Top Used License Plate Capture Cameras

By John Honovich, Published Aug 21, 2017, 10:22am EDT

Capturing license plates is a common video surveillance application.

But what cameras do integrators mostly commonly used? Special purpose LPC cameras or regular cameras? And would it be from the biggest camera manufacturers or from specialists?

150 integrators told IPVM what they use and what results they have. Inside this report, we breakdown the 5 top used cameras for license plate capture and share detailed commentary.

Top ******* ****

Conventional ******* *** ******* ******* ***

***** * ****** ** dedicated, ******* ******* ******* exist ** *** ******, overwhelmingly *********** *** **** are ***** ************, '*******' IP ******* *** ******* plate *******. *** *** 2 ********* ******** ******* these ******* ** ******* is ** ********** (** higher) *** ********** **. Both **** ***** ** allow *** ***** **** (i.e., **) *** ** handle ***** **** ******** issues *** *** ****** shutter ***** ****** *** cars *******.

** ******* *** *** number ** ******** ******** license ***** ******* ******** cameras ** * ************ factor ** *********** ***** conventional *******.

#1 - ****

**** ****** ** ********** mainly ******* **** ******* offer * ******** *** camera ***** *** ** they ******* ****** ******* plate ************. ** *** other ****, ***** ***** relatively ***** **** *** common *** *** ****** projects, **** ******* *** understandable. *** ****** *** Axis * *** * series ******* **** *** most ******** *****, **** as:

  • "**** *****. **** ******** the ******* *********** *** config."
  • "* ********* *** *** Axis *****-*. ** **** as *** **** **** scene ***** ** **** than ** **** ** the ********* ******* ***** capture ********, *** **** have ****** ****** ** target ** *** **** you ****."
  • "** **** ************ **** Axis ***** *** **** good ******* **** ****. As ***** ** *** set *** ****** ** correctly **'* ****** **** to *** * *****. We **** ****'* ******* about ******* ******* ****, this ** ****** ******* lots."
  • "** *** ********** * project ***** *** ***** the **** ******* **** a ******* ***, *** it ***** ** **** well"
  • "** **** *** **** luck ***** *** **** Q1765 *** *** *** to **'* ***** **** lens *** ***** ** emitters. ** **** *** good *** *** ***** capturing *********."
  • "**** ***** ** ***** they **** **** **** with *** * ****** a ****** *******"
  • "**'** **** *** **** Q15/16 ******. *** *** low ***** **** **** very ******* ** *********."
  • "**** ******* *** **** Q1615. **** **** ****** well, ********* ** ******* speed"
  • "**** ***** + ******** lens ** **** *** the ******* ***** *********. It ***** ******* ****** day ****, **** *** external ***** ****** *** the ***** ****."
  • "******** *'** **** ***** the **** *****-*** ** II ** ** * great ****** ***** **** 850nm **"

****, ******* *****-**** *** ** **** highest-end, **** ********* (~$*,***) fixed *******, * ***** HD, *** ****, ********** IR, ******* ******.

#2 - *******

*******'* ********* *** ** outlier ***** *** ***** top ******* ***** ********* conventional *******. ********, *******'* offering ** *** *** (recognition, *.*., ********) ******* of **** ********* ******* plates, **** **** ***** choices ****. *******, *******'* cameras *** *** **** expensive **** ************ ******* (see******* ****** *** ****** Tested) *** **** *** supported ** *******'* ***.

*********** ******** ********:

  • "******* *****. ***** ** outstanding."
  • "******* ** ***** *** a ***** ****** **** Genetec, ** **** ***** from *****. ***** *** Inex ***** **** ****, the ******** ** *****, and *'** **** ******* with ****. * ***** the ***** ******* *** better ***** *******."
  • "** ****** **** **** use *** ******* ****** system *** ******* ******. it ** * **** key ******** *** ** works. ** **** ***** in *** **** ******* cameras/solutions *** ***** *** real ******* **** *** of ****."
  • "******* *** **. ** is ********. ******* ** much ******."
  • "** *** *** ******* SharpX *** ****** *******, which ** **** *** very **** ******* ****. The **** **** **** the ***** ****** *** that **** *** ***** lenses, ** ****** ****** was * *********. *** newer ****** ******* **** this * ******, *** part ******, *** ** works ******** *** **** it (**** ** **** high ******)"
  • "**'** *** **** ******* with *** ******* ******* which *** ***** *****, I *******."
  • "***** ***'* **** ** Genetec- ** **** **** really ****** ***** **** them."
  • "** *** ******* ****** and **** *** ******** specifically *** ****. **** work ********** ****."
  • "*******, **** **** **** but *** * *** pricey"

