SlideShare a Scribd company logo
1 of 31
Laboratory Seminar
A Multi-Sensor based Uncut Crop Edge Detection Method
for Head-Feeding Combine Harvester
2013-06-26
Wonjae Cho*
Lab. of Field Robotics
Div. of Environmental Science and Technology
Graduate School of Agriculture
Kyoto University
Wonjae CHO, Michihisa IIDA, Masahiko SUGURI,
Ryohei MASUDA, Hiroki KURITA
0Kyoto University
Index
1. Introduction
2. Materials and Methods
3. Results and Discussion
4. Summary and Conclusions
0Kyoto University
Introduction
Ground plane
rd
cd
Uncut crop edge
RTK-GPS
LRF
GPS compass
Body information
(Travel speed, Header position, Etc.)
No. Sensor Figure
1 Laser Range Finder
: SICK Inc., LMS111
2 GPS Antenna
: Topcon LEGACY+
3 GPS Compass
: Hemisphere V110
0Kyoto University
Materials and Methods
1) Experimental setup – Navigation sensors
1
2
3
The head-feeding combine harvester
(VY50CLAM, Mitsubishi Agricultural Machinery Co., Ltd.)
Technical Specification of the sensors
0Kyoto University
Materials and Methods
1) Experimental setup – Control system
0Kyoto University
Materials and Methods
1) Experimental setup – Integrated sensor control platform (ISCP)
Sensor Interface
(Machine Vision, LRF)
ECU-KUx Interface
(Combine Body, RTK-GPS, GPS compass)
Driver
Layer
Platform
Layer
Steering
(Speed, Heading)
Sensor
Fusion
Image
Processing
Combine Body Control
(Header, Unloading Auger, Etc.)
Algorithm
Layer
Path
Planning
Uncut Crop Edge Detection 3D Mapping
Task Definition
(Harvest, Turn)
User
Interface
Layer
Sensors Display
(Machine Vision, LRF, RTK-GPS)
Combine Body
Display
3D Map
Display
Remote
Monitoring
The ISCP can be accessed via the following address
https://github.com/FiroKyoto/IntegratedSensorControlPlatform.git
0Kyoto University
Materials and Methods
2) 3D terrain mapping – The extrinsic parameters of the navigation sensors
Parameter Value
Tilt angle ( , °) 50.5
The height from ground ( ,m) 2.5
‘Longitudinal’ distance of
LRF from the origin of xy Plane ( ,m)
0.45
‘Transverse’ distance of
LRF from the origin of xy Plane ( ,m)
0.59
‘Longitudinal’ distance of
the divider from the LRF ( ,m)
0.6
‘Transverse’ distance of
the divider from the LRF ( ,m)
0.07

h
al
bl
cl
dl
0Kyoto University
Materials and Methods
2) 3D terrain mapping – Sensors data fusion
a. The polar coordinates system data ( , ) from LRF need to be transformed
into Cartesian coordinates system data (x, y, z).
b. The absolute position obtained from RTK-GPS should be transformed
into a position on the Transverse Mercator (TM) coordinates system.
c. Each of the transformed data of the navigation sensors should be applied
to equation and expressed as 3D points on space W.
cos
sin sin
sin cos
i i i
LMS
i i i i i
i i i i
x
p y
z h
 
  
  
   
       
      
_ _
_ _
_
cos sin 0
sin cos 0
0 0 0 1
MAP TM Heading Laser
i i i
TM LMS
x i c x i
TM LMS
y i a y i
LMS
z i
P P R P
P l p
P l p
p
 
 
 
    
           
        

0Kyoto University
Materials and Methods
2) 3D terrain mapping – The discrimination of the 3D points
LRF data distribution expressed on xz plane.
     
