Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
342 views

Box Counting Method

1) Fractals are patterns that are self-similar across different scales and are created by repeating a simple process over and over in an ongoing feedback loop. 2) The box-counting method calculates the fractal dimension of an object by counting how many boxes of different sizes are needed to cover the shape. A non-integer fractal dimension indicates qualities of different dimensional objects. 3) Specifically, the box-counting dimension is found by taking the log of the number of boxes needed versus the log of the box size, with the slope of the resulting line graph being the fractal dimension.

Uploaded by

Fatherg
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
342 views

Box Counting Method

1) Fractals are patterns that are self-similar across different scales and are created by repeating a simple process over and over in an ongoing feedback loop. 2) The box-counting method calculates the fractal dimension of an object by counting how many boxes of different sizes are needed to cover the shape. A non-integer fractal dimension indicates qualities of different dimensional objects. 3) Specifically, the box-counting dimension is found by taking the log of the number of boxes needed versus the log of the box size, with the slope of the resulting line graph being the fractal dimension.

Uploaded by

Fatherg
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Box-Counting Method

Self-similarity
Self similarity means that parts of the object are self similar (they contain parts of
themself). Self similar structures contain parts which are similar to themselves. It is
even possible that the whole structure can be found within the object (i.e.
Mandelbrot Set). Iteration and recursion describe the process of repeating steps.
Self-similarity is a property of the object, not of the steps used to build the object.

Example:
One of the basic properties of fractal images is the notion of self-similarity. This idea is
easy to explain using the Sierpinski triangle. Note that S may be decomposed into 3
congruent figures, each of which is exactly 1/2 the size of S! See above figure. That is
to say, if we magnify any of the 3 pieces of S shown in figure by a factor of 2, we obtain
an exact replica of S. That is, S consists of 3 self-similar copies of itself, each with
magnification factor 2.

Fractals
Fractals are infinitely complex patterns that are self-similar across
different scales. They are created by repeating a simple process
over and over in an ongoing feedback loop. So, fractals are natural
or artificial structures which offer a high level of scale invariance or
self-similarity.
Scale invariance relating to fractals means that the scale is not
important for the result. Structures which are scale invariant can be
observed from different distances but the object always looks the
same.

many natural phenomena are better described using a dimension


between two whole numbers. So while a straight line has a
dimension of one, a fractal curve will have a dimension between
one and two, depending on how much space it takes up as it twists
and curves.

Dimensions
By defining dimension in terms of the scaling properties of a shape, we
will come up with a quantitative way of describing fractals. This definition
of dimension is known as the self-similarity dimension, because it tells us
how many small self similar pieces of an object fit inside a large piece.
So what does a non-integer dimension mean, anyway?
There are several ways to think about this. First, a dimension between 1
and 2 means that the shape has some qualities of two-dimensional
objects and some of one dimensional objects. The snowflake is twodimensional in the sense that it resides in two-dimensional space. It rests
on the surface of a piece of paper, which is two-dimensional. However, as
the process of building the snowflake proceeds, the shape becomes more
and more edgy, in the sense that its perimeter grows and grows.

A fractal dimension is a ratio providing a statistical index


of complexity comparing how detail in a pattern (strictly speaking,
a fractal pattern) changes with the scale at which it is measured. It
has also been characterized as a measure of the spacefilling capacity of a pattern

Fractal Dimension
There are two definitions of fractal dimension:
Fractal dimension is a ratio providing a statistical index of
complexity comparing how detail in a pattern changes with the
scale at which it is measured.
Fractal dimension is a measure of the space-filling capacity of a
pattern that tells how a fractal scales differently than the space
it is embedded in.

A fractal dimension does not have to be


an integer.

Definition of Box-Counting Dimension


For different side lengths r we count N(r), the smallest number of boxes of side length r
needed to cover the shape. But, how does N(r) depend on r?
If the shape is 1-dimensional, such as the line segment, we have seen N(r) = 1/r
If the shape is 2-dimensional, such as the (filled-in) unit square, we have seen N(r) =
(1/r)2
If the shape is 3-dimensional, such as the (filled-in) unit cube, we expect N(r) = (1/r)3.
For more complicated shapes, the relation between N(r) and 1/r may not be so clear.
If we suspect that N(r) is approximately k(1/r)d, a power law relation, how can we find
d?
Taking Log of both sides of N(r) = k(1/r)d, we obtain
Log(N(r)) = Log(k) + Log((1/r)d) = dLog(1/r) + Log(k)
With the expectation that the approximation becomes better for smaller r. Solving for
d and taking the limit as r 0 gives db = limr 0Log(N(r))/Log(1/r) (Note as r 0 we
have 1/r , so Log(1/r) and Log(k)/Log(1/r) 0.)
If the limit exists, it is called the box-counting dimension, db, of the shape.
This limit may be slow to converge; an alternate approach is to notice
Log(N(r)) = dLog(1/r) + Log(k) is the equation of a straight line with slope d and yintercept Log(k).
So plotting Log(N(r)) vs Log(1/r), the points should lie approximately on a straight line
with slope db. This is the log-log approach to finding the box-counting dimension.

Box-Counting Dimension of a Gasket

Box-Counting Dimension Practice Problems

Box-Counting Dimension Practice Problems

Sierpinski carpet

First stage
Calculate next stages similarly

Option 1

Koch curve

Option 2
1.

3.

2.

Sierpinski Triangle

Stage 2
Calculate D for all stages one by one:

Stage 3

Stage 4

Stage 5

Similar Dimension

Stage 1

Stage 2

Stage 1-2 The shape can be decomposed into


N = 3 pieces, each scaled by a
factor of r = 1/2,
so ds = Log(3)/Log(2)

You might also like