#3 - ********

**** *** ********. ******** offers*********** *** ********** ************ **** *** more **** **** ** Avigilon *********** **** ************ Avigilon ******* ** ******* license ******:

  • "******** - **** *** excellent!! ***** ***** *** no ****** **** ******* the **."
  • "***** *** ******** ****** cameras * **. ***** well"
  • "****** ******** * ***"
  • "******** *** ****** *** most ******** ****."
  • "********. ********* *** * or * ** ** Bullet ****** ******** ****** detail. **'** **** **** Avigilon's ******* ***** *********** camera *** * *** customers *** **** *** worked ****."
  • "******** ******** *******, ******* when ** *** *** them."

Next - *********

*********, ******** ** *** overall ****, *** *** receive *********** ******* *** we ******* *** ******* impacted ****: (*) ***** they ***** *** *** LPC ******* ** ********* parts ** *** *****, they ** *** (** least ***) ** ***** America *** (*) ********* dealers **** ** **** to ******* **** ***** license ***** ******* ** less ******** ******.

Finally - *****

**** ** ** ***** ago, ***** *** *** of *** ******* ******* for ******* ***** *******. They ***** ***** ******* plate ******* ******* *** have ******* ****** **** common ** ****, ******** and ******* **** ***** overall. ***** *** ******* a ***** ****** ** choices *** *** *** their ******.

Specialist ******* **** ********* *** ******

********** ******* **** ******* manufacturers (*.*.,********* ** ******* ***** Camera ****,****** *** *** ****** Test) ******** * ******* of ********. ***** **** worked **** ** *** tests, ** ******* *** fact **** ******* ***** capture *** ***** ** reasonably ******** **** ************ cameras *** ****** ****** (e.g., ****** ******* ******), that *********** ****** ** stay **** ***** **** camera ********* ****** **** use * ********** / other *****.

Comments (44)

Every time we've tried a conventional IP camera in an LPC scenario, it has failed. Including "big names" promising they'll give us the correct settings for the camera. Has IPVM done a test with say an Axis camera for the purpose of LPC, with no external lighting and in dark conditions? I'd be very interested in seeing those shots and hopefully some video clips, and more importantly, all the setting modifications used.

"Every time we've tried a conventional IP camera in an LPC scenario, it has failed."

Can you be specific how it failed? In what way?

We have a number of tests including: License Plate Capture 4K TestLicense Plate Capture ShootoutLow Cost License Plate Capture Shootout, etc.

We can certainly give more feedback if you explain what ways it has failed for you. One immediate response, 'with no external lighting and in dark conditions', you have to have IR and you have to make the shutter faster.

Seemed we could never get the shutter speed and exposure just right. The pictures ended up too dark or the it had enough light but was too blurry because exposure was too long.

I know there has to be some kind of light, whether it be external to the camera or with it's built in IR. I meant without having to add external lighting. Have you guys found external lighting, whether white light or IR, always has to be added? Or can it be done with the camera's built in IR? If it can be done with the built in IR, that would be great so that it kept installation simpler.

Thanks for the links. I'll will take a closer look at them. I know they have some spec's in there, I guess I just have to dig further to try and glean all the settings needed for the specific camera. But most of those pictures (from a cursory look), don't seem to provide very good tag shots, and least not compared to a LPC camera.


The pictures ended up too dark

Do you mean the entire picture, including the plate, was dark or just the background?

don't seem to provide very good tag shots, and least not compared to a LPC camera.

LPC cameras, by design, are going to deliver a totally dark image at night except for the license plate.

But most of those pictures (from a cursory look), don't seem to provide very good tag shots

Most of those pictures are examples of what not to do (too slow shutter speed or low end cameras that struggle, etc.). Take a look at the details and let us know.

Entire picture is dark, not just the tag. But I don't have expectations to get a lot of background at night when the main objective is to get a good tag shot.

LPC cameras, by design, are going to deliver a totally dark image at night except for the license plate.

I know. Like I said, I don't have a lot of expectation for the surrounding shot detail. And maybe I just have unrealistic expectations. I'm looking for a tag shot from the IP camera that is so close to a LPC camera that even if it's not exactly the same clarity, is pretty close that it almost looks identical at first glance. Maybe that is just an unrealistic expectation.

Take a look at the details and let us know.

Will do.

