Weed Detection over Between-row of Sugarcane Fields using Machine Vision with Shadow Robustness Technique for Variable Rate Herbicide Applicator

gurgleplayAI and Robotics

Oct 18, 2013 (4 years and 21 days ago)

141 views

International Conference on Science, Technology and Innovation for Sustainable Well
-
Being
(STISWB), 23
-
24 July 2009, Mahasarakham University, Thailand

Weed Detection

over

Between
-
row of
Sugarcane Fields

using

Machine Vision
with
Shadow

R
obustness
Technique
for
Variable Rate Herbicide Applicator

A.
Muankasem, S.
Thainimit

EE Department

Kasetsart University

50 phaholyothin Rd.,
Jatujak Bangkok 10900 Thailan
d

E
-
mail:

uripot@hotmail.com
,

fengsyt@ku.ac.th


R. Keinprasit

NECTEC 112 Thailand Science Park
Phahon Yothin Rd., Klong 1, Klong Luang, Pathumthani 12120,
Thailand

E
-
mail:

r
achaporn.keinprasit@nectec.or.th

Abstract


U
niformly herbicide rate is used as
a
conventional practice

in Thailand

for controlling
weeds

in
sugarcane

fields
.

Since w
eeds

usually

grow in certain areas

with
non
-
uniform
ly
distribution,
u
niform
herbicide
rat
e

approach
is not suitable and
non
-
sustainable

agricultural technique

both in
terms of economic an environmental aspect.


To address the
s
e

issues
, v
ariable herbicide rate (VHR)
was introduced
. The VHR composes of two main
component
s
, which are
weed moni
to
ring and real
-
time spraying. This paper
deal
s

with
a
development of a

fast and robust weed monitoring

system for
VHR using over between
-
row of sugarcane field
s
.

The proposed method is designed to work under
natural illumination condition. The near
-
ground

images are captured using a typical web camera
without any assistant light diffuser.

The proposed

weed monitoring

is a machine vision based
approach. The offset excessive green (OEC) technique is used to initially separate green weeds from
soil backgrou
nd.

Then, a
dditional

greenness under highlights and shadows are analyzed using
t
wo
new techniques.

The
techniques
exploit

variations among three triplets, which are red, green and blue

under bright and dull lighting condition
. The proposed method is ev
aluated using 250 field images
captured under variation conditions such as sunny and rainy day.
From our
experiments
, the
developed system

show
s

promising results. Weeds under different lighting conditions are reliably
detects.
The
method is fast and rob
ust, suitable for using
in natural field environment.



Keyword
:
machine vision,
greenness, threshold
level,

greenness under
shadows,

Variable rate
applicator