   
2 2
n h h k h k n h
k
n h n h
x x y y x x y y
d
x x y y
    

  
(Kimberling, 1998)
Parameter Signification
z of LRF data set becomes
maximum.
End points of LRF data set.
Perpendicular to the data
set.
Outermost boundary point
of the uncut crop area.
hp
np
kp
kd
0Kyoto University
Materials and Methods
2) 3D terrain mapping – Expressed 3D map using OpenGL
0Kyoto University
Materials and Methods
3) Uncut crop edge detection – Using RANSAC algorithm
RANdom SAmple Consensus (RANSAC) algorithm is a method that predicts the model parameters
from the original data that has severe measurement noise (Fischler and Bolles, 1981).
Pseudo code
Given:
data – a set of observed data points
model – a model that can be fitted to data points
m – the minimum number of data values required to fit the model
N – the maximum number of iterations allowed in the algorithm
t – a threshold value for determining when a data point fits a model
d – the number of close data values required to assert that model fits well to data
Return:
bestfit – model parameters which best fit the data (or nil if no good model is found)
iterations = 0
bestfit = nil
besterr = something really large
0Kyoto University
Materials and Methods
3) Uncut crop edge detection – Using RANSAC algorithm
while iteration < N {
% Hypothesis Step
maybeinliers = m randomly selected values from data
maybemodel = model parameters fitted to maybeinliers
alsoinliers = empty set
for every point in data not in maybeinliers {
if point fits maybemodel with an error smaller than t
add point to alsoinliers
}
if the number of elements in alsoinliers is > d {
% this implies that we may have found a good model
% now test how good it is
bettermodel = model parameters fitted to all points in maybeinliers and
alsoinliers
thiserr = a measure of how well model fits these points
if thiserr < besterr {
bestfit = bettermodel
besterr = thiserr
}
}
increment iterations
}
return bestfit
0Kyoto University
Materials and Methods
3) Uncut crop edge detection – Using RANSAC algorithm
a. data – Calculated set of the outermost boundary points ( ) of the uncut crop area,
configured in a dynamic array.
b. m – The number of sample data as 5.
c. t – The threshold value is set as 0.1.
d. N – The number of repeats of RANSAC algorithm can be calculated from the equation.
[Where]
The probability ( ), where at least one sample set includes valid data, as 0.99.
Probability ( ) of the data validity as 0.5.
kp
log(1 )
log(1 )m
p
N




p

0Kyoto University
Materials and Methods
3) Uncut crop edge detection – Expressed uncut crop edge on W space
0Kyoto University
Results and Discussion
1) Results – Experimental conditions
Traveling path of the head-feeding combine harvester
on Google Maps.
Date 2012-09-20
Location Nantan city, Kyoto Prefecture, Japan
Field Rice paddy field
Harvester
Model
VY50CLAM
Harvester
Control
By an operator (human)
Traveling
Speed [m/s]
0.6
Movement
Direction
Counter-clockwise
Data set Four dataset (A, B, C, D)
0Kyoto University
Results and Discussion
1) Results – The result movie Clip
The result movie clip can be accessed via the following address
http://youtu.be/juyYuQfvgkk
0Kyoto University
Results and Discussion
1) Results – Uncut crop edge extracted by using RANSAC
Section (A) Section (B)
Section (C) Section (d)
0Kyoto University
Results and Discussion
1) Results – By the proposed method
3D terrain map of the experiment field.
Dataset Movement
Direction
Average of
Crop Height
[m]
Lateral
Offset
[m]
A South 0.514 0.294
B East 0.503 0.139
C North 0.549 0.067
D West 0.580 0.119
Average 0.537 0.154
Average processing speed of 35 ms
0Kyoto University
Results and Discussion
2) Discussion
0Kyoto University
Summary and Conclusions
a. The proposed method was able to generate 3D maps of the terrain to be
harvested at a processing speed of 35 ms by using navigation sensors.
b. The averages of lateral offset value of uncut crop edge and crop height,
extracted by the proposed method, were 0.154 m and 0.537 m, respectively.
c. While the proposed method of the study was robust in extracting the uncut
crop edge in general.
d. Nevertheless, it’s performance displayed a tendency to decline when the target
path was obscured by the lodging of rice plants.
e. Hence, in order to enhance the performance of the proposed method, the
mounting position of LRF needs to be modified to allow the accurate scanning
of the target path, together with the addition of an algorithm that can adjust
for the error in results.
Remote Monitoring
0Kyoto University
Results and Discussion
1) Results – The result movie Clip
The result movie clip can be accessed via the following address
http://youtu.be/eL6cX5_9o-0
小麦の収穫実験
2013-06-10 (月)
0Kyoto University
Introduction
Header height of head feeding combine
In case header height is lower than ground
In case header height is too high
0Kyoto University
Materials and Methods
1) Header control – Range designation
Travel Speed [m/s] = V
Distance [m] = S = Vt
Elapsed time [sec] = t
Start position Destination position
0Kyoto University
Materials and Methods
1) Header control – Relationship between header height and potentiometer
y = 596.2x + 365.7
R² = 0.977
0
100
200
300
400
500
600
700
800
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Potentiometer(ADvalue)
Hedaer Height (m)
p1
pNpk
d
l1
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
-3 -2 -1 0 1 2 3 4
Height(m)
Lateral direction (m)
Ground Scope (1m)
0Kyoto University
Results and Discussion
1) Results – Experimental conditions
Traveling path of the head-feeding combine harvester
on Google Maps.
Date 2013-06-10
Location 奈良県大桜井市西地区
Field Wheat field
Harvester
Model
VY446
Harvester
Control
By an Autonomous mode
Traveling
Speed [m/s]
0.5
Movement
Direction
Counter-clockwise
Data set One dataset
0Kyoto University
Results and Discussion
1) Results – The result movie Clip
The result movie clip can be accessed via the following address
0Kyoto University
Results and Discussion
2) Discussion
Rank weeds
Low density
0Kyoto University
Future work
a. Auto-Steering algorithm
b. Turn algorithm
c. Advanced header control algorithm
d. Autonomous guidance system for head-feeding combine harvester
e. Advanced Remote Monitoring Application
: Participation of Telerobotics Summer School 2013 (Support by IEEE)
 July 8 - 12 2013 in Yokohama, Hiyoshi Campus, Keio University, Japan
 https://sites.google.com/site/teleroboticssummerschool2013/
f. Migration Integrated Sensor Control Platform (ISCP)
: Visual C# to C++ (native C++ or visual C++)
0Kyoto University
Contact us
Thank you for listening to my presentation.
If you have any questions, please electronic mail me!
 cho@elam.kais.kyoto-u.ac.jp

