automated coin grading

I haven't seen anything about automated coin grading. I would bet this could be done with with today's technolog. Scan the coin and compare it against the standard and come up with the grade. The grade system could even be 100 point or even 1000 point. This, for me, would make TPGs acceptable because it could eliminate the variability introduced by humans doing the grading. Grade differences would still exist between grading companies because they all would have their own standards programmed in.

 

Log in or Sign up to hide this ad.

Hmmm, I'd like to see something like this. It would be a positive step in the right direction; uniformity and a standard all TPGs would be required to adhere to.

 

Great idea, but how does software determine "eye appeal"?

I can't even fathom how a great work of art could be appraised by a computer... it just can't be done.

 

I've mentioned this before here. I think it would be a great idea and would love to see it come to that. I think they will be able to judge eye appeal as well some day. Examples could be input into the system for it to go buy. Hundreds or thousands if necessary. All under the exact same lighting.

They have computers sorting produce that can tell in a flash how ripe tomatoes or apples are just by going through a scanner, and then by weight they'll get sorted again. These computers analyze the produce for color too. Then machines will kick them off the lines into the appropriate boxes.

It would be a monumental effort to get this done but I think it could be done. Many counterfeits could be caught by weight and size in split seconds.
Then they could be sent to a human grader to verify for sure that it's not authentic, before being destroyed at the factory.

I still wouldn't mind humans scanning everything that went through, as a double check before they would get sent back out.

 

I did a search and didn't find anything. Can you post a link or two please? I'll read up.

Eye appeal is meaningless. Many like those bright red Lincolns. I don't. I prefer chocolate brown. There is no one right eye appeal so you might as well ignore that factor. The idea is a measurable & repeatable assessment of condition. Some sort of spectrum analysis might even give coin color on some scale.

 

WOW - I think eye appeal is important and can vary from coin to coin. With that thought a well dipped coin could have 64 details and be graded 64. And we still get back to the beggining - look at the coin so you buy the coin not the holder. So what appeals to one may not appeal to another. And we did not even discuss the roughness or weak strike of the older planchets.

So now you introduce humans doing the programming for the recognition software - are you going to make them all use the same software? I am sure there will be competitors. So now you have coding differences - you just can't eliminate humans from the equation. And if they can not agree on grading standards now what makes you think they will use the same standards when doing the coding? I mean I can do the coding for some of them now - all morgans MS67 and all moderns MS70.

Part of the fun of this hobby is looking at and picking the coins you like. Yes you may have to look at a lot, but why does it matter what is on the holder? I have passed on plenty of coins that might have been technically graded correct, but held no eye appeal for me or I did not think they matched my opinion of the grade. So again - buy the coin not the holder no matter who (or what) did the grading.

 

Eye Appeal is huge when comparing coins.

Take any 10 1909-SVDB Lincolns, all graded PCGS MS-63RB ( you will find more than 10 of these at any major show ) and compare them. All MS-63 RB's are not the same and will vary in price by $500 or more.

Eye Appeal is huge when it comes to price tags as well.

A certain 1945-S Micro S Merc dime comes to mind

 

for now give me some seed money and i will get a team going and make this work. ITS your chance to become the greatest force in the coin industry for like ever dont let it go GD

 

we are agreeing more these days must be a sign of the changing times

 

i can get it done the technology is there just hasnt been used on coins. if we can scan your eye balls and analyze your thumbprints we can do it for your coins but we need to make it cost effective. luckily you coin collectors are loaded

 

With today's technology, there's no reason they cannot make it work. Eye appeal will always be up to the buyer and the grader and everyone else who sees the coin. Why leave eye appeal only up to the grader?

 

Right - why leave appeal up to the grader. And then again why leave grading upto someone else(or even a computer)? To me this sounds like something someone would want just so they did not have to learn how to grade.

 

It would be more proper to say it "didn't" work. Technology doesn't stand still and I think it's a virtual certainty that advances in optical and computing technology will enable technical grading to take place with more consistency than any human can hope to achieve. If machines can measure flaws down to micron levels and quantify them, coin grading based on detection of wear and contact marks isn't an overly difficult task. Eye appeal will probably remain a human grader's function.

It is very possible that this can be done today except for the fact that there is no financial reason to spend R&D dollars in this area because there is no profit in it. Someday, the technology will become so inexpensive and commonplace that human grading will become obsolete. Then it becomes a matter of preference for the various computer grading software packages.

 

Why not have a two step process? Electronic/computerized grading and then have a final check done by a knowledgeable person who has the final say...

Take Care
Ben

 

I would agree that you would be better off employing a method like this. Also with modern technology the surface of the coin could be analized to a greater degree.

The problem is to employ the technology needed to this kind of stuff I don't think it would be cost effective for the submitters.

If I am correct dipping was proved to remove metal from the surface of coins only after close inspection under an extremely powerful microscope.

Do do this to all coins would costs (i'm guessing) exorbitant amount of money.

 

automated............................

hmmm, i can see the future, look for the coin with a CAC sticker AND a memo inscribed on the holder.......................with computer agreement!!!

 

It doesn't work yet. Same was true of manned flight - some 19th century people said it was impossible. They were wrong.

I work in technology, and have pondered this automated grading thing. Computer imaging and image recognition tech would be awesomely complicated, but not impossible.

Bware of the idea that it would be completely objective and remove subjectivity. That would require widespread concensus on the subjective standards. That means the entire community, not just the TPGs. We can't automate standards until we agree on standards, and I don't think the hobby / industry has ever truly done that.

It would also require concensus on the impact of color on the grade, which is hugely subjective (as proven by this thread and others). I don't think that will ever enter into the automation.

And considering the awesome development costs, the automated grading would still be done by TPGs, and possibly only by the TPG that developed / perfected it. No way the average guy would buy the technology. The developers would own the patent, and may or may not choose to license it to others (including major companies, or other TPGs).

I doubt they would ever consider licensing to a competitor.

In short, there are two major hurdles - technology and people politics.

The latter is more challenging !