21
Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa 1

Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Embed Size (px)

Citation preview

Page 1: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

1

Human Gesture Recognition Using Kinect Camera

Presented by Carolina Vettorazzo and Diego Santo

Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Page 2: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

IntroductionThis work proposes a comparison

of human gesture recognition using data mining classification methods

The gestures where chosen to be the knowledge base of a smart home system which monitors and detects the fall motion of the elderly or hospital patients.

Page 3: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

IntroductionHuman gesture

◦Hands, arms, and body◦Movements of the head, face, and

eyes

Performance of recognition methods◦Light conditions◦Shadows◦Camera angle◦Occlusion

Page 4: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect

Page 5: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect - depth imageA pattern of IR dots is projected

from the sensor

These dots are detected by the IR camera

The dots will change position based on how far the objects are from the source.

Page 6: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect - depth image

Page 7: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect - depth image

Shotton et al, CVPR(2011)

Page 8: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect - Skeleton

Page 10: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

The Kinect - applicationsRehabilitation.

Improvement of athletes performance.

Interactive surfaces.

3D modeling.

Augmented reality

Page 11: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

MethodologyData mining classification

◦It is the process of extracting valid, previously unseen or unknown, comprehensible information from large databases

◦Algorithms can involve artificial intelligence, machine learning, statistics, and database systems.

z-score normalization◦improve the accuracy and efficiency of

mining algorithms

Page 12: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsIn this study, were selected four popular data

mining classification method were selected :◦ Back Propagation Neural Network (BPNN)◦ Support Vector Machine (SVM)◦ Decision Tree◦ Naїve Bayes

To identify three human gestures:◦ Stand ◦ Sit down◦ Lie down

Page 13: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsProcess of Classification

Figure 1: Overview of the proposed system

Page 14: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsProcess of Classification

◦1,200 input vectors for each of the three human gesture classes in input data

◦3,600 input vectors (x,y,z) for each distance setting as shown (Stand, Sit down, Lie Down).

◦7,200 input vectors in total for both camera distance settings (2m and 3m)

◦1,200 vectors for both camera distance settings (2m and 3m)

◦The output data contain 3,600 vectors in total

Page 15: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsBackpropagation Neural

Network(BPNN)◦ BPNN is a multilayer feed forward neural

network, which uses backpropagation algorithm in its learning.

Page 16: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsSupport Vector Machine (SVM)

◦ In machine learning, support vector machines

(SVMs, also support vector networks)

are supervised learning models with associated

learning algorithms that analyze data and recognize

patterns

Page 17: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsDecision Tree (DT)

◦Decision Tree is used to classify data from class label

Page 18: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Classification MethodsNaïve Bayes (NB)

◦Is a statistical classification which predicts class membership based on conditional probabilities.

Page 19: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Human Gestures

Page 20: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

ResultsBPNN 100%SVM 99.75%DT 93.19%NB 81.94%

Page 21: Human Gesture Recognition Using Kinect Camera Presented by Carolina Vettorazzo and Diego Santo Orasa Patsadu, Chakarida Nukoolkit and Bunthit Watanapa

Questions???