More Related Content

What's hot

my projects
my projectsmy projects
my projects
fferrieu
 
Poster: Monash Research Month 2007
Poster: Monash Research Month 2007Poster: Monash Research Month 2007
Poster: Monash Research Month 2007
Mahfuzul Haque
 
Slope Task 7.3 – Trials and validation cycle (by BOKU)
Slope Task 7.3 – Trials and validation cycle (by BOKU) Slope Task 7.3 – Trials and validation cycle (by BOKU)
Slope Task 7.3 – Trials and validation cycle (by BOKU)
SLOPE Project
 
Alexandra Karamitrou.ppt
Alexandra Karamitrou.pptAlexandra Karamitrou.ppt
Alexandra Karamitrou.ppt
grssieee
 
24th JCAART 2009 Conference
24th JCAART 2009 Conference24th JCAART 2009 Conference
24th JCAART 2009 Conference
suvonvorn
 
Multi Inertial Measurement Units (MIMU) Platforms: Designs & Applications
Multi Inertial Measurement Units (MIMU)  Platforms: Designs & ApplicationsMulti Inertial Measurement Units (MIMU)  Platforms: Designs & Applications
Multi Inertial Measurement Units (MIMU) Platforms: Designs & Applications
oblu.io
 

What's hot (18)

my projects
my projectsmy projects
my projects
 
"The effect of angles and distance on image-based three-dimensional reconstru...
"The effect of angles and distance on image-based three-dimensional reconstru..."The effect of angles and distance on image-based three-dimensional reconstru...
"The effect of angles and distance on image-based three-dimensional reconstru...
 
Poster: Monash Research Month 2007
Poster: Monash Research Month 2007Poster: Monash Research Month 2007
Poster: Monash Research Month 2007
 
Paszto - Fractal dimension of the Czech road network
Paszto - Fractal dimension of the Czech road networkPaszto - Fractal dimension of the Czech road network
Paszto - Fractal dimension of the Czech road network
 
Svm based cbir of breast masses on mammograms
Svm based cbir of breast masses on mammogramsSvm based cbir of breast masses on mammograms
Svm based cbir of breast masses on mammograms
 
Slope Task 7.3 – Trials and validation cycle (by BOKU)
Slope Task 7.3 – Trials and validation cycle (by BOKU) Slope Task 7.3 – Trials and validation cycle (by BOKU)
Slope Task 7.3 – Trials and validation cycle (by BOKU)
 
MediaEval 2018: Fine grained sport action recognition: Application to table t...
MediaEval 2018: Fine grained sport action recognition: Application to table t...MediaEval 2018: Fine grained sport action recognition: Application to table t...
MediaEval 2018: Fine grained sport action recognition: Application to table t...
 
Interphase Cells Removal from Metaphase Chromosome Images Based on Meta-Heuri...
Interphase Cells Removal from Metaphase Chromosome Images Based on Meta-Heuri...Interphase Cells Removal from Metaphase Chromosome Images Based on Meta-Heuri...
Interphase Cells Removal from Metaphase Chromosome Images Based on Meta-Heuri...
 
