Gardening
a box labelled Garden Unit. The output from both is mixed inside the box and comes out as a complex signal that is received by the Observer. This is illustrated in Diagram 5
adapted to a garden situation
simulation of a visual image with a number of components. These components may
be grouped into two cathegories: wanted and unwanted. The wanted ones are ascribed
to the infosource and the unwanted to the noise source. Both are mixed inside the
system - garden unit (the GU is actually a subsystem of the subsystem Garden as
presented in Diagram 1).
At this point we should remember that it is the Observer, not the Gardener,
the one who does the wanting (the one who decides which elements are wanted and which unwanted). However, we assume that in most situations there is agreement
in the judgements of both protagonists.
The signal coming from the infosource may arise from foliage,flowers,wood
and any decorative element inside the field of vision. These in turn generate perception
of shapes,textures, colors, etc. The signal coming from the noise source arises from
unwanted elements in the field of vision, from whatever the Observer may consider litter or garbage, from weeds, dried branches, etc (The informed reader may have noticed the similarity between the definition of Noise as any unwanted signal in the presence of wanted ones and that of a weed as"any unwanted plant in the presence
wanted ones" .
We propose two main cathegories of noise: garden noise and extraneous
(incidental) noise. The first comes from garden elements, i.e. dried branches,
yellow patches in lawns, unsightly shapes of bushes, unsettling combinations of shapes or colors, etc. All of these can be traced to the Gardener mistakes or inadequate treatment or placement of plants. The second class, extraneous noise, comes from
elements introduced into the garden by the public or others, i.e.all sorts of garbage or
refuse, assorted vandalism, etc. Though these do not arise from gardening peformance,
it is the responsability of the Gardener to eliminate or minimize this class of noise;
the Observer will not judge according to our classification but as to whether the whole
picture is pleasant or not.
We may ascribe variable intensities to the signals produced at the information and noise source. The output intensity is
a complex nonlinear summation of the signal intensities coming from the holons of
each source and as such is processed at the Observer's brain/mind.
[ I realize that the later sentence might appear a bit abstruse; I'm just tryng to be brief. Further clarifications or descriptions regarding visual perception and the use of Intensity instead of Power are
given in another place with appropiate links to learned places ]
Here we just note that that Signal and Noise arrive together to the Observer's
eyes and are processed simultaneously by his brain. Signal and noise intensity are
variables of our model and no one, least of all myself, is ascribing real existence
to them.
Within a given gardening unit the factors determining the intensity of the out-
coming signal may be discussed in terms of contributions from shapes, colours, textures, etc. form individual plants or parts of them. At a higher hierarchical level
in terms of contributions from combinations of colors and shapes and next according
to the composition of the image( the way the various elements included in the visual
field are arranged in space). It is our contention that the common Observer ( the one
that hasn't a trained gardening's eye) will not analyze his visual sensation according
to " annuals, perennials, geophytes, evergreen or decidous trees, flowering bushes,
etc. ", but in terms of the above mentioned shapes, colours or textures without
caring much for our garden cathegories. ( examples are given in annot.,empty)
In the case of the intensity of the outcoming noise signal, the situation may be
more neat or less diffuse. Noise was defined as an unwanted message in the presence
of a wanted one. An unwanted message will arise from any element that, acc. to the
Observer's judgment, doesn't belong to the set of elements that constitute the
gardening unit. Bottles, tin cans, newspapers, plastic bags, etc etc., clearly do not
belong, as neither will wilted flowers or broken branches. The intensity of the outcoming noise signal could be assumed to be directly proportional to the fraction
of the visual field occupied by those unwanted elements, with a proportionality
constant reflecting how unwanted a given element is.
Arbitrarly we chose an scale of 1 to 100 for both the signal from the infosource and that from the noise source. The numerical values of the SNR
resulting from various levels of infosource and noise signals are presented below:
1
2
3
4
5
6
7
8
9
10
20
50 40 30 20 10
33 27 20 13 7
25 20 15 10 5
20 16 12 8 4
17 13 10 7 3
14 11 8 6 3
12 10 7 5 2
11 9 7 4 2
10 8 6 4 2
5 4 3 2 1
increasing Noise signals so that for Noise values higher than 20 all the ratios are close
to zero. I have underlined in the Table SNR values of 10 to point out to a proposed
threshhold value over which SNR correlates with a pleasing effect and under which
correlate with displeasure. Until such correlation is supported the threshold value
is, of course arbitrarly placed; it may be justified on the basis that is 10 times smaller
than the maximum and that for values lower than 10 the Noise signals are of the
same order of magnitude than the infosource signal.
The behaviour of the ratio function is even more clear from Chart 1 where
the SNR values for three signal levels is presented.
The aplication of the model to some concrete garden situations is presented in
detail in the Section Model Application and here only the main points are presented.
A value of SNR close to 100 may be ascribed to a flowering bush, like a bouganvillea,
in full bloom covering the entire field of vision and with the minimum Noise signal.
(It should be kept in mind that the choice of values is not based on aesthetic grounds
but merely on the intensity of the signal). It may be seen from Table 1 that in such a case high levels of Nose are required to decrease the SNR to the threshold value.
The second example is that of an average lawn, again covering the vision field, with
a maximum SNR of 40 and for which rather low Noise values (i.e. N=4) decreases
them to the threshold. The third, and intermediate, example that will be discussed
will be that of flat beds of lantana mont. where values of the maximum SNR is
varied from 20 to 60 depending on the area fraction covered by flowers and on
their color.
Signals as function of
noise values.
brown bars: signal = 80
yellow bars: signal = 60
