0

Some banks accept photos of checks for deposit the checks. But they require some smartphone with some particular software. e.g. https://www.fidelity.com/mobile/android

I wonder why they would require some particular software to be installed and used on customers' martphone.

Why not using any camera to take photos of checks, and send/upload it to the banks through my computer?

Thanks!

|improve this question
  • I can scan and upload, but the scanner is their particular software too (some Java app for my bank). I think it must be a security issue, but I don't know enough about the software to be sure. – MrChrister Sep 23 '12 at 5:51
3

This is probably to ensure the integrity of the image. By using their software, they can make sure that the image is not altered after taken by the camera/scanner. Makes sense, security-wise, that's the only way for them to ensure the check does exist physically and haven't been photoshoped. Of course, you can photoshop a check, print it, and then scan it using their software, but that would be forgery.

|improve this answer
  • you can take a picture of a picture of a check and it works and as you point out you can create a fake easily enough that would get through the "Integrity check". – user4127 Sep 24 '12 at 13:55
  • My bank also tries to read the text to make sure its all legible, forcing you to retake the image if it isn't. – Andy Dec 13 '14 at 15:26
2

The software does more than take pictures of the check. It has some OCR and validation built in so that helps to reduce the number of submissions that fail.

Imagine if you could just submit any image to the bank. Someone wants to do a DOS(Denial of Service) attack all they have to do is flood the service with random images pulled from the internet. The server has to spend time trying to process images that are not even checks.

Instead the image is processed through a software that can do the initial processing to make sure that the image being submitted is an actual check. I do not have access to the specs but I would suspect that the software takes a first guess at what the check should be, sends that to the server for it to check against. This will help detect attacks and fraud attempts as well.

In addition there are different compression utilities that can be used on images. The images can appear practically identical when rendered but looking at the data that the computer processes they are completely different files. Since the computer has to work with the data part of the file having a standard image format greatly reduces the complexity of making the software usable for end users.

|improve this answer
  • THanks! THen why not install software on computers instead of smartphones? – Tim Sep 24 '12 at 13:27
  • @Tim - because smart phones have camera built in on a standard interface, They are computers (my droid 3 is a dual core 1.2ghz), people want to use there smart phones because they are smaller and easier to use, and there is such a variation in hardware and supporting structure that it is actually harder to write a user app for a computer than it is for smartphones because of the standardization that went into the design Android. – user4127 Sep 24 '12 at 13:50
  • No OCR is done on the phone when you submit the check. You can verify it with the simple network sniffing: the image goes to the bank and the bank responds with the data that you see on the next verification screen (it least so it seems, there's connection going on, only then you see the data). – littleadv Sep 24 '12 at 17:20
  • @JoeTaxpayer - I actually worked on a system that accepted the input from the OCR and sent the information into the main accounting system. The phone doesnt do the OCR because you could hack your phone. The image is recieved and cached, then sent to the OCR to system to be read, if the OCR is unable to read it then it is put into a queue for manual entry/rejected depending on if the system can get enough information to meet the criteria for queuing. Then the system I created takes that information and puts the numbers in the right fields and submits it to the mainframe. – user4127 Jan 4 '13 at 14:32
  • 1
    "Someone wants to do a DOS(Denial of Service) attack all they have to do is flood the service with random images pulled from the internet." As opposed with the app, where they can do a DOS attack with images pulled from the internet. – NPSF3000 Mar 4 '15 at 11:51
0

like most "security measures" in consumer banking, it is a false sense of security

the banks don't really care about what the signature looks like or anything, as their OCR check scanning software is just converting it into an ACH transaction that simply looks at the account, routing number and amount.

|improve this answer
  • how does it answer the question? – littleadv Sep 23 '12 at 19:06
  • the answer is that the special software is not required, and the requirement is an illusion of security. I decided that my answer sufficiently alluded to that conclusion – CQM Sep 23 '12 at 23:29
  • The fact is that it is required, and the question is why. Saying that the fact is an illusion shows you're not attached to the reality. You can argue the requirement is unnecessary, but you cannot say there's no such requirement. – littleadv Sep 23 '12 at 23:47
  • regarding "proprietary software for check scanning" tell me what regulation requires this for "security reasons". I don't think any federal regulator has addressed this at all they are pretty slow about technology regulations – CQM Sep 24 '12 at 1:02
  • I'm sorry, did I say "Federally required"? I don't think so. Not everything the banks require from their clients is mandated by Federal regulations. – littleadv Sep 24 '12 at 1:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.