENGE.
I mage and M edia U nderstanding L aboratory for Performance Evaluation of Vision-based Real-time Motion Capture Naoto Date, Hiromasa Yoshimoto, Daisaku.
New Features Jan 2014.
Introduce about sensor using in Robot NAO Department: FTI-FHO-FPT Presenter: Vu Hoang Dung.
REAL ROBOTS. iCub It has a height of 100 cm, weighs 23 Kg, and is able to recognize and manipulate objects. Each hand has 9 DOF and can feel objects almost.
Jan SedmidubskySeptember 23, 2014Motion Retrieval for Security Applications Jan Sedmidubsky Jakub Valcik Pavel Zezula Motion Retrieval for Security Applications.
Abstract Design Considerations and Future Plans In this project we focus on integrating sensors into a small electrical vehicle to enable it to navigate.
Kinect Part II Anna Loparev.
Safe Mobile Devices Designs Kristen Kuron JMA464; Dr. Gibbs Prototype A Prototype B.
REU Project RGBD gesture recognition with the Microsoft Kinect Steven Hickson.
Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa.
Hand Tracking for Virtual Object Manipulation
1 Computational Vision CSCI 363, Fall 2012 Lecture 31 Heading Models.
Interactive Spaces Huantian Cao Department of Computer Science The University of Georgia.
An Information Fusion Approach for Multiview Feature Tracking Esra Ataer-Cansizoglu and Margrit Betke ) Image and.
Under construction Caroline Steiner & Michael Beham.
A New Fingertip Detection and Tracking Algorithm and Its Application on Writing-in-the-air System The th International Congress on Image and Signal.
Online Kinect Handwritten Digit Recognition Based on Dynamic Time Warping and Support Vector Machine Journal of Information & Computational Science, 2015.
Class 13 LBSC 690 Information Technology More Multimedia Compression and Recognition, and Social Issues.
By: Lee Vang. First attempt to make a humanoid Robot by Honda was in 1986 (Model E0) History:
EEC 490 GROUP PRESENTATION: KINECT TASK VALIDATION Scott Kruger Nate Dick Pete Hogrefe James Kulon.
SciFest Overview Neil Gannon. Outline Demonstrations using a Microsoft Kinect sensor – Image Manipulation Real-time Invisibility Background removal (green.
Wiimote/Kinect Lab Midterm Update Senior Design December 2011, Group 16 Adviser: Dr. Tom Daniels Brenton Hankins Rick Hanton Harsh Goel Jeff Kramer.
Chapter 1 Fitness and Wellness for ALL OBJECTIVES Define Physical Fitness, Health and Wellness. Describe some of the benefits of Fitness, Health and Wellness.
REU Project RGBD gesture recognition with the Microsoft Kinect.
ROBOGRAPHERS FACIAL EXPRESSION RECOGNITION USING SWARMS SPONSORED BY: DR. KATIA SYCARA TEAM : GAURI GANDHI SIDA WANG TIFFANY MAY JIMIT GANDHI ROHIT DASHRATHI.
Big Data and IOT Laboratory by Dr. Dan Feldman. Asaf Slilat & Omri Cohen Fly With Me Zone 1 Zone 2 Target Anchor Point – Odroid U3 – PS Eye Camera –
Faculty of Information Technology, Brno University of Technology, CZ
Creative Coding & the New Kinect
Hand Gestures Based Applications
Body Tracking and Gesture Recognition Aaron Pulver
Demo.
Computer vision: models, learning and inference
First-person Teleoperation of Humanoid Robots
Attention Tracking Tool
Automated Detection of Human Emotion
Implementing Localization
Depth estimation and Plane detection
Charles Cole Mark Cerritelli Matthew Fister Mine Yalcinalp.
A language assistant system for smart glasses
Structured Predictions with Deep Learning
NBKeyboard: An Arm-based Word-gesture keyboard
ED STROKE ALERT Competency
Video-based human motion recognition using 3D mocap data
OPERAcraft + Kinect = Cinemacraft
Efficient Deformable Template Matching for Face Tracking
Jia-Bin Huang Virginia Tech ECE 6554 Advanced Computer Vision
Kinect for Creative Development with open source frameworks
Multi-Sensor Soft-Computing System for Driver Drowsiness Detection
CS2310 Zihang Huang Project :gesture recognition using Kinect
AHED Automatic Human Emotion Detection
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
Global Challenge Love Heart Lesson 1.
AHED Automatic Human Emotion Detection
Anarghya Mitra and Zelun Luo
AHED Automatic Human Emotion Detection
MAPVI: Meeting Accessibility for Persons with Visual Impairments
Bidirectional LSTM-CRF Models for Sequence Tagging
Sign Language Recognition With Unsupervised Feature Learning
CS2310 Milestone2 Zihang Huang Project: pose recognition using Kinect
Presentation transcript:

