Avigilon 5MP bullet

[cglaeser 0]

Not sure why people keep trying to re-invent the wheel. There are cameras designed specifically for License Plates but yet people continue to try and cut corners and try to make a camera not designed for this purpose.

 

Which LPR camera do you recommend that can acquire license plates for two or three lanes?

 

Best,

Christopher

[ssmith10pn 0]

Which LPR camera do you recommend that can acquire license plates for two or three lanes?

 

None,

I would do 3 cameras.

[cglaeser 0]

Which LPR camera do you recommend that can acquire license plates for two or three lanes?

 

None,

I would do 3 cameras.

 

Well, there's one answer to your question. I could give you several more. Many off-the-shelf solutions are D1, are B&W, have limited distance, and of the IP solutions are not ONVIF compliant and therefore record continuously, to name a few. I've had projects where budget was not an issue and I would have much preferred an off-the-shelf solution, but sometimes you have to roll your own.

 

Best,

Christopher

[buellwinkle 0]

Maybe I can translate as I speak Spanglish as I'm a citizen of both countries. By the way, good pictures of your install at Iguazú National Park. We went during a violent thunderstorm that left the Sheraton in the park without air conditioning, my wife was not happy but turned around when we walked the beautiful park and falls. We then fled to Ushuaia to cool off.

 

What viginetip is trying to show is the clearer sharper results he gets with his camera and has nothing to do with LPR but is using license plates to show the clarity differences between it an the Avigilon H3.

[thewireguys 3]

He is comparing a JPEG2000 image to H.264. There is no contest the JPEG2000 will be always sharper then H.264. I do agree the Bullet cameras need to be sharper but you will never be as good as the JPEG2000 cameras.

 

I while ago when I was testing the 5MP JPEG200 camera I wanted to see how it compared to a 5MP h.264 camera with the same lens. I ordered a Axis P1347 5MP camera and in H.264 mode I could only read about a 1/3 of the plates compared to the Aviglion 5MP JPEG2000 camera. When I switched the camera to MJPEG mode I could ID about 2/3s of the plates. JPEG2000 was much better option.

[buellwinkle 0]

I can see why their Pro cameras would totally outperform the bullets, it has a much larger sensor and a real photographic Canon lens not to mention higher resolution. Also, on the same camera I feel h.264 at a high enough bit rate can surpass mjpeg. Also depends on mjpeg compression levels.

 

In the products they show on their website, it appears to be their own creation, not sure they are use Avigilon Pro although in the pictures at Iguazu, it shows one in a box.

[thewireguys 3]

He is not using the PRO cameras and I was not referring to the PRO cameras. He is using the now discontinued JPEG2000 5MP Box camera.

[viginetip 0]

The comparison is between

Bullet 5Mpx H3

5Mpx Minidomo JPEG2000 (not the PRO line)

[viginetip 0]

We are reviewing the issue together with the representative in Argentina and technical support in CANADA, are looking at the issue at this very closely.

After the Review of Avigilon CANADA, the findings promise

[buellwinkle 0]

I see that now, was confused by Avigilon's website where they show domes and box cameras with h.264 and no mention of jpeg2000 but I now see that they once made a jpeg2000 dome (5.0MP-HD-DOME-DN).

 

Also, with h.264 and mjpeg there's tuning options where you can decide on quality choices. So are we comparing to the highest possible h.264 or mjpeg quality to lossless jpeg2000?

 

So why did they switch back from jpeg2000 to h.264 if it was better?

[thewireguys 3]

Because H.264 is a cool thing to do and most people don't understand the power of JPEG2000. Now I am not saying every camera needs or should be JPEG2000 but it is a tool and it works very well when you need it.

[buellwinkle 0]

What's interesting is that jpeg2000 didn't appear to make it to digital cameras. I would think that someone buying say a Nikon D4 for six grand for body only would want the absolute best quality image possible, but no JPEG2000 in that camera (or any DSLR I just checked).

 

From what I googled, I found that setting JPEG to minimal compression, the equivalent of 10 in Photoshop produced pretty close to artifact free images and setting JPEG2000 to it's lowest compression product marginally better results. So it could be that setting the Avigilon from JPEG2000 to MJPEG may have different compression settings.

 

I'll have to try it on my Nikon DSLR, take a picture in raw and save it to JPEG2000 in Photoshop and save to JPEG and see if there's a big difference in artifacts. I would expect that there may be a difference, not sure it's to the same extent as show in viginetip's images. My thoughts are the cameras have different image quality and JPEG2000 is one reason, but not the complete reason and probably not even the major reason but I'll reserve judgment until I can run the test.

