Image Processing Reference
In-Depth Information
Fig. 13.2
The six different
types of coins
“How does the robot know where a coin is?” Mick asked. Both Fred and SB
looked at him with surprise until he realized why. “Ahh, through image processing,
got it,” he said smiling.
“What are those small boxes for?” SB asked.
“That is where the robot places the sorted coins. Type one goes into the first box,
type two into the second and so on. I have six different types of coins (see Fig.
13.2
)
and hence 6 small boxes.”
“So the concept is the following,” SB recapitulated, “you have a program running
that can tell the box to release a portion of coins onto the table. The table then
vibrates in order to avoid coins laying on top of each other. The camera then captures
an image of the table and finds the coins and determines their type. The type and
position of each coin is then send to the computer that controls the robot to pick
up the coins one at a time and place them in the small boxes. When the image
processing cannot find any more coins a new portion of coins is released onto the
table. This is repeated until no more coins remain in the box. The robot, the vibrating
table and the box are already operational and you want us to make some image
processing finding the type and position of the coins, correct?”
“Yes, can you do it?”
“Of course” Mick and SB said at the same time. SB continued cocky, “after the
success we had with the system we built for your uncle this should be a-walk-in-
the-park.”
“Careful now,” Mick warned, “this is a different problem.”
“But we still have
the
topic,” SB replied smiling. They revisited the old system
they built for the bartender to find inspiration. After a few initial tests they decided
on the block diagram in Fig.
13.3
.
