Robocup Vision Tracking with Xetal Processor

1
Robocup Vision Tracking with Xetal Processor
Edge and colour-based object detection
Sebastien Pierrot
Supervisors Harry Broers (CFT),
Anteneh Abbo, Richard Kleihorst (NATLAB)
2
Outline

• Introduction
• Vision System
• Object Tracking
• Future Work

3
Introduction (1)
Robocup
4
Introduction (2)
Machine Vision
5
Outline

• Introduction
• Vision System
• Robocup vision system
• Xetal Architecture
• Task division
• Object Tracking
• Future Work

6
Vision system (1)
Actual Robocup vision system
7
Vision system (2)
Xetal Architecture Block Schema
8
Vision system (3)
Repartition Tasks

9
Outline

• Introduction
• System Vision
• Object Tracking
• Color-based detection
• Edge detection
• Future Work

10
Color-based detection (1)
Object Tracking (1)
RGB
3-D RGB cube
11
Color-based detection (2)
Object Tracking (2)
YUV color space
Y Luminosity U,V Chromatic components
Y0.3R0.58G0.12B U0.17R-0.33G0.5B V0.5
R-0.42G0.08B
12
Object Tracking (3)
Color-based detection (3)
HSV color space
V Value S Saturation H Hue
V ( R G B )/3 S ( 1 - min(R,G,B)/ V ) H
0 (G-B)/? if max is R 1/3
(B-R)/? if max is G 2/3 (R-G)/? if
max is B ? is the (max-min) of the RGBs
13
Color-based detection (4)
Object Tracking (4)
HSI color space
I Intensity S Saturation H Hue
??/2 if GgtB ?3?/2 if GltB H1
if GB
14
Color-based detection (5)
Object Tracking (5)
V

• Segmentation examples

U
Linear
S
Constant
H
15
Object Tracking (6)
Color-based detection (6)
Orange YUV segmentation
16
Object Tracking (7)
Color-based detection (7)
Orange HSI segmentation
17
Color-based detection (8)
Object Tracking (8)

• Implementation discussion
• HSV
• 4 variable divisions
• HSI
• One variable division
• Arc tangent function
• Conclusion
• YUV linear segmentation for quicker processing
• HSI constant segmentation for tuning facility and
  better color density

18
Object Tracking (9)
Edge detection (1)

• Goal
• Strong intensity contrast detection
• Divide the image into areas
• corresponding to different objects
• Reducing image informations
• Computation
• Estimated with the maximum of the 1st derivative
  or with the zero crossing of the 2nd derivative

19
Object Tracking (10)
Edge detection (2)

• Sobel edge detector
• Approximation absolute gradient magnitude at
  each point in an input grayscale image
• ? a pair of 33 convolution kernels
• Advantage Simple implementation
• Drawback Sensible to the noise

20
Object Tracking (11)
Edge detection (3)

• Canny edge detector
• More sophisticated multi-stage process
• Advantages
• Simple thresholing
• Lower sensibility to the noise
• Large patterns 55,77
• Drawbacks
• Larger code program

21
Object Tracking (12)
Edge detection (4)

• Example 77 pattern elaboration
• Shifts
• Sum of intermediate
• Results
• Gx/y(0) G(0)G(2)
• Gx/y(1) G(1)G(3)

77 Kernel
G(2)
G(0)
22
Object Tracking (13)
Edge detection (5)
Sobel detector
Results
Canny detector
23
Outline

• Introduction
• System Vision
• Object Tracking
• Future Work
• Edge detection tuning
• Data compression

24
Compression

• Goal Reducing space information
• Proposed format
• Features
• Processor identification (PID)
• Statistic information delivering from serial
  processor
• Shifts for blank elimination