Alexandra Karamitrou.ppt
Alexandra Karamitrou.pptAlexandra Karamitrou.ppt
Alexandra Karamitrou.ppt
 
Image verification assistant. By Polychronis Charitidis of CERTH-ITI
Image verification assistant. By Polychronis Charitidis of CERTH-ITIImage verification assistant. By Polychronis Charitidis of CERTH-ITI
Image verification assistant. By Polychronis Charitidis of CERTH-ITI
 
OGC's 20th Anniversary Dinner Slide Set
OGC's 20th Anniversary Dinner Slide SetOGC's 20th Anniversary Dinner Slide Set
OGC's 20th Anniversary Dinner Slide Set
 
Anton Korsakov - Determination of an unmanned mobile object orientation by na...
Anton Korsakov - Determination of an unmanned mobile object orientation by na...Anton Korsakov - Determination of an unmanned mobile object orientation by na...
Anton Korsakov - Determination of an unmanned mobile object orientation by na...
 
MINNI_ENEA_mihaela_mircea
MINNI_ENEA_mihaela_mirceaMINNI_ENEA_mihaela_mircea
MINNI_ENEA_mihaela_mircea
 
24th JCAART 2009 Conference
24th JCAART 2009 Conference24th JCAART 2009 Conference
24th JCAART 2009 Conference
 
Airport Runway Detection Based On ANN Algorithm
Airport Runway Detection Based On ANN AlgorithmAirport Runway Detection Based On ANN Algorithm
Airport Runway Detection Based On ANN Algorithm
 
Multi Inertial Measurement Units (MIMU) Platforms: Designs & Applications
Multi Inertial Measurement Units (MIMU)  Platforms: Designs & ApplicationsMulti Inertial Measurement Units (MIMU)  Platforms: Designs & Applications
Multi Inertial Measurement Units (MIMU) Platforms: Designs & Applications
 
Gabriele Candela_Image-based 3d reconstruction of a bamboo-steel spatial trus...
Gabriele Candela_Image-based 3d reconstruction of a bamboo-steel spatial trus...Gabriele Candela_Image-based 3d reconstruction of a bamboo-steel spatial trus...
Gabriele Candela_Image-based 3d reconstruction of a bamboo-steel spatial trus...
 
2014 IEEE Int. Symposium on System Integration (SII) : Project on Development...
2014 IEEE Int. Symposium on System Integration (SII) : Project on Development...2014 IEEE Int. Symposium on System Integration (SII) : Project on Development...
2014 IEEE Int. Symposium on System Integration (SII) : Project on Development...
 

Viewers also liked

ソフトウェアプラットフォームを用いたコンバインロボットの制御
ソフトウェアプラットフォームを用いたコンバインロボットの制御ソフトウェアプラットフォームを用いたコンバインロボットの制御
ソフトウェアプラットフォームを用いたコンバインロボットの制御
Institute of Agricultural Machinery, NARO
 
Isabella 4 Ano B
Isabella   4 Ano  BIsabella   4 Ano  B
Isabella 4 Ano B
guestad57c9
 
Detection of genetically modified crops in animal feed in Serbia.PDF
Detection of genetically modified crops in animal feed in Serbia.PDFDetection of genetically modified crops in animal feed in Serbia.PDF
Detection of genetically modified crops in animal feed in Serbia.PDF
Gordana Zdjelar
 
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
ASHOKKUMAR RAMAR
 
Delegation & empowerment presentation
Delegation & empowerment presentationDelegation & empowerment presentation
Delegation & empowerment presentation
Kamal Prasad
 

Viewers also liked (20)

Therft alert system
Therft alert systemTherft alert system
Therft alert system
 
ソフトウェアプラットフォームを用いたコンバインロボットの制御
ソフトウェアプラットフォームを用いたコンバインロボットの制御ソフトウェアプラットフォームを用いたコンバインロボットの制御
ソフトウェアプラットフォームを用いたコンバインロボットの制御
 
Empowerment
EmpowermentEmpowerment
Empowerment
 
Meiosis
MeiosisMeiosis
Meiosis
 
072a - Who is the head of the Ordine Gesù Redentore
072a - Who is the head of the Ordine Gesù Redentore072a - Who is the head of the Ordine Gesù Redentore
072a - Who is the head of the Ordine Gesù Redentore
 
OxyContin Teachback
OxyContin TeachbackOxyContin Teachback
OxyContin Teachback
 
Isabella 4 Ano B
Isabella   4 Ano  BIsabella   4 Ano  B
Isabella 4 Ano B
 
Predictive analytics roadshow
Predictive analytics roadshowPredictive analytics roadshow
Predictive analytics roadshow
 
