How to choose surveillance camera resolution: 2 million, 4 million, 8 million pixels
Common gears and screen sizes
| Commonly known as | Resolution | Total number of pixels | Typical uses |
|---|---|---|---|
| 1 million / 720P | 1280×720 | 920,000 | Corridors, elevators and other small areas |
| 2 million / 1080P | 1920×1080 | 2.07 million | Mainstream configuration, suitable for indoor and outdoor use |
| 4 million / 2K | 2560×1440 | 3.69 million | Entrances, cashiers and other details that need to be identified |
| 8 million / 4K | 3840×2160 | 8.29 million | Squares, parking lots, requiring large scene details |
Key indicator: pixel density
What really determines "whether you can see clearly" is not the total pixels, but the pixel density per unit area. Commonly used references in the industry:
- See outline/behavior: about 40~60 px/m
- See the human face clearly (recognizable): about 120~160 px/m
- See the license plate clearly: about 150~200 px/m
For example: If you need to recognize a face in a 3-meter-wide area, the horizontal dimension must be at least 3 × 140 ≈ 420 px, and 2 million pixels (1920 wide) is completely sufficient; If the area is 8 meters wide, you need 8 × 140 = 1120 px. 2 million is too tight, so it should be 4 million or 8 million.
Cost: Bandwidth and storage increase simultaneously
The resolution is doubled, and the code rate is roughly doubled. Taking H.265 as an example, 2 million is about 2 to 3 Mbps, and 4 million is about 4 to 6 Mbps. 8 million is about 8~12 Mbps. This means that storage costs, switch bandwidth, and NVR decoding capabilities must all be evaluated simultaneously.
Selection conclusion
- Ordinary indoor passages and elevators: 2 million is enough, the most cost-effective.
- Entrance, front desk, cashier: 4 million, taking into account details and costs.
- Large scenes (squares, parking buildings, perimeters): 8 million or a combination of dome camera + gun camera, taking into account both panoramic and close-ups.
- Do not unify 8 million for the entire site, which will cause a huge waste of storage and bandwidth.