JAI Wave Series

• Reveals material properties invisible to visible-spectrum cameras: moisture distribution, polymer composition, and subsurface defects
• Enables continuous inline inspection at production speed with compact 1K and 2K models 
• Distinguishes materials with near-identical visible appearance through high spectral sensitivity and 14-bit output
• Reduces host-side processing requirements through on-camera FFC, DPC, and Destripe image correction
• Fits standard line scan installations: GigE Vision, C-mount optics, compact 60 × 60 mm form factor

Polymers with identical visual appearance show distinct SWIR spectral signatures, enabling automated material classification.

Silicon is partially transparent to SWIR, allowing subsurface cracks and wafer defects to be detected without contact and without damage

Line scan cameras build the image line by line as the object moves through the field of view, the suitable architecture for continuous inspection on conveyors, web material and production lines.

SWIR line scan combines spectral sensitivity to material properties with the throughput needed for inspection at production speed. Line scan architecture is the appropriate choice when inspection targets are always in motion: conveyor belts, web material, and reel-to-reel processes. The JAI Wave Series combines both.

Quality classification based on internal condition rather than surface appearance alone. Moisture content, organic compound distribution, and foreign material identification are reliably detectable. The wide field of view achievable with the 12.8 mm sensor and C-mount optics allows a single camera to cover parallel conveyor lanes, reducing hardware count on high-throughput sorting lines. 

Silicon transmits SWIR wavelengths, enabling subsurface crack detection, layer uniformity inspection, and alignment verification without contact or destructive testing.  

Polymers with identical visible appearance carry distinct SWIR signatures. Line scan imaging classifies mixed plastic streams at sorting speeds, including black and dark plastics not detectable by NIR-based systems. 

SWIR backlight illumination reveals the seal zone clearly, making seal width and seal integrity directly measurable. Zipper closures and foreign material inclusions are detectable in the same pass. 

Firmware configuration, camera pre-settings, and commissioning documentation, reducing on-site setup time for Wave Series installations. 

Feasibility testing validates SWIR spectral contrast using actual production material samples before system specification. Technical coaching covers illumination selection, spectral filter pairing, and acquisition parameter optimisation. System design support is available for complete SWIR line scan installations. 

For multi-camera web inspection layouts, application-specific subsystem development, or SWIR imaging within existing production line control infrastructure: project qualification and development through to series readiness. 

SWIR (short-wave infrared) covers the 900–1700 nm wavelength range, beyond the visible spectrum (400–700 nm) and the NIR range accessible to extended-silicon sensors (700–900 nm). Unlike visible and NIR imaging, SWIR captures molecular-level differences in materials, including moisture absorption, organic compound signatures, and polymer composition, that produce no contrast at shorter wavelengths. Standard silicon sensors have no response in SWIR; InGaAs sensors are required.

SWIR line scan cameras are appropriate when the inspection decision depends on material properties producing no useful contrast in the visible or NIR range, and when the production process is continuous. Representative applications include moisture content mapping, polymer classification on conveyor belts, subsurface defect detection in silicon wafers, and fill-level or seal verification through opaque packaging. Where the inspection target is surface appearance or geometry, a standard monochrome or colour camera is generally more cost-effective. 

The 1K configuration provides 1024 pixels per line at up to 29 kHz. The 2K configuration provides 2048 pixels per line at up to 40 kHz using a dual-row pixel-shift design: two offset 1K sensor rows are combined in-camera in real time, producing an effective pixel pitch of 6.25 × 6.25 µm and enabling sub-pixel defect detection without a native 2K InGaAs sensor. The 2K image is synthesised entirely on-camera, with no external acquisition card required. Both configurations share the same sensor width, lens mount, and GigE Vision interface; upgrading from 1K to 2K requires no changes to optics, illumination, or working distance. Selection between the two depends primarily on the spatial resolution required in the scan direction and the inspection line speed. 

SWIR illumination must emit in the 900–1700 nm range. Common sources include SWIR LEDs, tungsten-halogen lamps, and laser diodes at SWIR wavelengths. Line scan setups require uniform illumination across the scan line; spectral bandpass filters are sometimes used to target specific material absorption features. STEMMER IMAGING supports illumination selection and layout through its Engineering Services. 

Line scan cameras build an image line by line as material moves through the field of view, making them the correct choice for conveyor, belt, and web inspection where targets are in continuous motion. Area scan cameras capture a full frame in a single exposure and are better suited to discrete object inspection in controlled stop-and-inspect setups. 

Both configurations use GigE Vision (1000BASE-T) and are compatible with any GigE Vision-compliant acquisition framework. STEMMER IMAGING’s CVB Common Vision Blox provides GigE Vision acquisition, image processing, and analysis in a single development environment. JAI SDK is also available directly.

Yes. STEMMER IMAGING provides feasibility testing through its Engineering Services, validating spectral contrast against actual production material samples before system specification. This is particularly relevant for SWIR applications, where achievable contrast depends on the molecular composition of specific materials and can vary between nominally similar samples.