Pose Recognition using Kinect for Home Care System Zihang Huang CS2310 Project presentation

Goal Detect human being in a room using Kinect Segment each parts of the body, such as torso, hands, head Diagnose symptoms based on video stream Falling detection Identify hand gestures to decide data flow

System Overview 3D skeleton Diagnose Symptom frames OpenPose status ……. ……. ……. OpenPose

Component Diagram Falling Detection Index Cell M2 M3 Pose Estimation index cell Status Index Cell M0 M1 M4 Diagnose Symptoms Index Cell M5 Frame message(M0) is sent to pose estimation index cell generating 3D skeleton message(M1). Status index cell sends initial status(M2 and M5) to both falling detection index cell and diagnose symptoms index cell. Falling detection index cell will return detection result(M3) back to status index cell. When diagnose symptoms index cell detects symptom, it will send symptom message(M5) to status index cell updating current status.

Emergency Officer(EO) Diagnose Symptoms T1 T2 Emergency Officer(EO) T3 P0 P5 P1 P2 P3 T4 T1: Headache T2: Heart Attack T3: Stomachache T4: Left Knee Pain T5: Right Knee Pain P0: initial status P1:Symptom diagnosed P2: Show symptom text P3: Send message to EO P4: Confirm gesture P5: message fetching P4 T5

Emergency Officer(EO) Falling Detection T1 Emergency Officer(EO) P0 P5 P1 P2 P3 T2 T1: Movement velocity per 10 frames T2: The distance between the middle point of the body and ground P4 P0: initial status P1: Falling detected P2: Show “falling detected” text P3: Send message to EO P4: Confirm gesture P5: message fetching

Headache demo -headache

Headache Demo -confirm

Headache Demo -cancel

Falling Demo

Falling Demo

Demo https://www.youtube.com/watch?v=xthx0zhqqA0

Conclusion So far, this system can detect five symptoms, headache, heart attack, stomachache, left and right knee pain. And each of components are able to run individually. A valid symptom detection requires a specific hand confirmation gesture which is a designed fist gesture. Walking around or jumping in the room will not affect falling detection.

Future Work Explore multiview application Improve detection accuracy and precision Integrate face emotion detection to diagnose symptoms Record symptom times, duration and frequency and apply deep learning to predict potential risk of diseases

Reference Paper: [1] Cao, Z., Simon, T., Wei, S. E., & Sheikh, Y. (2016). Realtime multi-person 2d pose estimation using part affinity fields. arXiv preprint arXiv:1611.08050. [2] Simon, T., Joo, H., Matthews, I., & Sheikh, Y. (2017). Hand Keypoint Detection in Single Images using Multiview Bootstrapping. arXiv preprint arXiv:1704.07809. GitHub: OpenPose https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenKinect https://github.com/OpenKinect/libfreenect2 OpenCV https://github.com/opencv/opencv My project repository https://github.com/derekwong66/pose-estimation-in-home-care-system