IR Viewer

[rbibler]

Getting the images to line up has more to do with the IR Lens and a standard camera Lens. they are not the same. My BX320 Has a 45Degree and my camera must be I don't 50 or 55Degrees...

Best

Ron

[David A. Andersen]

I think you're right David!
There is nothing in that photograph that corresponds to what really shows up in an infrared scan!

Just another one of those bogus IR things out on the Internet!

I did a job the other day for someone that wanted an infrared photograph to put in a magazine on a product they are selling. As I was providing different options for the client, they kept telling me they wanted to see this, they didn't want to see that, the color wasn't cool enough or hot enough in certain areas, things are not as clear as they had hoped etc.

I had to remind them that they are getting an infrared scan to verify performance. If they want a graphic dramatization rather than an infrared scan they should be using Photoshop and make it look however they want.

I am sure the program operates as depicted in advertising, however the end result will not look like this! It actually will look much less dramatic in most cases.

Also, this program will really stink with low resolution cameras.

[David A. Andersen]

Originally Posted by rbibler

Getting the images to line up has more to do with the IR Lens and a standard camera Lens. they are not the same. My BX320 Has a 45Degree and my camera must be I don't 50 or 55Degrees...

Even when you adjust your digital camera to the proper focal length, it is not the same. It gets better though.

The applications that are out there want to put a small infrared scan in the middle of a wide angle view so you can see the perspective of where the anomaly is. So basically you're using a telephoto infrared lens with a wide-angle digital camera. The fisheye effect in wide-angle lenses will distort and prevent alignment.

If I adjust to 45° with a digital camera, the size of the IR scan and the digital picture will be the same and will overlap very closely. However, this is not how the two cameras were manufactured. Nor is it how it was intended to be used.

You can correct the distortion effect of different lens focal lengths so that they will match with computer software, but it is still very difficult.

[Dominic J. Maricic]

Originally Posted by dvalley

Dominic,

I've got it figured out.

The guy who is selling the IR Viewer simply took the digital image and changed the color palette to the overlapped image. Then he probably photo-shopped the defect area.

Can you try this (oh mighty one)?

Yeah I noticed this right away and it's just a Photoshop effect with a little polish. The problem is that it doesn't show anything this way!

[Dominic J. Maricic]

David, what if you took several IR pictures of an area closer up, then a digital from farther back. That way the IR would have higher resolution as we stitch them together to match the digital photo?

[dbelmont]

Just another gimmick IMHO. The question is will it cause a consumer to decide to call me? If so then it would be a good gimmick. But again IMHO I don't see consumers being impacted 9one way or another n my market. I emphasize IR on my website and so far the static pictures seem enough. Given that other inspectors in the area aren't pushing IR at all.

I could do most anything with a website but find that the real key to a successful site is focusing on SEO and not getting hung up on gimmicks.

Just my thoughts.

Don

[David P. Valley]

Originally Posted by dmaricic

David, what if you took several IR pictures of an area closer up, then a digital from farther back. That way the IR would have higher resolution as we stitch them together to match the digital photo?

Dominic,

I will not be able to accomplish this technique until we start getting cold weather. I need adequate Delta T for great IR images.

I can turn on my personal central A/C and try it, but the defects won't stand out as good as the ones I can obtain on the road.

Anyone have any bright ideas?

[Chuck Evans]

I should be able to create the source images using the same method that I use to create "fusion" style images with my current setup (my EX 320 does not do fusion). I just need to take a little more care and perhaps use a tripod.

Here's the fundamental process:

1. Capture the infrared image being careful to capture a couple static reference marks (e.g., HVAC registers, corners, light fixtures, etc.) to aid in registering the IR image to the VB image. Also being careful to hold the imager level. Use the laser to ID the center of the image;
2. Capture a VB image of the same spot with the digital camera. I do this with the camera set at high resolution and wide angle. I do not try to duplicate the focal length of the IR camera (it's much easier to do this with software). The digital image will be much wider field, much higher resolution but must be taken from the same spot and lens height as the IR camera;
3. Using FLIR Quick Report, tune the IR image. This will increase the image resolution from 320x240 (my imager) to 640x480 (I assume it's interpolated);
4. Open the VB image in image editing software, I use PaintShop Pro, but I'm sure Photoshop can do everything necessary. Correct any tilt discrepancies in the VB image so that it matches the IR image for level;
5. Resize the VB image so that the reference points on the IR image match the same points on the VB image (this will correct for differences in the focal distance). This requires some trial and error. For my cameras, I resize the wide angle VB image to somewhere between 55-60% of the original;
6. Paste the IR image over the VB image as a separate layer and adjust position so that it is properly indexed over the VB image (i.e., all of your reference points line up in both images). This may be easier if you make the IR layer semi-transparent while you do this. The IR image will only cover a portion of the VB image as the VB image still has a wider field of view and overall resolution;
7. Once the images are properly registered with each other, crop the image to the size of the IR layer (640x480 in my case), discarding the excess area of the VB layer. This leaves you with a single, two layered image (VB and IR) with both layers showing the same composition, resolution and apparent focal length;
8. Split the two layers into separate single layer images and save them individually. This should give separate matching VB and IR images that you can use for the see-through effect but now using a real IR image.

I know this sounds complicated, but it's not really that difficult, if you are only working a few images.

[Dominic J. Maricic]

Chuck, if you could get me a few good, lined up images I'd appreciate it. That's the hardest part.

[David P. Valley]

Chuck,

I'm looking forward to your results....

[Chuck Evans]

I have an IR inspection today. I'll see what I can come up with. It's new construction, so there may not be any dramatic pictures.

[Dominic J. Maricic]

I'll put a few different ones together and see what we can create.

[Chuck Evans]

Originally Posted by dmaricic

I'll put a few different ones together and see what we can create.

First attempt: Here are a couple of matched VB and IR images of an electrical panel to experiment with. There are no special thermal anomalies in these, but they should be good for experimenting with a see-through style IR viewer.

Go for it Dominic...

[Dominic J. Maricic]

Thanks Chuck, good line ups. I'm going to play with this over the next week or two and see what I can come up with!

[David P. Valley]

Originally Posted by dmaricic

Thanks Chuck, good line ups. I'm going to play with this over the next week or two and see what I can come up with!

Dominic,

Have you found time to determine if you can design a similar IR Viewer for websites. I'd pay to have one of these on my IR website.