ULW and IAFIS
-
Charles Parker
- Posts: 586
- Joined: Mon Jul 04, 2005 6:15 am
- Location: Cedar Creek, TX
ULW and IAFIS
I am looking for any comments (good and bad) on the ULW and IAFIS?
Does anyone have any tips, hints or suggestions for improving the number of hits in IAFIS?
I have heard that some agencies access it in LEO and that it is easier to do that way?
If you are a single source state is there anyway to go through LEO or are you stuck with going through the state switcher?
Looking for other peoples experiences with the ULW in accessing IAFIS?
Does anyone have any tips, hints or suggestions for improving the number of hits in IAFIS?
I have heard that some agencies access it in LEO and that it is easier to do that way?
If you are a single source state is there anyway to go through LEO or are you stuck with going through the state switcher?
Looking for other peoples experiences with the ULW in accessing IAFIS?
Knuckle Draggin Country Cousin
Cedar Creek, TX
Cedar Creek, TX
-
Whisler
- Posts: 22
- Joined: Tue Jul 05, 2005 5:19 pm
- Location: Borger, Texas
- Contact:
ULW
I've been using the ULW for several years, however it is less now that we have our own local AFIS system. One of the requirements for the ULW is that an agency has to search their State’s database before launching a search against the FBI’s IAFIS. If there is no hit at State level, then you can submit to IAFIS.
The most significant problem I run into is the responses I receive from the State. I do not know if it is an issue with the compression software or some other factor, but the quality of the images at State level is poor. Nearly 70-80% of the images on the response are of quality that prohibits comparison against a latent. Many images could be categorized as level 1 prints. It appears the resolution is set too low, as extreme pixilation occurs with any attempts to zoom in. I can provide sample images by email upon request.
The candidate images returned from IAFIS are much better quality; however the ULW is not a very friendly program when you’re trying side-by-side comparisons.
In response to your questions, about ways to increase hits on IAFIS, I’d have to say that overcoming the State’s shortcomings would be the first obstacle. The second would include some enhancements to the ULW software that would allow viewing the responses simultaneously with the latent image and some minor tools to allow interaction between the two (such as being able to draw a simple line between the two).
I realize that the free ULW program is not designed to compete with the features of Cogent, Morpho, and other AFIS/APIS systems, however I believe there should be some basic user friendly comparison features.
In a perfect world our AFIS system would be able to directly interface with the State and IAFIS systems, bypassing the limited features of the ULW. The proprietary barriers with the different AFIS vendors will be difficult to overcome, although I understand there is some work be conducted on that very issue (but that is a topic for another post sometime)
The most significant problem I run into is the responses I receive from the State. I do not know if it is an issue with the compression software or some other factor, but the quality of the images at State level is poor. Nearly 70-80% of the images on the response are of quality that prohibits comparison against a latent. Many images could be categorized as level 1 prints. It appears the resolution is set too low, as extreme pixilation occurs with any attempts to zoom in. I can provide sample images by email upon request.
The candidate images returned from IAFIS are much better quality; however the ULW is not a very friendly program when you’re trying side-by-side comparisons.
In response to your questions, about ways to increase hits on IAFIS, I’d have to say that overcoming the State’s shortcomings would be the first obstacle. The second would include some enhancements to the ULW software that would allow viewing the responses simultaneously with the latent image and some minor tools to allow interaction between the two (such as being able to draw a simple line between the two).
I realize that the free ULW program is not designed to compete with the features of Cogent, Morpho, and other AFIS/APIS systems, however I believe there should be some basic user friendly comparison features.
In a perfect world our AFIS system would be able to directly interface with the State and IAFIS systems, bypassing the limited features of the ULW. The proprietary barriers with the different AFIS vendors will be difficult to overcome, although I understand there is some work be conducted on that very issue (but that is a topic for another post sometime)
J. Whisler
-
josher89
- Posts: 509
- Joined: Mon Aug 21, 2006 10:32 pm
- Location: NE USA
I've recently finished a training class (sponsored by the FBI/CJIS Division) on the use of ULW. It seems that right now, version 5.0 has a fews bugs that have been worked out and the release of that version is coming forthwith. The current version, v3.3, seems to be what the majoirty of ULW users are currently using. Additions to version 5.0 are the introduction of filters, similar to Photoshop, that can be used to enhance a latent directly in the program.
Version 5.0 (as I as told) was designed to be a stand-alone program for searching latents through IAFIS, hence the addition of the filters and other, smaller, changes. Bundled with 5.0 comes uPrint, a small program that now allows for the direct printing of a 10-print card based on FBI number. Before (in v3.3), an image request (IRQ) had to be created and sent off to the FBI. One could select one fingerprint block or several or the entire card and when the request was sent back, one could print to a FPC printer with the specified fingerprint blocks.