You can really be given the settings to us as they will be different for every setup.  You have to play and understand how to set them up for proper exposure of the plate. 

We have 50+ Avigilon 3MP 9-22MM bullet cameras in the field to ID license plates.  All with different lighting conditions and different settings.  Even have the Axis Q1765 setup at our office which has different settings because of different lighting conditions. 

I know they may be somewhat different, but they shouldn't be that different. Dark is dark, 20ft to 30ft is a pretty clear distance specification and 30mph or less should be doable. And I have worked with the manufacturer's on the settings and sent them back pictures. They send me back an adjustment to make, and it still doesn't look near as good as an LPC camera.

Some of the shots are 30ft out others are 100ft.  Some have a lot of ambient light others have all IR.   Some have cars going 40+ MPH and others have cars at a stop sign or speed bump.  

Distance and speed are one thing. But if it's a lot of adjustment that has to be made for ambient light, I probably would not be comfortable with that. Let's say there is a lot of street light that you have to adjust for, and you do and get a good shot. But if the light goes away one night, does the shot not work anymore? A LPC camera would work in all light conditions. I would not like being dependent on having a good shot because the lighting did not work one night or just isn't consistent for other reasons. So maybe that goes to my point about my expectations not being realistic.

Let's say there is a lot of street light that you have to adjust for, and you do and get a good shot. But if the light goes away one night, does the shot not work anymore?

Yes, it will work because even though there is ambient lighting we still have IR but you will have to adjust when the camera switches to B/W mode as sometimes there is enough light to keep the camera in color.   This is why I said all settings are specific to the install. 

Michael, thanks so far for you input, it is appreciated. If I could trouble you further, have you tried keeping the camera in night mode and seeing if it works in the daylight as it does at night, or do you think it would work that way? I know the question sounds a little odd. I can explain further, but I was just wondering.

we've had great success specking "normal" cameras for LPR scenarios. Their are so many configurable settings on a good "normal" camera that you can indeed turn it into an LPR camera. Some major things that play into getting a good license plate

- optical zoom. 
- shutter speed
- IR lighting 

and sometimes you have to play with the other image settings to really nail it down. The image will be very dark due to the fast shutter speed, it will actually look pretty crappy, but very effective at capturing plates. One thing I do know is that its not a one and done deal, It takes some tweaking to get it 100%, you may be playing with it for a few days before you get it right because you have to test the settings and make sure they work under many different conditions. Their are so many variable that come into play with this scenario.

Granted, ive never had to speck out cameras capturing license plates that are going 70 MPH such as a turnpike camera, but for normal street and gate cameras, works great.

3  basic rules apply. Lighting, Speed and Resolution.

Resolution....Most manufactures suggest 75 pixels per foot horizontally. I prefer higher. Assuming that there are no lane barriers, a typical Hi Quality 1080P camera will do the job. They work well in low light with technologies like ZIPstream and SNAPstream. Don't go 5MP or 4K, they are great with resolution. but they are only good from sunrise to sunset.

Speed.... Slow the car down and install speed bumps. pretty simple. install the cameras at a 10 degree angle if possible. In California, it is illegal not to have front license plates. Look for the plate in the rear. Brake lights are easier than headlights when it comes to ALPR software.

 Lighting...LED'S or IR. Frankly I prefer LED's as the are energy efficient and they tend to keep the bad guys away. I would rather escort a female co-worker to a safe place under light than under light that only cameras can see.

We recently tested half a dozen cameras for a particularly troublesome LPR project. Plates are from another locality and have very little contrast. We tested several avigilons, Messoa, Geovision LPR, Dahua, UNV, all with and without additional IR power. It was a very strange test, The avigilons did not cut it regardless of settings, which blew my mind. The LPR specific cameras were dissappointing at best. Geovision would not play nice with Avigilon VMS, and in the end, we got the best result for this particular application out of a UNV conventional camera with a very fast shutter speed and tons of IR power.


This test challenged everything I know about LPR capture and made one thing perfectly clear, results will vary greatly depending on the layout and colors and reflectivity of the plate being captured.

for a particularly troublesome LPR project...results will vary greatly depending on the layout

Armando, at a high level, what was the layout? Distance? Angle? Issues?

We had a clear shot of the plate with minimal offset, and even tested the cameras from a direct overhead shot.

The main problem was the plate itself did not provide enough contrast to make the plate readable consistently in the dark. We were struggling with lighting at the site, so we decided to do extended testing at our own site, replicating the lighting with a lux meter. (very very dark) and the plate would bloom with most cameras. The LPr specific cameras would still do it. We brought in our avigilon engineer to test with us and he had the same problems. There were other concerns, but the plate itself was the harbinger of the biggest issues.

