Critical Infrastructure Surveillance: Radar & Thermal Approach

Wu Kang | Radar System Specialist | Updated September 2026

Critical infrastructure surveillance combines radar detection with thermal and visible-light imaging so a facility can flag both an approaching intruder and a developing equipment problem, day or night, without relying on a person watching a camera feed continuously.

Two Different Problems Under One Label

“Critical infrastructure surveillance” covers two distinct jobs that get bundled together in procurement conversations, and separating them changes what a facility should specify:

• كشف التسلل المحيط: detecting a person or vehicle approaching or crossing a boundary, day or night, in weather that would defeat a camera-only system.
• Equipment and process anomaly monitoring: catching a developing thermal problem, such as an overheating connection or a hot spot on process equipment, before it becomes a failure.

Both use radar and thermal sensing, but they answer different questions. A radar-vision fusion system built for perimeter detection is optimized to find a moving target crossing a line. A thermal camera used for equipment monitoring is optimized to measure temperature on a fixed asset. Buying one to do the other’s job usually means paying for capability that goes unused, or missing the coverage that was actually needed.

Why Radar Alone or a Camera Alone Falls Short

A camera-only perimeter needs either enough fixed units to cover every approach angle or an operator actively watching, and its usefulness at night or in fog depends entirely on the imaging mode. Radar covers wide areas continuously regardless of light or most weather conditions, but a radar track by itself doesn’t confirm what triggered it: a person, a vehicle, or blowing debris.

Radar-vision fusion closes that gap by having the radar cue a camera automatically. When the radar detects a track, the system slews the camera to that bearing so an operator (or an AI recognition layer) gets a visual confirmation without manually panning to search for it. This is the mechanism behind Midradar’s MR-RD02SE-DMS15-AU Radar-EO Detection System, which combines radar detection, visible-light imaging, infrared thermal imaging, AI target recognition, and servo pan-tilt tracking through the RGS2000 platform for coordinated search, tracking, identification, and alarm handling.

Thermal Imaging for Perimeter Detection: What’s Published

For perimeter-facing thermal imaging, the MR-HTVC6511-2132 is the representative model with published figures:

القدرة	النطاق
اكتشاف المركبات	8.6 كم
Human detection	3.4 كم
Vehicle recognition	2.4 كم
Human recognition	900 m
Visible lens range	8–320 mm
Thermal lens range	21–105 mm (also available in 75/105/155 mm)
Thermal resolution	640×512
Environmental rating	IP66

Detection and recognition are different thresholds: detection means the system flags that something is there, recognition means an operator or algorithm can identify what it is. The 3.4 km/900 m gap between human detection and human recognition range is the practical distance an operator has to confirm a target visually once radar or thermal has already flagged it.

The Published Data Gap in Radar-Vision Fusion Systems

This is worth stating plainly rather than glossing over: Midradar’s radar-vision fusion product line (including MR-RD02SE-DMS15-AU and related MR-AUS-prefixed fusion configurations) is described on its category page by capability, “360° target search, tracking, identification, platform monitoring,” without full per-model distance parameters published. The standalone thermal camera above has concrete detection and recognition figures; the fusion system built around it currently does not, at least not on the public-facing category page.

For a procurement officer sizing coverage against a specific facility perimeter, that gap matters more than it would for a general product overview. Request the current datasheet or a pre-sales technical consultation for the specific fusion configuration under consideration before assuming its effective range matches either the radar component’s or the camera component’s published figures in isolation, since a fused system’s real-world detection range depends on how the two sensors are integrated, not just the sum of their individual specs.

A Misconception Worth Correcting: Equipment Monitoring Isn’t the Same Product Category

Facilities with high-temperature equipment, pipelines, or hazardous-material processes commonly want continuous temperature-anomaly monitoring, catching a developing hot spot on a valve, pipe joint, or electrical connection before it becomes a fire or equipment failure. This is a real and distinct requirement from perimeter intrusion detection, even though both use thermal sensing.

