558
6 Reverse Engineering of Microcircuits
Fig. 6.35 Typical example of offset value distribution: average offset and deviation (a); offset value
(solid line) and normal probability law (b)
When calculating the offset value, experts recommend selecting a specific line
and a specific column directly from the fragments of the aligned region. Offset values
can be either positive or negative, and the offset sign determines the direction along
the alignment axis. If the alignment results turn out to be unsatisfactory (as it usually
happens), it is possible to extend the borders of the comparison region along and
across, thus increasing the number of the necessary iterations. After increasing the
number of such parameters, it is necessary to repeat the alignment process. At the
same time, experts need to realize that an increase in the number of such parameters
inevitably leads to an increase in the time required for alignment.
Usually, extreme amounts of hardware resources are required to calculate numeric
values for each such pair of aligned frames, since the number of the aligned frames
far exceeds a thousand. In order to reduce the required time, experienced experts
suggest calculating offset for a number of frames selected by the expert as opposed
to calculation for all frames. In order to reduce the time of alignment of a separate
group of bitmap frames into a single panoramic image, it is possible to use the
numeric value of the average offset for all frames included in the panoramic image.
6 Reverse Engineering of Microcircuits
Fig. 6.35 Typical example of offset value distribution: average offset and deviation (a); offset value
(solid line) and normal probability law (b)
When calculating the offset value, experts recommend selecting a specific line
and a specific column directly from the fragments of the aligned region. Offset values
can be either positive or negative, and the offset sign determines the direction along
the alignment axis. If the alignment results turn out to be unsatisfactory (as it usually
happens), it is possible to extend the borders of the comparison region along and
across, thus increasing the number of the necessary iterations. After increasing the
number of such parameters, it is necessary to repeat the alignment process. At the
same time, experts need to realize that an increase in the number of such parameters
inevitably leads to an increase in the time required for alignment.
Usually, extreme amounts of hardware resources are required to calculate numeric
values for each such pair of aligned frames, since the number of the aligned frames
far exceeds a thousand. In order to reduce the required time, experienced experts
suggest calculating offset for a number of frames selected by the expert as opposed
to calculation for all frames. In order to reduce the time of alignment of a separate
group of bitmap frames into a single panoramic image, it is possible to use the
numeric value of the average offset for all frames included in the panoramic image.