Warehouse Camera Selection GuideBy Ethan Ace, Published May 20, 2015, 12:00am EDT
Warehouses can contain many high-value items yet often do not have people inside them watching for issues.
Expanding on our Camera Selection Guide, we tested a warehouse with different camera types, shown below:
Inside, we show the results of the following experiments:
- Varying Resolution: What is the impact of SD vs. HD vs. 5MP vs 4K vs 10MP?
- Adjusting the Field of view
- Using IR vs. super low light cameras
- Panoramic camera use
At least 1080p resolution is recommended in even small warehouses, since long aisles result in longer distances to target than many scenes. SD and 720p provide few details even at short ranges. However, if higher resolution cameras are to be used (5MP and up), moderately dim lux levels often found in warehouses may cause increased gain and noise due to these cameras' poor low light performance.
Integrated IR is recommended to capture activity at night, unless the facility operates 24/7. Since lights are typically shut down after hours, even super low light cameras will have difficulty detecting subjects.
If only scene/activity overview is desired, panoramic cameras may provide the best option as they cover a very wide area, including both floor and racking space. However, they will not provide reliable identifying details due to their high mounting height and distance from target, and may even miss small objects left or removed from shelves. If higher detail is required, multiple standard cameras are recommended.
Note that this sample scene is typical of a smaller facility, ~60' from end to end. Details at far ends of longer warehouse aisles would be less clear, unless camera FOV is tightened (losing horizontal FOV) or cameras moved closer (requiring more cameras).
SD vs. HD vs. MMP
The image comparison below shows our test subject when he first enters the warehouse area. At a FOV this wide (~45' HFOV), PPF in the SD and 720p cameras is too low to provide strong identifying details. The 1080p and 5MP cameras provide rough details of the subject's face, while he is easily identifiable in the 4K model. The 10MP camera's image quality suffers from increased noise in this scene due to the somewhat dim (~35 lux) warehouse lighting, while others do not have this issue.
Integrated IR vs. Super Low Light
Nighttime lighting in most warehouses is limited, unless the facility is staffed 24/7, making low light imaging a key factor in detecting incidents after hours.
The comparison below shows relative performance of integrated IR vs. super low light cameras using the same resolution and field of view at night, a very dark ~0.02 lux level. The integrated IR bullet provides easy detection of the subject and scene, while the super low light model is essentially black.
The comparison below shows this difference in detail:
Due to their large size and potential for activity to occur nearly anywhere in the FOV, panoramic cameras are sometimes recommended. These cameras allow users to see an entire aisle or aisles (depending on layout) from a single point, though with reduced details compared to typical fixed lens cameras.
This panorama image shows an example of a fisheye camera viewing a warehouse aisle. The subject is visble as he enters, takes an item from a shelf, and exits.
However, few details are available due to the low PPF and high mounting angle of the camera, as shown in the side by side comparison below, with no identifying information aside from rough height and clothing style. It may also be difficult to detect removal of small items with PPF this low.
This drop in details is even more noticeable at night. Even using a panoramic model equipped with integrated IR, the subject is difficult to detect except when very close to the camera, seen below.
For more reading on panoramic cameras and their related issues, see:
For more coverage on other common scenarios, like hallways, entrances and rooms, see our Camera Selection Guide.
1 report cite this report:
Back to Top