When customers compare thermal imaging cores, the discussion often starts with familiar specifications:
256, 384 or 640 resolution? 12 μm or 8 μm pixel pitch? 25 mK or 40 mK NETD? 30 Hz or 60 Hz?
These parameters matter.
But in real projects, they are rarely the best place to start.
The first question should be:
What is the thermal imaging system actually expected to do?
Security surveillance and industrial temperature measurement are a good example.
Both use infrared detectors. Both may use similar resolutions and optics. Both may even use a very similar thermal core architecture.
But the selection priorities are fundamentally different.
Security Surveillance: The Priority Is Detection
In security and monitoring applications, the main objective is usually to detect, recognize and continuously observe a target.
The system needs to answer questions such as:
- Is there a person, vehicle or other target in the scene?
- At what distance does the target need to be detected?
- How much detail is required?
- Does the system need to support tracking or AI analysis?
- Is the thermal camera installed on a fixed or moving platform?
This changes how we evaluate the thermal core.
1. NETD: Can the System See Small Thermal Differences?
For surveillance applications, thermal sensitivity is critical.
A lower NETD allows the detector to distinguish smaller temperature differences between the target and the background.
This becomes especially important in low-contrast environments, for example:
- High humidity
- Foggy conditions
- Early morning
- Long-distance observation
- Targets with temperatures close to the surrounding environment
For security applications, the question is often not whether the system can measure the target temperature accurately.
The question is whether the target can be clearly separated from the background.
That is an important distinction.
2. Lens Selection Often Matters More Than Resolution Alone
Resolution is important, but optics directly determine how those pixels are used.
A 640×512 detector combined with an unsuitable lens may not perform as well as expected in a specific monitoring application.
Short focal lengths provide a wider field of view and are more suitable for:
- Perimeter monitoring
- Indoor monitoring
- Close-range observation
- Wide-area coverage
Longer focal lengths provide a narrower field of view and place more pixels on distant targets.
This is why we normally recommend starting from three parameters:
Target size + detection distance + required field of view
Only after these are clear does focal-length selection become meaningful.
3. Frame Rate and Latency Matter in Dynamic Systems
For a fixed security camera, very high frame rates are not always necessary.
For moving platforms, robotics, pan-tilt systems or AI tracking applications, the situation is different.
In these projects, engineers often pay more attention to:
- Stable 30 Hz or 60 Hz video
- Low latency
- Fast image response
- Reliable data transmission
- Real-time tracking performance
The video interface is equally important.
Depending on the host platform, the thermal core may need to support interfaces such as:
MIPI, USB/UVC, CVBS or other digital video outputs.
Interface compatibility should be confirmed early in the project, not after hardware design has already started.
Industrial Temperature Measurement: The Priority Is Accuracy
Industrial thermal imaging has a different objective.
The system is not only trying to identify where a hot object is.
It needs to determine:
How hot is it?
That single requirement changes the thermal core selection process significantly.
1. A Thermal Image Is Not the Same as Temperature Data
A thermal imaging core can produce an excellent image without necessarily being suitable for accurate temperature measurement.
For industrial thermography, the system usually needs radiometric capability.
That means the detector output must be converted into usable temperature information.
Depending on the application, the system may require:
- Full-frame temperature data
- Point temperature measurement
- Maximum and minimum temperature
- ROI temperature analysis
- Alarm thresholds
- Continuous thermal monitoring
This is very different from a thermal surveillance camera that only needs a stable video stream.
2. Measurement Accuracy Becomes a Core Specification
For industrial applications, engineers usually need to evaluate additional parameters such as:
- Temperature measurement range
- Measurement accuracy
- Repeatability
- Calibration stability
- Operating temperature
- Working distance
- Environmental compensation
For example, detecting a 100°C component is relatively easy.
Determining whether that component is actually at 98°C, 100°C or 103°C is a different engineering problem.
That is where calibration becomes critical.
3. Emissivity Cannot Be Ignored
Industrial temperature measurement is influenced by the physical properties of the target.
Different surfaces have different emissivity.
Metal, plastic, painted surfaces, glass and other materials can produce different measurement behavior.
The final temperature result may also be affected by:
- Reflected temperature
- Atmospheric transmission
- Measurement distance
- Ambient temperature
- Window materials
- Optical configuration
A higher-resolution detector does not automatically solve these problems.
For a serious industrial thermal system, calibration and compensation are just as important as detector specifications.
The Same Specification Can Have Different Meanings
This is one of the most common misunderstandings in thermal core selection.
Take 640×512 resolution as an example.
In a surveillance system, higher resolution may provide:
- More target detail
- Better long-distance observation
- More pixels on target
- More flexibility for AI processing
- Better digital zoom capability
In an industrial system, higher resolution may provide:
- More temperature measurement points
- Better spatial temperature distribution
- More detailed inspection of small components
- Better localization of abnormal heat
But higher resolution still does not automatically mean higher temperature accuracy.
The same applies to NETD.
A lower NETD is useful in both applications.
But the reason is different.
For surveillance:
Lower NETD improves target visibility.
For industrial thermography:
Lower NETD helps distinguish smaller temperature differences.
Same specification.
Different engineering objective.
Software and Integration Requirements Are Also Different
The difference does not stop at optics and detector performance.
The host system usually requires completely different data.
For security applications, customers may need:
- Video output
- Low-latency image transmission
- AI object detection
- Target tracking
- Perimeter alarms
- Multi-sensor fusion
- SDK integration
For industrial applications, customers may need:
- Temperature matrices
- Radiometric data
- ROI analysis
- High-temperature alarms
- Temperature curves
- Calibration parameters
- Measurement APIs
This is why interface selection should not only answer:
“Can this core output video?”
A better question is:
“What data does the host system need from the thermal core?”
That question often determines whether USB, UVC, MIPI, UART, I²C or another communication interface is required.
Our Approach to Thermal Core Selection
At Panora Precision, when we discuss a thermal imaging core project with a customer, we normally do not start by recommending the highest resolution model.
We first try to understand the application.
For surveillance projects, we usually focus on:
- Target type
- Detection distance
- Field of view
- Platform movement
- Required frame rate
- Latency
- Video interface
- AI or tracking requirements
For industrial temperature measurement, the discussion shifts toward:
- Temperature range
- Required accuracy
- Target material
- Emissivity
- Working distance
- Measurement area
- Calibration method
- Temperature data output
Once these requirements are clear, detector resolution, pixel pitch, NETD, lens, frame rate and interface selection become much more straightforward.
In many OEM projects, the best thermal core is not the one with the most impressive specification sheet.
It is the one that fits the complete system.
The key difference is simple:
Security thermal imaging is primarily about seeing and detecting targets.
Industrial thermography is primarily about measuring temperature reliably.
They may use similar detectors.
But they should not be selected using the same engineering logic.
For OEM and system-integration projects, defining the application first usually saves more time than comparing another ten datasheets.
— Panora Precision Thermal Imaging Cores & OEM Integration