CARÁCTER DE CRISTO 1
CARÁCTER DE CRISTO 1CARÁCTER DE CRISTO 1
CARÁCTER DE CRISTO 1
 
Music video analysis
Music video analysisMusic video analysis
Music video analysis
 
Detection of genetically modified crops in animal feed in Serbia.PDF
Detection of genetically modified crops in animal feed in Serbia.PDFDetection of genetically modified crops in animal feed in Serbia.PDF
Detection of genetically modified crops in animal feed in Serbia.PDF
 
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
IEEE PROJECTS ABSTRACT-A zigbee based wireless sensor network for monitoring ...
 
my anti t.a.s
my anti t.a.smy anti t.a.s
my anti t.a.s
 
herramientas para construir preguntas tipo pruebas saber.
herramientas para construir preguntas tipo pruebas saber.herramientas para construir preguntas tipo pruebas saber.
herramientas para construir preguntas tipo pruebas saber.
 
Crop monitoring system
Crop monitoring systemCrop monitoring system
Crop monitoring system
 
Impact of government school to-land agricultural project on rural development...
Impact of government school to-land agricultural project on rural development...Impact of government school to-land agricultural project on rural development...
Impact of government school to-land agricultural project on rural development...
 
Projeto de Integração do Rio São Francisco
Projeto de Integração do Rio São Francisco Projeto de Integração do Rio São Francisco
Projeto de Integração do Rio São Francisco
 
Implementation of agricultural automation system using web &amp; gsm technolo...
Implementation of agricultural automation system using web &amp; gsm technolo...Implementation of agricultural automation system using web &amp; gsm technolo...
Implementation of agricultural automation system using web &amp; gsm technolo...
 
Agriculture
AgricultureAgriculture
Agriculture
 
Delegation & empowerment presentation
Delegation & empowerment presentationDelegation & empowerment presentation
Delegation & empowerment presentation
 

Similar to A multi-sensor based uncut crop edge detection method for head-feeding combine harvester

ERC_EGUE_FINAL_Aug 12_PJB
ERC_EGUE_FINAL_Aug 12_PJBERC_EGUE_FINAL_Aug 12_PJB
ERC_EGUE_FINAL_Aug 12_PJB
Paul Brodbeck
 

Similar to A multi-sensor based uncut crop edge detection method for head-feeding combine harvester (20)

Adaptive Hyper-Parameter Tuning for Black-box LiDAR Odometry [IROS2021]
Adaptive Hyper-Parameter Tuning for Black-box LiDAR Odometry [IROS2021]Adaptive Hyper-Parameter Tuning for Black-box LiDAR Odometry [IROS2021]
Adaptive Hyper-Parameter Tuning for Black-box LiDAR Odometry [IROS2021]
 
IRJET- Proposed System for Animal Recognition using Image Processing
IRJET-  	  Proposed System for Animal Recognition using Image ProcessingIRJET-  	  Proposed System for Animal Recognition using Image Processing
IRJET- Proposed System for Animal Recognition using Image Processing
 
IRJET- Surveillance of Object Motion Detection and Caution System using B...
IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...IRJET-  	  Surveillance of Object Motion Detection and Caution System using B...
IRJET- Surveillance of Object Motion Detection and Caution System using B...
 
BPSO&1-NN algorithm-based variable selection for power system stability ident...
BPSO&1-NN algorithm-based variable selection for power system stability ident...BPSO&1-NN algorithm-based variable selection for power system stability ident...
BPSO&1-NN algorithm-based variable selection for power system stability ident...
 
IRJET- Expert Independent Bayesian Data Fusion and Decision Making Model for ...
IRJET- Expert Independent Bayesian Data Fusion and Decision Making Model for ...IRJET- Expert Independent Bayesian Data Fusion and Decision Making Model for ...
IRJET- Expert Independent Bayesian Data Fusion and Decision Making Model for ...
 
B.Tech Thesis
B.Tech ThesisB.Tech Thesis
B.Tech Thesis
 
1708027_1708037.pptx
1708027_1708037.pptx1708027_1708037.pptx
1708027_1708037.pptx
 
rbs - presentation about applications of machine learning.
rbs - presentation about applications of machine learning.rbs - presentation about applications of machine learning.
rbs - presentation about applications of machine learning.
 
