39
Target Tracking a Non- Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

  • View
    231

  • Download
    11

Embed Size (px)

Citation preview

Page 1: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Tracking a Non-Linear Target Path

Using Kalman Predictive Algorithm

byJames Dennis Musick

Page 2: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Agenda

• Introduction

• Problem Definition

• Centroid Algorithm

• Kalman Filter

• Target Discrimination

• Conclusion

• Future Work

Page 3: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Introduction

• In the field of biomechanical research there is a subcategory that studies human movement or activity by video-based analysis

• Markers used– Optical

– RF

– Passive reflective

– Etc…

• Video based motion analysis

• 2D Analysis

• 3D analysis

• Golf swing example

Page 4: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Problem Definition

• In order to track the following have to be accomplished– Centroid calculation– Prediction– Discrimination

Page 5: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Problem Definition cont.

• Trials used– Walking Trial– Jumping Trial– Waving Wand Trial– Increasing complexity

Page 6: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm• Introduction

• Scanning scheme

Page 7: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• 640 x 480– ~ 307200 pixels

• 8-bit Gray-scale

• Block diagramThreshold X/Y

address location

Target Discrimination Buffer

Logic control and centroid calculation

Centroid ValueMemory

Page 8: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Threshold

Page 9: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• x/y addressing

Page 10: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Target Pixel Discrimination Buffer – x_sum, y_sum, LS_target, RS_target,

Bot_target, target_pixel_num

Page 11: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Logic Control and Centroid Calculation

}iin target pixels{

}iin target pixels{

}iin target pixels{

}iin target pixels{ ,),(

kk

kk

kk

kk

iii n

y

n

x

Cyx

Page 12: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Centroid Memory Buffer – Once a target is completed (defined as no pixels within

the search criteria at the row just below the target), then the centroid data is stored in a memory array until the data is read out at the end of the number of pictures that are being analyzed.

– The array would be structured in the following manner if there were three targets in each of 5 pictures:

• Target_Centroid_Array = (xy,Target #, Picture #) => (1:2, 1:3, 1:5).

Page 13: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Examples

Page 14: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Performance and Limitations – Three targets simultaneous– Total number

Page 15: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Centroid Algorithm cont.

• Measurement Uncertainty

• Correct (3.5,4) Correct (3.5,3)

• Blue missing (3.5,4) Red missing (3.8,3.17)• Red missing (3.64, 4.21)

Page 16: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter

• Introduction – State Space representation

kkk xTxx 1

Velocityxk

k

k

k

k

x

xT

x

x

10

1

1

1

Page 17: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont.

kkkk xT

xTxx 2

1 2 wk ux

T

2

2

wkkk uxTxx 1 Velocityxk onAcceleratixk

wkk BuApp 1

vkk DuCps ),0(~ 2ww Nu ),0(~ 2

vv Nu

Page 18: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont

Page 19: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont

Page 20: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont

• Target Models:– Noisy Acceleration model

Page 21: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont

• Target Models:– Noisy Jerk model

Page 22: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont

• Selection of update time:• T = 1

Page 23: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• b

Page 24: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 25: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 26: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 27: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 28: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 29: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Kalman Filter cont• Operation of the Kalman Filter

Page 30: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Introduction– Goal

Page 31: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Example

Page 32: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Example cont

Page 33: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Operation of algorithm

Page 34: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Operation of algorithm cont

Page 35: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Operation of algorithm cont

Jumping Trial

Page 36: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Operation of algorithm cont

Page 37: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Target Discrimination

• Occluded targets

Page 38: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Conclusion

• Centroid algorithm

• Kalman filter– Model

• Discrimination

Page 39: Target Tracking a Non-Linear Target Path Using Kalman Predictive Algorithm by James Dennis Musick

Future Work

• Hardware implementation

• 3D application

• Other biomechanical target discrimination (segmentation, etc.)

• Other tracking application (space, robotics, etc.)