Counting the number of banknotes in a stack using the optical technology

Hi Ankit, Apparently everyone's being distracted by issues about how voltage doublers work. Does this currency counter need to be 100% accurate? I imagine it would. You'd need to be certain that all the note edges are in pretty accurate alignment with each other, or the imaging system could miss some. What about if the edges of some were dirty, causing the detection system to see them as gaps? All the bank note counters I could find on the net appear to work on the principle of mechanically separating them and counting each one individually.

Bob

Are the banknotes of the one denomination?, if so why not accurately weigh the notes to get the total. This assumes the notes are all the same and don't need to be verified as real ones.

I have made laser contour scanners that do just as you want. Basically its very simple. No way will a CCD do what you want, I can tell you that right now, the notes are too fine.

I use a highly sensitive laser meter. The laser bounces of the paper/notes and comes back. It is pulsed and the time of flight is measured to record the distance from the laser imession to the paper. This allows us to see the tiniest of gaps/curves where two notes/paper meet. it is accurate to 0.1 mircometer dept. Scanning the laser up the side of the stack of notes will give the computer a contour and will count the notes. Our program can count notes/papers and articles as thick as whole envelopes. We typically use it to count stacks of mail or stacks of paper.

Depending on how big your stacks are is how much it will cost. We have stacks that are about 600mm high (0.6m). A special drive motor moves the laser head up the stack, a PC is used to digitally process the signal and get a count.

The total cost of the unit is around $15,000 AUD ea. The main expense is the about $6,000 for a laser. $5,000 for drive assembly. Then $1,000 for a computer and hardware box etc.

This can be made even cheaper depending on how fast you need it to scan. We scan notes at about 0.6m stack in around 30 seconds but we can increase speed with more expensive laser to do it in about 3 seconds. If the stack is made smaller obviously this helps. With bank notes you can probably do very small stacks (like 1 or 2 inches) and have each stack counted in a 1 second. Accuracy is 100% perfect since it is a perfect image of the stack.

That is obviously 1 off prototype expensense. If you mass produce the cost would be considerably lower.

That's how NAB do it. I carried a fat wad from an ANZ branch to a NAB branch the other day. ANZ counted it out for me (probably for my benefit) but NAB put it on the scales and that was that.

Been there, done that. Weight is good for coin, note is too light and variable. Only way to achieve accuracy is to use mechanical methods to count note singularly. Even this method can fail, but if one is counting notes in such a fashion then there should be ways of reconciling the amount.

That's also the way large numbers of tiny electronic components are counted. As long as the scales are accurately calibrated, there's no reason why it shouldn't be a very cost-effective solution. Much less to go wrong than using an optical method.

"Wayne Reid"

It is almost certain the same notes were later processed through a note counting machine and visually inspected.

BTW

You related to Wayne Kerr ??

....... Phil