As far as tips searching in ULW, there are several that I can offer. The first is to read the article printed in the JFI, July/August 2007, entitled "A Latent Print Examiner's Guide to IAFIS" (If you need a copy, I can PDF it to anyone!). It was s study conducted by the FBI's Latent Print Division in Quantico and their research in performing streamlined and efficient searches. Very good article.
The next tip is to utilize the features your local machine's extractor uses. When the latent print is viewed by the computer, it utilizes a thin-image format. This thin-image will sometimes mark a bifurcation when it's actually a ridge ending (a feature is a feature, right?). Let the machine search it that way. It's up to us to verify a match anyway.
The third tip, really use the ridge counting feature. I work in Nebraska and our AFIS system is new (updated from AFIS 2000). We're still Motorola branded but the new system is smooth (plus we have palm searches!). Our local searches do not require a ridge counting feature. ULW, on the other hand, does. It attempts to count the ridges between features and allows the user to accept its automated count or change it. Count the ridges and input the correct number if it's off. I've searched the same latent twice (once I accepted the automated ridge count number and the other I inputed the correct number). Both searches yielded the suspect, but the one with the correct ridge counts came back the top choice by a lot while the other came back around 15 or so on the canidate list.
One other tip for searching: The penetration level needs to be below 30%. Anything you know about the suspect or finger is helpful. If you can't get below the 30% penetration level, cheat. Send the print two or three times with different fingers marked (i.e. first search, mark it as the number one, two, and three fingers/second search, mark it as four, five, and six, etc).
Here's a little disclaimer: some states and local agencies use Outlook to send in latent searches. Nebraska goes through the state patrol (state AFIS repository) and from there, a LAN connection straight into the FBI. Every search we do has to go through the state (which is nice because we can search juvenille records, civil prints, and sex offenders as well as criminals) and from there, to the FBI. I got a LEO account after the ULW training because of the potential latent print search feature. Well, it doesn't exist yet (or at least no one I've talked to at the Help Desk knows about it!). I just got back from FBI/CJIS in Clarksburg, WV and talked to the rep from my area and she said the same thing; it's not up and running in LEO just yet.
If you don't have a LEO account, send me an email and I will forward you the application; it's no where on the web yet! It's worth having even if you're not a latent examiner. (STEP)--That's me getting off my LEO soapbox!
One bad thing about ULW (and this might be my state's issue) is that when I send off a search, I do not have the ability to pull up the image I sent off to compare with when I get my results. The file is saved in a network folder (programmed by the state), sent to the FBI, and a file is placed back in the network folder with my results. If I try and save the file in another location, it will not get sent properly. So, my suggestion is to save the latent image as a JPG (maybe with dots marking features), and then doing an Alt+Tab dance with ULW and Photoshop when you get results. Again, I think that it's a state issue; at the ULW class I was able to send the searches off, pull the image back up, with my results, (all in ULW) and I cascaded my windows in ULW to show both the searched image and the respondents.
As far as experiences, they've all be good. I've found with very little tweaking, the minutae the state AFIS program extracts needn't be changed. If you see one that isn't marked, by all means, mark it.
One thing to keep in mind is the number of searches per state per day. It wasn't until I attended some training at FBI/CJIS that I was able to pin down a number from a person in the know. Nebraska (I'm told) is allowed 12 searches per day and the system will hold five days in queue. We haven't had a problem yet and I work nights and I can usually get results back (IAFIS) in 10-15 minutes tops. Not sure about daytime.
Well, let me stop and catch my breath. This is the first time I've posted here in this forum and I don't know how much I should or shouldn't type. If anyone has any more specific quesitons, please email me or ask in this thread.
Josh
Version 5.0 (as I as told) was designed to be a stand-alone program for searching latents through IAFIS, hence the addition of the filters and other, smaller, changes. Bundled with 5.0 comes uPrint, a small program that now allows for the direct printing of a 10-print card based on FBI number. Before (in v3.3), an image request (IRQ) had to be created and sent off to the FBI. One could select one fingerprint block or several or the entire card and when the request was sent back, one could print to a FPC printer with the specified fingerprint blocks.
As far as tips searching in ULW, there are several that I can offer. The first is to read the article printed in the JFI, July/August 2007, entitled "A Latent Print Examiner's Guide to IAFIS" (If you need a copy, I can PDF it to anyone!). It was s study conducted by the FBI's Latent Print Division in Quantico and their research in performing streamlined and efficient searches. Very good article.
The next tip is to utilize the features your local machine's extractor uses. When the latent print is viewed by the computer, it utilizes a thin-image format. This thin-image will sometimes mark a bifurcation when it's actually a ridge ending (a feature is a feature, right?). Let the machine search it that way. It's up to us to verify a match anyway.