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUESNEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
 
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUESNEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUES
 
IRJET- Different Data Mining Techniques for Weather Prediction
IRJET-  	  Different Data Mining Techniques for Weather PredictionIRJET-  	  Different Data Mining Techniques for Weather Prediction
IRJET- Different Data Mining Techniques for Weather Prediction
 
IRJET- Survey Paper on Human Following Robot
IRJET- Survey Paper on Human Following RobotIRJET- Survey Paper on Human Following Robot
IRJET- Survey Paper on Human Following Robot
 
Design of Kalman filter for Airborne Applications
Design of Kalman filter for Airborne ApplicationsDesign of Kalman filter for Airborne Applications
Design of Kalman filter for Airborne Applications
 
A Review Paper on Elderly Fall Detection
A Review Paper on Elderly Fall DetectionA Review Paper on Elderly Fall Detection
A Review Paper on Elderly Fall Detection
 
Object Tracking By Online Discriminative Feature Selection Algorithm
Object Tracking By Online Discriminative Feature Selection AlgorithmObject Tracking By Online Discriminative Feature Selection Algorithm
Object Tracking By Online Discriminative Feature Selection Algorithm
 
Udacity-Didi Challenge Finalists
Udacity-Didi Challenge FinalistsUdacity-Didi Challenge Finalists
Udacity-Didi Challenge Finalists
 
ERC_EGUE_FINAL_Aug 12_PJB
ERC_EGUE_FINAL_Aug 12_PJBERC_EGUE_FINAL_Aug 12_PJB
ERC_EGUE_FINAL_Aug 12_PJB
 
C1802022430
C1802022430C1802022430
C1802022430
 
Failure Prediction Using Interaction between Parallel Links of FA Equipment
Failure Prediction Using Interaction between Parallel Links of FA EquipmentFailure Prediction Using Interaction between Parallel Links of FA Equipment
Failure Prediction Using Interaction between Parallel Links of FA Equipment
 
A Novel Feature Selection with Annealing For Computer Vision And Big Data Lea...
A Novel Feature Selection with Annealing For Computer Vision And Big Data Lea...A Novel Feature Selection with Annealing For Computer Vision And Big Data Lea...
A Novel Feature Selection with Annealing For Computer Vision And Big Data Lea...
 

More from Institute of Agricultural Machinery, NARO

More from Institute of Agricultural Machinery, NARO (7)

ML基本からResNetまで
ML基本からResNetまでML基本からResNetまで
ML基本からResNetまで
 
Feature Pyramid Network, FPN
Feature Pyramid Network, FPNFeature Pyramid Network, FPN
Feature Pyramid Network, FPN
 
Review: Deep contextualized word representations
Review: Deep contextualized word representationsReview: Deep contextualized word representations
Review: Deep contextualized word representations
 
LeNet & GoogLeNet
LeNet & GoogLeNetLeNet & GoogLeNet
LeNet & GoogLeNet
 
The Integrated Sensor Control Platform for Head-feeding Combine Harvester
The Integrated Sensor Control Platform for Head-feeding Combine HarvesterThe Integrated Sensor Control Platform for Head-feeding Combine Harvester
The Integrated Sensor Control Platform for Head-feeding Combine Harvester
 
ロボットコンバインのためのソフトプラットホーム開発
ロボットコンバインのためのソフトプラットホーム開発ロボットコンバインのためのソフトプラットホーム開発
ロボットコンバインのためのソフトプラットホーム開発
 
100519 ver final
100519 ver final100519 ver final
100519 ver final
 

Recently uploaded

Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
vu2urc
 

Recently uploaded (20)

TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Advantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your BusinessAdvantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your Business
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...
 
Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century education
 
GenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdfGenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdf
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?
 

A multi-sensor based uncut crop edge detection method for head-feeding combine harvester

