02-06-2012, 12:45 PM
AUTOMIZATION OF UMPIRING IN CRICKET USING FUZZY LOGIC
[attachment=23977]
INTRODUCTION
FUZZY LOGIC
Fuzzy logic was introduced by Prof. Lofti Zadeh in 1965 as a mathematical
way to represent uncertainty in everyday life. It can provide the means to represent
vague and fuzzy information, manipulate it, and to draw inferences from it. In
ordinary mathematics, information is of a crisp kind. It belongs to a set or it does
not. The choice of a yes-or-no answer is possible and usually applied, but
information could be lost in such a choice, as the degree of belonging is not taken
into consideration. A fuzzy model is the idea of a fuzzy set. A fuzzy set differs from
conventional (crisp) sets in its semi-permeable boundary membrane. Instead of a
characteristic function that has 2 states, inclusion (1) or exclusion (2), the fuzzy set
has a function that admits a degree of membership in the set from complete
exclusion (0) to absolute inclusion (1). The value zero is used to symbolize complete
non-membership, the value 1 is used to symbolize complete membership, and values
in between are used to symbolize intermediate degrees of membership. Membership
in a fuzzy subset should not be on a 0 or 1 basis, but rather on a 0 to 1 scale; that is
the membership should be an element of the interval [0,1].
WHY USE “FUZZY LOGIC” AS A TOOL?
Fuzzy Logic methodology, a branch of Artificial Intelligence is basically
characterized by three traits.
First, it does not consider whether something is true or false, but rather how
true the statement is.
Second, because it is similar to human reasoning, its implementation tends to
be based on natural language.
The third trait is that it is flexible and can model a complex, non-linear
system by using imprecise information.
The mode that has just been described provides an immediate output which
makes the fuzzy as the choicest tool for our problem.
FUZZIFICATION
A membership function acts on input variables usually from sensor data, in
what is known as a fuzzifier.
The fuzzifier output is referred to as a fuzzy-data value, which is the input to
the rule evaluator, which compares the fuzzy-data value to the value
established for each rule.
If one rule seems to be dominant explanation for the fuzzy-data value it is
considered to have ‘won’.
This news can be de-fuzzified for our real values.
FIELD SETUP
We install cameras at appropriate places, which are capable of providing the
fuzzy system with the input values angle, height and distance. These data are
then manipulated to determine the outcome of the delivery of a ball.
The cameras may be installed at the top of the stadium to provide a
panoramic view, which eases the task of measuring the height.
The cameras may also be installed around the ground to measure other
parameters like angle.
All the cameras will always be tracking the ball, so that we have a lot of
angles to look at the ball.
[attachment=23977]
INTRODUCTION
FUZZY LOGIC
Fuzzy logic was introduced by Prof. Lofti Zadeh in 1965 as a mathematical
way to represent uncertainty in everyday life. It can provide the means to represent
vague and fuzzy information, manipulate it, and to draw inferences from it. In
ordinary mathematics, information is of a crisp kind. It belongs to a set or it does
not. The choice of a yes-or-no answer is possible and usually applied, but
information could be lost in such a choice, as the degree of belonging is not taken
into consideration. A fuzzy model is the idea of a fuzzy set. A fuzzy set differs from
conventional (crisp) sets in its semi-permeable boundary membrane. Instead of a
characteristic function that has 2 states, inclusion (1) or exclusion (2), the fuzzy set
has a function that admits a degree of membership in the set from complete
exclusion (0) to absolute inclusion (1). The value zero is used to symbolize complete
non-membership, the value 1 is used to symbolize complete membership, and values
in between are used to symbolize intermediate degrees of membership. Membership
in a fuzzy subset should not be on a 0 or 1 basis, but rather on a 0 to 1 scale; that is
the membership should be an element of the interval [0,1].
WHY USE “FUZZY LOGIC” AS A TOOL?
Fuzzy Logic methodology, a branch of Artificial Intelligence is basically
characterized by three traits.
First, it does not consider whether something is true or false, but rather how
true the statement is.
Second, because it is similar to human reasoning, its implementation tends to
be based on natural language.
The third trait is that it is flexible and can model a complex, non-linear
system by using imprecise information.
The mode that has just been described provides an immediate output which
makes the fuzzy as the choicest tool for our problem.
FUZZIFICATION
A membership function acts on input variables usually from sensor data, in
what is known as a fuzzifier.
The fuzzifier output is referred to as a fuzzy-data value, which is the input to
the rule evaluator, which compares the fuzzy-data value to the value
established for each rule.
If one rule seems to be dominant explanation for the fuzzy-data value it is
considered to have ‘won’.
This news can be de-fuzzified for our real values.
FIELD SETUP
We install cameras at appropriate places, which are capable of providing the
fuzzy system with the input values angle, height and distance. These data are
then manipulated to determine the outcome of the delivery of a ball.
The cameras may be installed at the top of the stadium to provide a
panoramic view, which eases the task of measuring the height.
The cameras may also be installed around the ground to measure other
parameters like angle.
All the cameras will always be tracking the ball, so that we have a lot of
angles to look at the ball.