The third tip, really use the ridge counting feature. I work in Nebraska and our AFIS system is new (updated from AFIS 2000). We're still Motorola branded but the new system is smooth (plus we have palm searches!). Our local searches do not require a ridge counting feature. ULW, on the other hand, does. It attempts to count the ridges between features and allows the user to accept its automated count or change it. Count the ridges and input the correct number if it's off. I've searched the same latent twice (once I accepted the automated ridge count number and the other I inputed the correct number). Both searches yielded the suspect, but the one with the correct ridge counts came back the top choice by a lot while the other came back around 15 or so on the canidate list.
One other tip for searching: The penetration level needs to be below 30%. Anything you know about the suspect or finger is helpful. If you can't get below the 30% penetration level, cheat. Send the print two or three times with different fingers marked (i.e. first search, mark it as the number one, two, and three fingers/second search, mark it as four, five, and six, etc).
Here's a little disclaimer: some states and local agencies use Outlook to send in latent searches. Nebraska goes through the state patrol (state AFIS repository) and from there, a LAN connection straight into the FBI. Every search we do has to go through the state (which is nice because we can search juvenille records, civil prints, and sex offenders as well as criminals) and from there, to the FBI. I got a LEO account after the ULW training because of the potential latent print search feature. Well, it doesn't exist yet (or at least no one I've talked to at the Help Desk knows about it!). I just got back from FBI/CJIS in Clarksburg, WV and talked to the rep from my area and she said the same thing; it's not up and running in LEO just yet.
If you don't have a LEO account, send me an email and I will forward you the application; it's no where on the web yet! It's worth having even if you're not a latent examiner. (STEP)--That's me getting off my LEO soapbox!
One bad thing about ULW (and this might be my state's issue) is that when I send off a search, I do not have the ability to pull up the image I sent off to compare with when I get my results. The file is saved in a network folder (programmed by the state), sent to the FBI, and a file is placed back in the network folder with my results. If I try and save the file in another location, it will not get sent properly. So, my suggestion is to save the latent image as a JPG (maybe with dots marking features), and then doing an Alt+Tab dance with ULW and Photoshop when you get results. Again, I think that it's a state issue; at the ULW class I was able to send the searches off, pull the image back up, with my results, (all in ULW) and I cascaded my windows in ULW to show both the searched image and the respondents.
As far as experiences, they've all be good. I've found with very little tweaking, the minutae the state AFIS program extracts needn't be changed. If you see one that isn't marked, by all means, mark it.
One thing to keep in mind is the number of searches per state per day. It wasn't until I attended some training at FBI/CJIS that I was able to pin down a number from a person in the know. Nebraska (I'm told) is allowed 12 searches per day and the system will hold five days in queue. We haven't had a problem yet and I work nights and I can usually get results back (IAFIS) in 10-15 minutes tops. Not sure about daytime.
Well, let me stop and catch my breath. This is the first time I've posted here in this forum and I don't know how much I should or shouldn't type. If anyone has any more specific quesitons, please email me or ask in this thread.
Josh
"...he wrapped himself in quotations—as a beggar would enfold himself in the purple of emperors." - R. Kipling, 1893
-
josher89
- Posts: 509
- Joined: Mon Aug 21, 2006 10:32 pm
- Location: NE USA
One more thing...
Another thing I neglected to mention. Our state is 1000 dpi while the FBI is 500 dpi. Thus, when we get respondents back from an IAFIS search, we magnify 2:1 with our 1:1 image and voila, same-same.
"...he wrapped himself in quotations—as a beggar would enfold himself in the purple of emperors." - R. Kipling, 1893
-
David Grady
- Posts: 24
- Joined: Fri May 12, 2006 10:21 am
- Location: City of Worcester
ULW and IAFIS
I work for a city in a single source state. However, I e-mail my latent prints (processed in ULW) through the LEO JABS portal. A couple of years ago it took approximately 21 hours to get a response. I enetered one a couple of weeks ago and received a respsonse in about ten minutes. And it was a hit. We have very good success with IAFIS.
I open the image in Photoshop, make any necessary adjustments, calibrate it to the correct size and then set the resolution to 1000 ppi.
My agency uses a local AFIS first (Spex), then we utilize the state AFIS 2nd (we have an NEC FW-21 workstation). We enter into IAFIS only if there are no hits in the first two. As I said, we've had good success and have a decent amount of hits with it. It's definately worth the effort to set it up and send your own prints.
I open the image in Photoshop, make any necessary adjustments, calibrate it to the correct size and then set the resolution to 1000 ppi.
