SwiftIR | High-Framing Uncooled Thermal Camera by Axiom Optics
High-speed LWIR imaging
640 × 480 active pixels · 17 µm pitch · up to 120 fps in the high-speed configuration
Broadband thermal camera
8–14 µm spectral range
Shuttered NUC / FFC
Built-in mechanical shutter for one-point non-uniformity correction
Single-cable USB 3.0
UVC video, CDC camera control and 5 V power
IR-Vista and SDK
Windows x64 · macOS Apple silicon · Linux ARM64; C, C++ and Python
High-Framing uncooled thermal camera
The SwiftIR 640 is a High-Framing uncooled thermal camera from Axiom Optics, combining compact construction with high-speed LWIR thermal imaging. A 640 × 480 active array, 17 µm pixel pitch and broadband 8–14 µm detector response provide detailed infrared images for research, machine vision and dynamic scene observation. The high-speed configuration supports up to 120 frames per second over USB 3.0.
High-frame-rate LWIR imaging at up to 120 fps
For applications that need more temporal samples than conventional 30 or 60 fps imaging, SwiftIR 640 combines its detector with the high-speed proxy and USB 3.0 interface configuration. At 120 fps, frames are sampled approximately every 8.3 ms. The sensor’s nominal thermal time constant is 12 ms; the sampling interval and detector response time describe different aspects of imaging performance.
This high-frame-rate LWIR camera is suited to developing imaging systems for moving objects, laboratory experiments, process observation and infrared machine vision. Required sensitivity, motion contrast and system timing should be evaluated under the intended operating conditions.
Broadband uncooled thermal imaging
The microbolometer converts absorbed long-wave infrared radiation into an electrical signal. Objects emit infrared radiation according to their temperature and emissivity, while reflected radiation and the scene background also contribute to the recorded image. SwiftIR 640 displays variations in this detected radiance as grayscale or color contrast.
The broadband configuration has no methane bandpass filter. It is designed for general LWIR scene imaging. The standard 19 mm f/1.0 athermalized lens provides a nominal 32.0° × 24.2° field of view and 0.895 mrad IFOV. Other lens options are available on request.
Built-in mechanical shutter for shuttered NUC
A built-in mechanical shutter closes to provide a uniform reference for shuttered one-point non-uniformity correction (NUC), also called flat-field correction (FFC). This compensates pixel-to-pixel offsets and helps maintain a consistent image as the camera conditions change. The correction briefly interrupts scene imaging. Bad-pixel replacement and display enhancement complement the correction workflow.
USB 3.0 thermal camera with IR-Vista software
A single USB connection carries UVC video, CDC virtual-COM camera control and 5 V power. IR-Vista provides acquisition, corrected-image display, recording and offline analysis on Windows x64, macOS Apple silicon and Linux ARM64. C, C++ and Python SDK APIs support integration into custom applications.
White Hot and color lookup tables, min/max linear stretch, histogram equalization and CLAHE help visualize scene contrast. ROI statistics, histograms and line profiles support image analysis; raw .bin recordings can be browsed and replayed for offline review, with trim and processed MP4 export.
Output: corrected images. Radiometric temperature calibration is not specified, so this configuration does not provide calibrated temperature measurements. Sensor NETD is <30 mK at f/1, 300 K and 30 Hz; this condition does not establish camera sensitivity at 120 fps.
Explore Axiom Optics’ LWIR thermal cameras and cores, or contact us to discuss the SwiftIR 640 configuration for your application.
| Detector and optics | |
|---|---|
| Detector | Uncooled microbolometer; 640 × 480 active pixels |
| Pixel pitch | 17 µm |
| Detector spectral band | 8–14 µm; no methane bandpass filter |
| Maximum frame rate | Up to 120 fps with the high-speed proxy and USB 3.0 configuration |
| Thermal response | 12 ms nominal sensor time constant |
| Native sensor NETD | <30 mK at f/1, 300 K, 30 Hz; not a 120 fps camera sensitivity specification |
| Standard lens | 19 mm EFL, f/1.0; athermalized; M25 × 0.5 mount |
| Nominal field of view | 32.0° H × 24.2° V; 39.4° diagonal |
| IFOV | 0.895 mrad with the 19 mm lens |
| System and interface | |
|---|---|
| Video and control | USB 3.0 UVC video + USB CDC virtual COM over one cable |
| Image data | 14-bit digitization; 16-bit image containers |
| Shuttered operation | Built-in mechanical shutter for shuttered one-point NUC / FFC; briefly interrupts scene imaging |
| Processing | NUC / FFC, AGC and bad-pixel replacement |
| Power | 5 V USB-powered architecture |
| Housing envelope | Approximately 36 × 36 × 58.4 mm, excluding lens and cable |
| Environmental | IP67 applies to lens only; camera enclosure rating not specified |
| Software and image output | |
|---|---|
| GUI / SDK | IR-Vista; C, C++ and Python APIs |
| Operating systems | Windows x64; macOS Apple silicon; Linux ARM64 |
| Display and analysis | White Hot/color LUTs; min/max stretch, histogram equalization, CLAHE; ROI statistics, histograms and line profiles |
| Record / replay | Raw .bin recordings; recording browser; trim and processed MP4 export |
| Calibration / output | Corrected images; no radiometric temperature calibration specified |
| Ordering | |
|---|---|
| Configuration | SwiftIR 640 USB / 19 mm f/1.0; high-speed configuration |
| Lens options | Other lenses available on request |
High-speed thermal camera applications
- Laboratory infrared imaging and dynamic scene observation.
- Machine-vision development and moving-object inspection.
- Process visualization and thermal-contrast studies.
- OEM integration with USB 3.0 and C, C++ or Python software.
Application suitability depends on scene contrast, optics, motion and operating conditions. SwiftIR 640 provides corrected images; calibrated temperature measurement is not specified.
IR-Vista software for the SwiftIR 640 thermal camera
Acquire, display, record and analyze corrected infrared images with the common Axiom Optics camera platform.
Operating system compatibility and SDK APIs
| Platform | Supported architecture / interface |
|---|---|
| Windows | x64 |
| macOS | Apple silicon |
| Linux | ARM64 |
| SDK APIs | C, C++ and Python |
Acquisition, correction and display
- Camera control and shuttered NUC / FFC using the built-in mechanical shutter.
- Bad-pixel replacement, White Hot and color LUTs.
- Min/max linear stretch, histogram equalization and CLAHE.
- ROI statistics, histograms and line profiles.
Offline replay and recording management
Record raw .bin frames, browse recordings with scene thumbnails, sort by date/time and inspect metadata. Open files by double-clicking; organize subfolders, trim clips and export processed MP4 video. Live acquisition is licensed; offline .bin review needs no license.

SwiftIR 640 mechanical outline
Five orthographic views of the shared camera housing. Dimensions in millimeters; not to scale. Housing envelope approximately 36 × 36 × 58.4 mm, excluding lens and cable.

