Database Reference
In-Depth Information
The Life Forms system consists of many variables designed to represent key positions
as frame series, in accordance with the number of frames per second that the user stipulates.
Among them are the fi gure's physical dimensions, presentation mode, stage measurements, user
standpoint, location in relation to stage location, position relative to other fi gures, etc.
The human fi gure in the Life Forms application is composed of 22 limbs. The physical
position is characterized by indicating the location of a reference point in space, and the
position of each of the 22 limbs. The reference point represents three coordinates (X,Y,Z)
in the spatial region of the stage, applying to the pelvic region of the fi gure, or alternatively,
to the fi gure's lowest point, i.e., the limb or part of a limb that possesses the smallest Z
component. Furthermore, each of the limb positions is defi ned by means of three angles on
the Y, X and Z axes, indicative of the Pitch, Roll and Spin angles.
The Life Forms application uses 69 basic variables to represent the position of an im-
age consisting of 22 limbs, that is,
69 fl oating point variables
whose storage volume varies
according to the computer platform used.
Life Forms
Storage Volume
Table 3 presents Life Forms variables in a Macintosh environment. The volume of raw
data relating to fi gure animation variables (not including the fi gure environment) is compared
to the volume of raw data required by our binary model by means of sets of 7 bits per joint
per time unit of the database cube.
Lines 4 - 7 show the components for defi ning a position. Their total volume range
is 424 - 442 bytes, i.e., 3392 - 3536 bits. It should be noted that these volumes underwent
compression by the program. Even if we do not include all the possible location variation,
the volumes of all the 69 fl oating parameters for representation of the fi gure's limbs in the
position can alone reach 4416 bits (69*64 bits per fl oating variable). Adding 56 bytes to
represent the rest of the components (4 - 6) results in
4864
total bits for image position.
Note that the storage volume of the Life Forms application does not vary in accordance
with the situations indicated above; it depends on the number of positions defi ned, rather
than on the sampling frequency, motion type, or number of limbs that actually perform the
movement. The advantages of the binary model become even more evident when we compare
the volume consumed for motion storage, whose representation by means of the key frames
approach demands the defi nition and representation of an intermediate position.
Examples
1.
To represent and store
one second
of an arm movement through an angle of more than
180 degrees (as illustrated in Figure 9), three positions are required. Thus, the Life
Forms application requires
14,592 bits
(4,864 *3). The binary model, on the other
hand, requires between 324 and 1,540 bits (depending on the sampling frequency),
an advantage factor of 9.5 to 45.
2.
To represent motion where the forearm starts slightly after the arm and fi nishes the
movement slightly earlier, the Life Forms application requires fi ve key frames, that is,
24,320 bits
. The binary model, on the other hand, takes only 324 - 1540 bits (depend-
ing on frequency of sampling), giving a comparative factor of 15.8 to 75.
3.
Defi ning three different motion velocities, e.g., starting with acceleration, continu-
ing at constant velocity, and coming to a stop, means added storage. The Life Forms
application requires seven positions (key frames), or
34,048 bits
. The binary model