My agency uses a local AFIS first (Spex), then we utilize the state AFIS 2nd (we have an NEC FW-21 workstation). We enter into IAFIS only if there are no hits in the first two. As I said, we've had good success and have a decent amount of hits with it. It's definately worth the effort to set it up and send your own prints.
-
Norberto Rivera
- Posts: 33
- Joined: Mon Aug 15, 2005 7:10 am
- Location: Griffin, GA
- Contact:
David,
Do you use the IAFIS module in your Spex Printquest system? We've had a Printquest system for about two years now and I love it. I got the hardware from the FBI/CJIS for the IAFIS connection and had the analog line installed, but even if I can get the system to connect to the network my submissions don't show up in the IAFIS inbox when I send. I'm about to ship the system back to NJ again to see if they can correct the problem for me, but I was curious to hear your experience with it.
Do you use the IAFIS module in your Spex Printquest system? We've had a Printquest system for about two years now and I love it. I got the hardware from the FBI/CJIS for the IAFIS connection and had the analog line installed, but even if I can get the system to connect to the network my submissions don't show up in the IAFIS inbox when I send. I'm about to ship the system back to NJ again to see if they can correct the problem for me, but I was curious to hear your experience with it.
"We're all here 'cuz we ain't all there!"
"How long a minute is depends on what side of the bathroom door you're standing on."
"How long a minute is depends on what side of the bathroom door you're standing on."
-
josher89
- Posts: 509
- Joined: Mon Aug 21, 2006 10:32 pm
- Location: NE USA
Well, I figured out how to save my image so I can pull it back in to ULW when I get my results back. Just some more tweaking and playing to figure it out!josher89 wrote:One bad thing about ULW (and this might be my state's issue) is that when I send off a search, I do not have the ability to pull up the image I sent off to compare with when I get my results. The file is saved in a network folder (programmed by the state), sent to the FBI, and a file is placed back in the network folder with my results. If I try and save the file in another location, it will not get sent properly. So, my suggestion is to save the latent image as a JPG (maybe with dots marking features), and then doing an Alt+Tab dance with ULW and Photoshop when you get results. Again, I think that it's a state issue; at the ULW class I was able to send the searches off, pull the image back up, with my results, (all in ULW) and I cascaded my windows in ULW to show both the searched image and the respondents.
"...he wrapped himself in quotations—as a beggar would enfold himself in the purple of emperors." - R. Kipling, 1893
-
Lonnette Kendoll
- Posts: 5
- Joined: Thu Jul 07, 2005 10:10 pm
- Location: Richardson, TX
ULW and IAFIS
Charlie, we have been using the ULW software for a couple of years now. We've been without a state terminal (you know the story) ... so we are supposed to use it to search DPS first; however as previously mentioned, there are some major problems with that side of it. The ULW software is not the most user friendly; however since we use it on a daily basis, it's gotten to be old hat now. Proper ridge count and minutiae placement greatly increases scores and results. If presented with a smudge running through the print and detail on either side, search the two areas separately. I've hit on 5 minutiae. All in all, I've been happy with the results but will be greatly cheered when we again have a state terminal.
If you have any questions, give us a call.
Lonnette Kendoll
Police Dept; Richardson, TX
972-744-4870
If you have any questions, give us a call.
Lonnette Kendoll
Police Dept; Richardson, TX
972-744-4870
-
David Grady
- Posts: 24
- Joined: Fri May 12, 2006 10:21 am
- Location: City of Worcester
To Norberto
No, I haven't been able to do my IAFIS entries thru my SPEX AFIS module. But, in all fairness to SPEX, I got one of the earlier PrintQuest models. It has the module for IAFIS, but they stated up front when I bought it that the module was not yet up and running. I'm having the system upgraded this year and hope to fix this.
I spoke to someone here in New England, just a couple of weeks ago, and they told me they had the Spex with a a working IAFIS module and that it worked great. But I can't remember who I talked to (it sucks getting old).
I spoke to someone here in New England, just a couple of weeks ago, and they told me they had the Spex with a a working IAFIS module and that it worked great. But I can't remember who I talked to (it sucks getting old).
-
Norberto Rivera
- Posts: 33
- Joined: Mon Aug 15, 2005 7:10 am
- Location: Griffin, GA
- Contact:
Re: To Norberto
Tell me about it!David Grady wrote:...(it sucks getting old).
The software upgrade is not a big deal at all. (Assuming your service contract is current). I called Spex and they did it over the phone. You won't notice many differences from the user interface, except for the login screen. It gets hits left and right and as long as it keeps doing it I'll keep lovin' it. 8)
I hope after I get it back from NJ next month all the issues will be resolved.
"We're all here 'cuz we ain't all there!"
"How long a minute is depends on what side of the bathroom door you're standing on."
"How long a minute is depends on what side of the bathroom door you're standing on."