Vehicle Entrance Camera Selection GuideBy Ethan Ace, Published Jul 21, 2015, 12:00am EDT
When it comes to vehicle surveillance, the conversation regularly turns to license plate capture. But other information, such as identifying details of the driver, vehicle make and model, etc., is often just as important.
In this addition to our Camera Selection Guide, we look at a parking lot entrance/exit, seen here:
And test three key scenarios:
- Can I identify make and model of the vehicle, and any unique features?
- What are the tradeoffs between a wider and tighter FOV?
- Can I get identifying video of the driver, even through windows?
When covering outdoor vehicle access points, light handling is more important than resolution, as changing daylight levels and glare may cause washed out images in many cameras. Generally, across a 1-2 lane entrance, using a ~25' FOV, 1080p provided enough resolution to identify vehicle make and model (and possibly license plates) as well as driver details through windshield/windows when entering.
Polarizing filters (see our full test) may reduce glare and improve images of subjects through winshield and windows
Due to strong light produced by headlights and taillights, capturing vehicle make/model and driver details is very difficult at night with out adding significant light or heavily adjusting exposure. If nighttime capture is desired, we recommend a dedicated license plate capture camera or properly configured IR bullet be used. See our report License Plate Capture Shootout 2014 for these topics.
In the tests below we cover three key surveillance objectives:
- Obtaining vehicle information, such as make, model, and any unique features.
- The tradeoffs between wide (~90°) and tight (~35°) FOVs.
- Driver identification, viewed through windows and windshield, including handling sun glare.
Note: This section does not cover license plate capture or recognition. LPC/R are nuanced topics better covered in their own reports. To see more on license plate capture see our License Plate Capture Shootout 2014.
Field of View Selection
The decision between wide and narrow field of view depends on what levels of detail and overall scene awareness are desired.
Wide field of view: Using a wider field of view, users are able to capture more scene overview, showing direction of travel, cars on nearby roads, etc. However, details of the vehicle are reduced or in some cases lost, and may be limited, i.e., "red sedan" vs. "red Subaru Impreza 4-door."
Narrow field of view: Narrowing the field of view to the width of the entrance provides better details, with vehicle make and model more easily discernable, and potentially delivering further details such as wheel style, damage to the vehicle, etc.
Using the IPVM Camera Calculator we can see the difference in field of view, below:
For the purposes of our test, we used a medium field of view (~60 degrees), looking at capture of vehicle information as well as driver identification:
Cameras 1080p and above are reliably able to capture vehicle make and model in the daytime in this field of view. The 720p camera is borderline, with the Subaru logo visible, but not reliably clear. SD resolution is able only to identify vehicle style and color.
With the car approaching the cameras, 1080p resolution and above are needed to provide identifying details of the driver's face through the windshield/windows, though the 5MP camera is notably darker than the 1080p and 4K models, obscuring some details. 720p provides rough details, but is notably more pixelated than higher resolutions, while SD simply provides PPF too low to view details.
In some applications, polarizing filters may be used to cut down sun glare and increase driver identification performance. We cover this in detail in our Testing Polarizing Filters on HD Cameras report.
The comparison below, taken from that report, shows the difference in image quality between a camera with filter, without, and with a P-Iris lens, with significantly better images delivered by the camera using the polarizing filter.
Note that polarizing filter performance varies depending on mounting angle and angle of the sun, and may require repeat fine-tuning for best results.
While both of the above tasks are possible during the day, with good lighting, all cameras tested failed to capture both faces and vehicle information at night due to strong lighting caused by the vehicle's headlights and taillights. As mentioned, if nighttime capture is desired, we recommend a dedicated license plate capture camera or properly configured IR bullet be used. See our report License Plate Capture Shootout 2014 for these topics.
IR On vs Off
Using a 1080p IR bullet to cover the entrance, we see the obvious tradeoff between IR on vs. off. With IR off, the vehicle's tail lights wash out details of the vehicle as well as the license plate. However, with IR on and exposure sped to 1/250s to compensate, the scene becomes too dark to view vehicle details, though the license plate becomes visible. If capture of vehicle information is needed at night, additional visible light should be added to the scene to offset the bright tail light hot spots.
Though we were able to capture driver information during the day, at night the car's headlights wash out the scene. Even with IR on and using a fast exposure, only headlights are visible.
Back to Top