  • 1. Laboratory Seminar A Multi-Sensor based Uncut Crop Edge Detection Method for Head-Feeding Combine Harvester 2013-06-26 Wonjae Cho* Lab. of Field Robotics Div. of Environmental Science and Technology Graduate School of Agriculture Kyoto University Wonjae CHO, Michihisa IIDA, Masahiko SUGURI, Ryohei MASUDA, Hiroki KURITA
  • 2. 0Kyoto University Index 1. Introduction 2. Materials and Methods 3. Results and Discussion 4. Summary and Conclusions
  • 3. 0Kyoto University Introduction Ground plane rd cd Uncut crop edge RTK-GPS LRF GPS compass Body information (Travel speed, Header position, Etc.)
  • 4. No. Sensor Figure 1 Laser Range Finder : SICK Inc., LMS111 2 GPS Antenna : Topcon LEGACY+ 3 GPS Compass : Hemisphere V110 0Kyoto University Materials and Methods 1) Experimental setup – Navigation sensors 1 2 3 The head-feeding combine harvester (VY50CLAM, Mitsubishi Agricultural Machinery Co., Ltd.) Technical Specification of the sensors
  • 5. 0Kyoto University Materials and Methods 1) Experimental setup – Control system
  • 6. 0Kyoto University Materials and Methods 1) Experimental setup – Integrated sensor control platform (ISCP) Sensor Interface (Machine Vision, LRF) ECU-KUx Interface (Combine Body, RTK-GPS, GPS compass) Driver Layer Platform Layer Steering (Speed, Heading) Sensor Fusion Image Processing Combine Body Control (Header, Unloading Auger, Etc.) Algorithm Layer Path Planning Uncut Crop Edge Detection 3D Mapping Task Definition (Harvest, Turn) User Interface Layer Sensors Display (Machine Vision, LRF, RTK-GPS) Combine Body Display 3D Map Display Remote Monitoring The ISCP can be accessed via the following address https://github.com/FiroKyoto/IntegratedSensorControlPlatform.git
  • 7. 0Kyoto University Materials and Methods 2) 3D terrain mapping – The extrinsic parameters of the navigation sensors Parameter Value Tilt angle ( , °) 50.5 The height from ground ( ,m) 2.5 ‘Longitudinal’ distance of LRF from the origin of xy Plane ( ,m) 0.45 ‘Transverse’ distance of LRF from the origin of xy Plane ( ,m) 0.59 ‘Longitudinal’ distance of the divider from the LRF ( ,m) 0.6 ‘Transverse’ distance of the divider from the LRF ( ,m) 0.07  h al bl cl dl
  • 8. 0Kyoto University Materials and Methods 2) 3D terrain mapping – Sensors data fusion a. The polar coordinates system data ( , ) from LRF need to be transformed into Cartesian coordinates system data (x, y, z). b. The absolute position obtained from RTK-GPS should be transformed into a position on the Transverse Mercator (TM) coordinates system. c. Each of the transformed data of the navigation sensors should be applied to equation and expressed as 3D points on space W. cos sin sin sin cos i i i LMS i i i i i i i i i x p y z h                            _ _ _ _ _ cos sin 0 sin cos 0 0 0 0 1 MAP TM Heading Laser i i i TM LMS x i c x i TM LMS y i a y i LMS z i P P R P P l p P l p p                                 
  • 9. 0Kyoto University Materials and Methods 2) 3D terrain mapping – The discrimination of the 3D points LRF data distribution expressed on xz plane.           2 2 n h h k h k n h k n h n h x x y y x x y y d x x y y          (Kimberling, 1998) Parameter Signification z of LRF data set becomes maximum. End points of LRF data set. Perpendicular to the data set. Outermost boundary point of the uncut crop area. hp np kp kd
  • 10. 0Kyoto University Materials and Methods 2) 3D terrain mapping – Expressed 3D map using OpenGL
  • 11. 0Kyoto University Materials and Methods 3) Uncut crop edge detection – Using RANSAC algorithm RANdom SAmple Consensus (RANSAC) algorithm is a method that predicts the model parameters from the original data that has severe measurement noise (Fischler and Bolles, 1981). Pseudo code Given: data – a set of observed data points model – a model that can be fitted to data points m – the minimum number of data values required to fit the model N – the maximum number of iterations allowed in the algorithm t – a threshold value for determining when a data point fits a model d – the number of close data values required to assert that model fits well to data Return: bestfit – model parameters which best fit the data (or nil if no good model is found) iterations = 0 bestfit = nil besterr = something really large
  • 12. 0Kyoto University Materials and Methods 3) Uncut crop edge detection – Using RANSAC algorithm while iteration < N { % Hypothesis Step maybeinliers = m randomly selected values from data maybemodel = model parameters fitted to maybeinliers alsoinliers = empty set for every point in data not in maybeinliers { if point fits maybemodel with an error smaller than t add point to alsoinliers } if the number of elements in alsoinliers is > d { % this implies that we may have found a good model % now test how good it is bettermodel = model parameters fitted to all points in maybeinliers and alsoinliers thiserr = a measure of how well model fits these points if thiserr < besterr { bestfit = bettermodel besterr = thiserr } } increment iterations } return bestfit