The main problem was the plate itself did not provide enough contrast to make the plate readable consistently in the dark.

Was it a specific state of type of plate? Was the plate dirty or did it have some other issue? Was it just that plate alone or?

This project is still in play so I dont want to put too much out there. its a specific plate. When we ran the tests with other plates we did not have the same results. (the good cameras remained good, but the other cameras did not have nearly as much trouble as they had with the target plate). Plates were brand new provided by the locality for testing purposes. There were two test plate designs, one was not nearly as difficult but they are different enough that finding the right setting to capture both on the same choke point was difficult.

It was the combination of low light, vehicle lighting, plate reflectivity level/details and low contrast.


considerable additional IR was extremely helpful as it allowed for a faster shutter speed.

Florida green plates are the toughest. Anything red after that. We always use 850nm IR pass filters, force B/W, force contrast, brightness, sharpness, gamma and exposure. Always have IR illumination and program the camera on the fly for day, dusk and night. We set for darker, not lighter. Over exposure kills shots. Our ALPR software is tuned to darker.

It's an art, but in the end it's become a commodity, an very useful one but a commodity.

John, I think you underestimate the value of your expertise and the nuances of this type of work. We have done a good amount of LPR work, and have replaced a good number of illfunctioning LPR cameras installed by others. The significant opportunity for failure of design is there and its the knowledge and experience that makes it seem so matter of fact. I was surprised how much effort it took for this particular plate after already having a significant amount of experience. To me that makes it all art and science and not a commodity.

Armando - forgive me. I think it is a difficult art, best left to good painters. My commodity comment was directed at component costs, at the paint, not the painter. ;)

Same page

We use P1425-LE / P1425-LE MkII or P1435-LE for barrier solutions. We recommend to use the camera in night mode and IR on all day. It works well.


in the uk market and Europe I would say it's hikvision, I guess this data is based on the American market ?

#2, surely for a major Hikvision distributor, like yourself, Hikvision would be the top.

We did not break down by continent but your 'guess' is clearly impacted by what you sell.

For good plate images you need a competent installer, be it pure IR reflective images or a combination of normal vision with IR. 

My experience (regardless of producer) is that a good OCR engine must be highly customized/tuned for local plate types. Meaning the engine must have had hundreds/thousands of plates processed before any serious installation.

These installations require know-how, and the requirements of the given solution dictates the type of hardware, and number of cameras (front/back)

Also there is always a % of plates that will not be read in any given population, regardless of hardware. Dirt, snow, extreme rain/weather, non reflective, obscured for different reasons (extra lights/bumpers/stickers), foreign plates etc. 
Always communicate and plan for how to handle these cases.

My experience (regardless of producer) is that a good OCR engine must be highly customized/tuned for local plate types.

To emphasize, we only asked here about LPC, i.e., capturing plates not processing / recognizing them automatically (e.g., via OCR). What types of LPR / ANPR software people most use is also an interesting question but not one we covered here.

"(2) Hikvision dealers tend to skew to smaller jobs where license plate capture is less commonly needed." 

This comment is not indicative of the UK market that is more mature than many others in terms of public space CCTV. In most cases it's the installers who are running scared of pushing what is freely available despite the full support of the distributors. Indeed, anecdotally - larger projects tend not to have the "frills" of ANPR which is easier sold to to SMB's rather than corporate clients.

We have sold many brands of ANPR over the past 10 years, all at a high cost and exceptionally hard to set-up. CitySync and the re-branded Vista options are so critical on the camera height, angle and distance from target that it makes every installation nightmare and with far lower capture's than they claim. You also require a separate contextual overview camera. Hikvision smash it out of the park with their free offering. Multiple lanes off one camera, and full daytime contextual viewing. Tech support from DVS is exceptional, allowing the camera to set up at an oblique 30 degree angle with a 100% capture rate for day and night. All camera settings are available to make the installation right first time without hours of return trips such was our experience with other systems.

IPVM has not tested the Hikvision ANPR solution yet, which is shame as this would put a different slant on the story. Would be interesting to know the regional stats on the 150 participants.

The article should, perhaps, have more clearly differentiated between cameras that capture and store to a searchable dbase, than cameras that simple view and capture the image. Happy to say that Hiks had a simple, free, effective dbase search facility coupled with a great quality night time capture. 

The article should, perhaps, have more clearly differentiated between cameras that capture and store to a searchable dbase, than cameras that simple view and capture the image.