[thewireguys 3]

Your best bet would be to pick up a 5MP JPEG2000 box camera to compare the images.

 

Also I don't see how JPEG2000 would make any sense for DSLR or any other point in shoot camera. JPEG2000 helps in the CCTV world with the ability to dynamically adjust the image size depending on your screen size or matrix view.

[buellwinkle 0]

Didn't know that. All I read on JPEG2000 is that it's supposed to provide greater compression with less compression artifacts. Is this a general thing or just Avigilon's use of it?

 

It's like Mobotix, they had their own compression for years and H.264 came along and they said they were not going to follow the herd. Then I heard a "mooo" when I saw that their D14 was going to have H.264. Can't fight public opinion.

[cglaeser 0]

Didn't know that. All I read on JPEG2000 is that it's supposed to provide greater compression with less compression artifacts. Is this a general thing or just Avigilon's use of it?

 

It's like Mobotix, they had their own compression for years and H.264 came along and they said they were not going to follow the herd. Then I heard a "mooo" when I saw that their D14 was going to have H.264. Can't fight public opinion.

 

Copy and pasted from Wiki:

 

While there is a modest increase in compression performance of JPEG 2000 compared to JPEG, the main advantage offered by JPEG 2000 is the significant flexibility of the codestream. The codestream obtained after compression of an image with JPEG 2000 is scalable in nature, meaning that it can be decoded in a number of ways; for instance, by truncating the codestream at any point, one may obtain a representation of the image at a lower resolution, or signal-to-noise ratio – see scalable compression. By ordering the codestream in various ways, applications can achieve significant performance increases.

[ssmith10pn 0]

What's interesting is that jpeg2000 didn't appear to make it to digital cameras. I would think that someone buying say a Nikon D4 for six grand for body only would want the absolute best quality image possible, but no JPEG2000 in that camera (or any DSLR I just checked).

 

Pros shoot in RAW and Edit in RAW. They only convert to Jpeg for public consumption.

 

And there is night and day difference in a jpeg image and a RAW image quality.

[buellwinkle 0]

Cool, that makes sense then for surveillance to be able to read part of the stream for lower resolution. From what I read the biggest barrier to adoptance has been that it uses a lot of memory and cpu resource, meaning cameras have to have more powerful processors to make it work which may make them less competitive in the market.

 

As for picking up the live stream say for live viewing remotely, hasn't the industry already adopted dual streaming. Clearly not as dynamic solution as jpeg2000 but there has to be a substantial reason that many other surveillance camera companies aren't doing it. In this whitepaper from Axis on compression - http://www.axis.com/documentation/whitepaper/video/video_compression.htm sort of calls jpeg2000 fuzzy vs. jpeg blocky calling it a preference rather than the next new thing.

[what-da 0]

I did lots of study at University about compression and the best video server that I come across used JPEG2000 for recording and MPRG4 (h.264) for live view

And here is a article worth a read that compares the two

http://www.militaryaerospace.com/whitepapers/2012/02/curtiss-wright-controls-embedded-computing1.whitepaperpdf.render.pdf

[cglaeser 0]

I did lots of study at University about compression and the best video server that I come across used JPEG2000 for recording and MPRG4 (h.264) for live view

And here is a article worth a read that compares the two

http://www.militaryaerospace.com/whitepapers/2012/02/curtiss-wright-controls-embedded-computing1.whitepaperpdf.render.pdf

 

Informative article. Thanks for posting.

 

Best,

Christopher

[SportPlumber 0]

JPEG 2000 is a Wavelet compression compared to DCT macro block compression in MPEG. So it fails much more gracefully. If bandwidth is reduced, it will just appear less sharp in the image rather than the total loss of blocks in h264.

 

It is very popular in the Broadcast TV world for transmission of HD events back to the Studio. Most of the Telco digital lines were 270Mbs the SD-SDI bitrate and so compression was needed to get the 1.5Gbs HD stuff down the smaller pipe. 1080i HDTV over JPEG2000 at 100Mbs looks really good and you can actually run 2 streams down a 270 which could give you 3d or higher frame rates.

 

Also MPEG encoders tend to have a "footprint" due to the GOP and when you concatenate, they can look pretty bad especially on an ATSC MPEG2 box. So using the JPEG2000 gets rid of that.

 

It is really cool technology for video and never meant for stills. But I think h264 wins when bandwidth is small. So as stated by thewireguys, it is a specific tool for some circumstances.