  • 13. 0Kyoto University Materials and Methods 3) Uncut crop edge detection – Using RANSAC algorithm a. data – Calculated set of the outermost boundary points ( ) of the uncut crop area, configured in a dynamic array. b. m – The number of sample data as 5. c. t – The threshold value is set as 0.1. d. N – The number of repeats of RANSAC algorithm can be calculated from the equation. [Where] The probability ( ), where at least one sample set includes valid data, as 0.99. Probability ( ) of the data validity as 0.5. kp log(1 ) log(1 )m p N     p 
  • 14. 0Kyoto University Materials and Methods 3) Uncut crop edge detection – Expressed uncut crop edge on W space
  • 15. 0Kyoto University Results and Discussion 1) Results – Experimental conditions Traveling path of the head-feeding combine harvester on Google Maps. Date 2012-09-20 Location Nantan city, Kyoto Prefecture, Japan Field Rice paddy field Harvester Model VY50CLAM Harvester Control By an operator (human) Traveling Speed [m/s] 0.6 Movement Direction Counter-clockwise Data set Four dataset (A, B, C, D)
  • 16. 0Kyoto University Results and Discussion 1) Results – The result movie Clip The result movie clip can be accessed via the following address http://youtu.be/juyYuQfvgkk
  • 17. 0Kyoto University Results and Discussion 1) Results – Uncut crop edge extracted by using RANSAC Section (A) Section (B) Section (C) Section (d)
  • 18. 0Kyoto University Results and Discussion 1) Results – By the proposed method 3D terrain map of the experiment field. Dataset Movement Direction Average of Crop Height [m] Lateral Offset [m] A South 0.514 0.294 B East 0.503 0.139 C North 0.549 0.067 D West 0.580 0.119 Average 0.537 0.154 Average processing speed of 35 ms
  • 19. 0Kyoto University Results and Discussion 2) Discussion
  • 20. 0Kyoto University Summary and Conclusions a. The proposed method was able to generate 3D maps of the terrain to be harvested at a processing speed of 35 ms by using navigation sensors. b. The averages of lateral offset value of uncut crop edge and crop height, extracted by the proposed method, were 0.154 m and 0.537 m, respectively. c. While the proposed method of the study was robust in extracting the uncut crop edge in general. d. Nevertheless, it’s performance displayed a tendency to decline when the target path was obscured by the lodging of rice plants. e. Hence, in order to enhance the performance of the proposed method, the mounting position of LRF needs to be modified to allow the accurate scanning of the target path, together with the addition of an algorithm that can adjust for the error in results.
  • 22. 0Kyoto University Results and Discussion 1) Results – The result movie Clip The result movie clip can be accessed via the following address http://youtu.be/eL6cX5_9o-0
  • 24. 0Kyoto University Introduction Header height of head feeding combine In case header height is lower than ground In case header height is too high
  • 25. 0Kyoto University Materials and Methods 1) Header control – Range designation Travel Speed [m/s] = V Distance [m] = S = Vt Elapsed time [sec] = t Start position Destination position
  • 26. 0Kyoto University Materials and Methods 1) Header control – Relationship between header height and potentiometer y = 596.2x + 365.7 R² = 0.977 0 100 200 300 400 500 600 700 800 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Potentiometer(ADvalue) Hedaer Height (m) p1 pNpk d l1 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 -3 -2 -1 0 1 2 3 4 Height(m) Lateral direction (m) Ground Scope (1m)
  • 27. 0Kyoto University Results and Discussion 1) Results – Experimental conditions Traveling path of the head-feeding combine harvester on Google Maps. Date 2013-06-10 Location 奈良県大桜井市西地区 Field Wheat field Harvester Model VY446 Harvester Control By an Autonomous mode Traveling Speed [m/s] 0.5 Movement Direction Counter-clockwise Data set One dataset
  • 28. 0Kyoto University Results and Discussion 1) Results – The result movie Clip The result movie clip can be accessed via the following address
  • 29. 0Kyoto University Results and Discussion 2) Discussion Rank weeds Low density
  • 30. 0Kyoto University Future work a. Auto-Steering algorithm b. Turn algorithm c. Advanced header control algorithm d. Autonomous guidance system for head-feeding combine harvester e. Advanced Remote Monitoring Application : Participation of Telerobotics Summer School 2013 (Support by IEEE)  July 8 - 12 2013 in Yokohama, Hiyoshi Campus, Keio University, Japan  https://sites.google.com/site/teleroboticssummerschool2013/ f. Migration Integrated Sensor Control Platform (ISCP) : Visual C# to C++ (native C++ or visual C++)
  • 31. 0Kyoto University Contact us Thank you for listening to my presentation. If you have any questions, please electronic mail me!  cho@elam.kais.kyoto-u.ac.jp