Graphics Reference
In-Depth Information
Fig. 3.2
Left panel
Mono front view camera system;
Right panel
Stereo front camera system. At
a higher cost, stereo system is capable to detect objects more reliably. Preprocessing stages (image
rectification and disparity calculation) in stereo camera system are typically done in an FPGA or
ASIC
makes stereo camera system more accurate and robust than a mono camera system.
Typical block diagram of a front view stereo camera system is shown in right panel
of Fig.
3.2
.
In today's front camera stereo vision systems preprocessing (such as rectification
and disparity calculation) is typically hardwired and performed on an ASIC or FPGA
before moving data to the processor.
Although stereo front camera system comes at additional cost (increases overall
complexity of the system), the two imaging sensors provide means to calculate scene
depth which is crucial for reliable object detection.
System size reduction drives a constant demand for higher system integration
maximizing number of functions running on a single processor. This drives the need
for more compute performance. On the other hand, compute performance increase
must come without compromising overall system cost. The system needs to be pack-
aged in a miniature enclosure and must deliver maximum compute performance,
while dissipating minimum heat in order to operate at the extreme temperatures of
severe desert heat or Arctic freezing cold. The opposing requirements create a very
challenging environment.
3.3.2 Rearview Camera Systems
Passive backup video cameras are gaining popularity in the USA, especially on
large vehicles such as minivans, pick-up trucks, and SUVs. The rearview camera is
designed to help the driver avoid a backup collision while reversing the vehicle.