Midradar’s published thermal camera specifications (the MR-HTVC6511-2132 figures above) are oriented toward personnel and vehicle detection/recognition at range, which is the perimeter security use case. There are no published radiometric or fixed-point temperature-monitoring specifications in the current product catalog for process-equipment hot-spot detection. A facility that needs both perimeter security and equipment thermal monitoring should treat these as two separate specification exercises rather than assuming one thermal camera line serves both, and should confirm directly whether radiometric measurement capability exists in any specific model being quoted.

How This Fits Into a Facility’s Security Architecture

A typical layout separates the two problems by zone: radar-vision fusion units cover the outer perimeter and approach lanes, cued to alert on movement crossing a defined boundary; thermal or radiometric monitoring, where applicable, covers fixed high-risk assets inside the fence line. The two layers report into the same alarm-linked platform so an operator sees both intrusion and equipment alerts in one interface rather than switching between separate systems.

A Framework for Specifying Radar-Vision Fusion Coverage

1. Separate the perimeter and equipment-monitoring requirementsbefore writing a spec. They’re different product categories even when a vendor markets them together.
2. Map the facility perimeter length and approach anglesto determine how many radar-vision fusion units are needed for continuous coverage without gaps.
3. Use recognition range, not detection range, to size operator response time.Detection tells you something is there; recognition tells you what it is, and that gap is your actual reaction window.
4. Request the current fusion-system datasheet directlygiven the published data gap noted above, rather than estimating range from the radar or camera components separately.
5. Confirm radiometric/temperature-measurement capability explicitlyif equipment hot-spot monitoring is part of the requirement, since it’s not part of the published perimeter camera specs.
6. Plan the alarm-linked platform integration alongside sensor selection, so intrusion and equipment alerts land in one operator interface rather than two disconnected systems.

Integration and Platform Standards

Radar-vision fusion and thermal imaging systems connect to a facility’s existing security or SCADA-adjacent platform through open protocols: GB/T 28181, ONVIF, and RTSP for video/track interoperability, with RESTful APIs and SDKs available for deeper platform integration. Confirming which VMS, alarm platform, or command center the facility already runs before specifying hardware avoids discovering a compatibility gap during commissioning.

What This Costs

Midradar does not publish pricing for أنظمة اندماج الرادار والرؤية or thermal imaging cameras. Cost depends on perimeter length, number of fusion units, lens configuration, and whether equipment thermal monitoring is included as a separate scope item; a quote is required for an accurate number.

Trade-offs Worth Weighing

Radar-vision fusion reduces the operator workload of manually scanning a long, low-activity perimeter and holds up in weather conditions that degrade camera-only systems. The trade-off is that the fusion system’s own published specifications are thinner than the individual sensors that make it up, which means more pre-sales verification work for a procurement officer than a single-sensor spec sheet would require. Equipment thermal monitoring, if needed, is a genuinely separate line item rather than something a perimeter thermal camera line already covers, and treating it as bundled in risks a coverage gap on the asset side of the facility.

الأسئلة الشائعة

Is radar-vision fusion the same product as a perimeter intrusion detection radar?

No. A radar-vision fusion system adds camera cueing, AI recognition, and platform integration on top of a radar detection core. The radar component alone doesn't include the visual confirmation layer.

What's the difference between detection range and recognition range?

Detection means the system flags that a target is present. Recognition means an operator or algorithm can identify what the target is. For the MR-HTVC6511-2132, human detection reaches 3.4 km but recognition drops to 900 m.

Can the same thermal camera monitor equipment for overheating and watch the perimeter for intruders?

Not necessarily. Midradar's published perimeter thermal camera specs are oriented toward personnel/vehicle detection at range, not radiometric equipment temperature measurement. Confirm radiometric capability separately if equipment monitoring is required.

Are detailed specs published for the radar-vision fusion systems themselves? 

Not fully. The category page describes capability (360° search, tracking, identification, platform monitoring) without complete per-model distance parameters. Request a current datasheet for the specific configuration under evaluation.

Is pricing published for this equipment?

No. Radar-vision fusion and thermal camera pricing depends on perimeter length, unit count, and configuration, so a quote is required.

How does this integrate with an existing SCADA or alarm platform?

Through open protocols (GB/T 28181, ONVIF, RTSP) plus RESTful APIs and SDKs, allowing integration with an existing VMS, alarm system, or command platform rather than requiring a proprietary replacement.