💡 At a Glance
Coram's detection accuracy relies on optimal camera resolution and placement, and maximizing Pixels Per Foot (PPF) values to capture fine details. Evaluating your setup according to these guidelines is crucial for reliable, consistent detection across various surveillance scenarios.
⚡ Key Takeaways
Optimize Resolution & Placement
Use high-resolution cameras and position them appropriately to capture critical areas at optimal angles and distances.
Maximize PPF
Aim for 50–70 PPF (or higher) for detailed detection, by calculating effective horizontal resolution and FOV to ensure sufficient accuracy.
Consider Environment & Lighting
Ensure cameras are placed in well-lit, unobstructed areas to avoid interference from glare, low light, or obstructions, and at sensible angles to capture typical images of subjects.
Evaluate and Adjust
Regularly assess your camera's current PPF and reposition or upgrade hardware if targets aren't captured with enough detail.
Overview
Coram supports a range of detection types, from identifying smaller objects like firearms and facial details to recognizing motion and activity of people, animals, and vehicles. Achieving consistent, reliable accuracy depends on several key factors:
Optimizing Camera Resolution & Placement: Higher-resolution cameras and optimized mounting positions increase detail and, consequently, detection accuracy.
Maximizing Pixel Per Foot (PPF) Values: Ensure cameras capture enough detail for effective detection, especially for smaller objects or facial features.
Considering Environment & Lighting: Poorly-lit areas, or visual obstructions such as foliage or fencing, may reduce accurate recognition rates.
Pixels Per Foot (PPF) Fundamentals
Pixels Per Foot (PPF) indicates the number of pixels covering one foot of distance in a camera’s field of view. A higher PPF means more fine detail, allowing Coram's recognition models to better distinguish subtle details and features in the images captured by your cameras.
Why PPF Matters
Accuracy: When cameras capture sufficient pixel detail, Coram can differentiate between important detections and benign or unimportant objects.
Consistency: Higher PPF reduces missed detections and false positives by ensuring each frame has enough clarity.
Recommended Ranges
Coram generally recommends aiming for 50–70 PPF (or beyond) to capture enough detail for fine-detail detection tasks. The exact distance you can achieve this PPF at depends on the camera resolution and lens field of view (FOV).
Example Distances:
Camera | Resolution (px) | Horizontal FOV | Maximum Distance |
5MP Fixed lens (2.8mm @ F1.6) | 2560 x 1920 | 113° | 17 ft (at 50 PPF) |
4K Fixed lens (2.8mm @ F1.6) | 3840 x 2400 | 113° | 25 ft (at 50 PPF) 21 ft (at 60 PPF) |
5MP Optical zoom (2.8mm - 12mm) | 2560 x 1920 | 29.2° | 98 ft (at 50 PPF) 82 ft (at 60 PPF) |
4K Optical zoom (2.8mm - 12mm) | 3840 x 2400 | 29.2° | 147 ft (at 50 PPF) 123 ft (at 60 PPF) |
Tip: These maximum distance values are calculated using the following formula, which you can use to guide camera purchasing and installation decisions:
The formula can also be written as follows:
Horizontal Resolution ÷ (2 × PPF × tan((Horizontal FOV ÷ 2) × π ÷ 180))
To learn more about PPF for facial detection and firearm detection, refer to Introduction to Facial Recognition and Understanding Pixel Per Foot (PPF) Requirements for Firearm Detection.
Meeting PPF Requirements
Check Your Setup
Compare your camera’s current resolution and position to Coram’s recommended distances.
If you’re unsure of your current PPF, calculate it using the camera’s horizontal resolution and the effective horizontal FOV in feet.
Maximize PPF Where It Counts
Identify high-priority zones (e.g., entryways, corridors, or focal points).
Place cameras so that footage captured in these areas meets or exceeds your target PPF.
Move or Upgrade Cameras If Needed
If you can’t reach the ideal PPF by repositioning a camera, you may need higher-resolution hardware.
Consult your Coram representative if you're considering replacing or upgrading cameras.
Calculating the PPF Value for a Camera
To calculate the PPF for your camera configuration:
Measure the effective horizontal FOV in feet (area you want to monitor).
Identify the effective horizontal resolution (number of pixels covering that FOV).
Use the formula:
PPF = Effective Horizontal Resolution ÷ Effective Horizontal FOV
PPF Example Calculation
Below is a demonstrative example that shows you how to calculate the PPF at three different locations for the same camera.
Location | Effective Horizontal Resolution | Effective Horizontal FOV | PPF |
A | 475 px | 35 ft | 13.6 PPF |
B | 1425 px | 35 ft | 40.7 PPF |
C | 2135 px | 35 ft | 61.0 PPF |
Note: Only location C meets the minimum recommended PPF value. Locations A and B fall short, even though the effective horizontal FOV remains constant.
Positioning Cameras to Ensure Visibility
Ideally, place cameras within 15–20 feet of critical areas to improve accuracy for smaller items or facial detail, even with lower-resolution cameras.
Aim for reliable angles & eye-Level views, as cameras capturing near-frontal perspectives increase detection likelihood.
Try to position cameras with an unobstructed view of subjects, as the camera may not capture enough detail to offer sufficient detection accuracy, even if it is only partially blocked.
Where possible, subjects should be well-lit, as intense glare, low light, or extreme backlighting can mask key features.
Even if cameras meet PPF guidelines, its important to remember that harsh or uneven lighting, awkward or uncommon angles, and visual obstructions can significantly lower detection accuracy.
Recommendations for Facial Recognition
For best results when registering a new profile for a face that the system has not seen before, ensure the person is within 20 feet of the camera, looking nearly or perfectly face-on. This close, frontal perspective is crucial for the system to capture enough facial detail to establish an accurate initial match.
Once a face has been successfully registered, Coram can recognize it in future footage even if the person’s face is not perfectly front-facing.
If the camera is positioned high up, or if people rarely look directly at it as they pass, fewer new faces may be registered. In such cases, consider uploading a pre-existing front-facing image of the face in AI Analytics > Profiles to help the system recognize them.
Tip: To learn more refer to Managing Faces, Profiles, and Persons of Interest.
Recommendations for Firearm Detection
If you use firearm detection for any of your cameras, the associated PPF value is a critical factor that impacts the performance and accuracy of firearm detection.
Tip: To learn more about Coram's firearm detection feature, refer to Introduction to Firearm Detection.
Why PPF Matters in Firearm Detection
Meeting minimum recommended PPF values ensures that your cameras capture enough detail for Coram's AI model to accurately identify firearms, since firearms are often smaller and have less distinguishing features compared to faces or larger objects. A higher PPF also reduces the likelihood of the system mistaking other objects for firearms or missing a firearm altogether.
Recommended PPF Values for Firearm Detection
For accurate and reliable firearm detection, Coram recommends a PPF value of between 50–70 and beyond.