I think it was fairly clear to me which type was being discussed, both in the common vernacular that LPC, License Plate Capture, means image capture and not OCR (optical character recognition) reads, and given the article made no mention about the readability of the plates by the system for the purpose of making a text indexed database.

IPVM has not tested the Hikvision ANPR solution yet, which is shame as this would put a different slant on the story. 

No tests would have any impact on this 'story' as it shares survey results. We don't change survey results based on our testing, whether it is for or against any manufacturer.

Secondly, Hikvision ANPR is not yet shipping in North America. We asked Hikvision after you asked us about this earlier in the summer and they said they plan to release ANPR in North America in the near future. When they do, we will test it.

We have had one job that needed a LPR and we used a LTS CMIP7923LPR. It has worked very well and the customer has been extremely satisfied.


As an AVO employee I would like to share that in the UK & Ireland the most used camera for LPR is the Bullet with Adaptive IR (in AVO systems), especially as most applications now seem to be 'slow speed' applications such as Carpark entrances. The LPR solution is simply used to open barriers in most cases. We also supply a document for the SI's which explains how to configure the camera and which model to use (MP). But as Michael explained above no 2 scenario's are ever identical so small changes to the settings might be necessary and that's where the SI's own knowledge of cameras and camera settings is very helpful. I personally have been involved in selling and later assisting with commissioning LPR systems based on Bullets and all worked fine in the end. A challenge in some countries are the personalized (illegal) plates. With personalized I mean using a font that does not meet legal standards, not the combination of fonts such as in the UK. Another challenge can be the number of different plates that need to be recognized such as English, Northern Irish and Irish all in one data base, but we have that covered as well.

Ugh.. Avoid the Messoa's they just do not cut it. :(

Axis has been our best solution so far.. Genetec Sharp is great buy insanely expensive. 

Have not used Dahua and Hikvision's in the field yet but they are looking promising. 

We have just started using the ATV model IP camera.  It has done very well in limited environments in parking lots.

I'm not a fan of MOBOTIX but their Mxpeg codec, rather than traditional H264/5, is actually very good for LPC. It offers very crisp capture, little blur. Since they don't offer any cameras with IR you need an external illuminator, which for LPC, probably allows a brighter and better capture, but adds cost. 

When LPC/LPR is discussed and Integrators are mentioning their success stories, are they viewing the front or rear license plate?  It's my understanding that 18 states do NOT require a front license plate.  I would think that LPC/LPR on the rear would be easier at night then the front with headlights on.  So, are these success stories based on front or rear?

We did not ask about front vs rear in the survey.

Front is definitely harder though given the intensity of headlights. Related: How to Handle Headlight Surveillance Problems

Have used HIKVision for NPR application. One camera can do both Surveillance as well as NPR. The camera's video stream can be recorded as well as the plate number details captured from the OCR images. The camera has built in White-Light LED and can be linked to a Gate Barrier for automatic opening of Whitelist Vehicles. HIKVision NVR was used. Results were good Day/Night generating over 98% accuracy (Customer site had something like 100 staff vehicles) and solution was basically used for Vehicle Access Control. The barrier would open for an authorized vehicle in about 3-4 seconds.

I've heard about this before.  Curious.  Are you in an environment with snow?  When we get snow here, it can sometimes stick to license plates.  Also, when the snow melts, the license plates get covered in grime from the residue of the sand put on the roads when it is slippery out.  My assumption is that this would make the reading of the plate impossible.  I assume I would then be asking the customer to get out of their vehicle and wipe their plate off when they get to the door.  Or, am I missing something?

I'm sure one solution to that snow problem you mention would be to just stick a piece of cardboard on the back or front with magnets or something with the license plate number written on it. You don't even have to use your own plate number, you can use a neighbor's. :)

I think any camera could make a good LPC camera given enough white light in that case. But it's nice to have it in the camera rather than needing to add another component like an external light.

I just wrote a blog post on this topic last week. Here are my findings and a couple videos on conventional security cameras acting as LPR cameras, as well as LPR cameras. I have noted Ten Factors to Consider when Buying LPR cameras: 

Raymond, that's good info. The article mentions Messoa cameras, but a couple people here seem to comment that they did not find Messoa very effective. How have they been for you?

Watch the video on the Messoa link in the article. I'll post another video once the installation is complete.

Read this IPVM report for free.

This article is part of IPVM's 6,730 reports, 908 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

Already a member? Login here | Join now
Loading Related Reports