eSWIR cameras extend shortwave infrared imaging beyond the conventional 1,700 nm range, potentially helping industrial teams inspect materials and scenes that are difficult to assess with visible-light cameras. A reported 1920 × 1080 colloidal quantum dot imager demonstrates how solution-processed detectors integrated with CMOS readout circuitry can combine high definition with sensitivity into the extended SWIR band.
The technology does not make every SWIR application a better fit, and detector selection still depends on factors such as illumination, sensitivity and operating conditions. But its combination of spectral reach, pixel count and room-temperature operation offers system designers another option to evaluate.
How colloidal quantum dot eSWIR cameras work
Colloidal quantum dots are semiconductor nanoparticles whose optical response can be tuned by changing particle size. In the described sensor, lead-sulfide quantum dot photodiodes are deposited directly onto a silicon CMOS readout integrated circuit (ROIC). The photodiodes convert incoming photons into electrons, which the circuitry reads out at the pixel level.
This monolithic integration differs from approaches that bond a separate detector array to a readout circuit. The reported fabrication process uses semiconductor-industry deposition techniques and does not require hybridization, epitaxial growth, exotic substrate materials or pixel-level sensor patterning. The whitepaper identifies wafer-level fabrication as a potential scaling path; it also reports that arrays with pixel pitches as small as 3 µm have been demonstrated using the technology, although the evaluated eSWIR camera has a 15 µm pitch.
Changing quantum dot size can shift the detector response to longer wavelengths. For the evaluated 1920 × 1080 imager, the reported spectral responsivity extends from 300 nm to about 2.1 µm. The whitepaper describes the camera as sensitive through 2,000 nm. This spectral coverage may be relevant where additional wavelength access can reveal useful material or scene information.
What the reported performance means
The imager was evaluated using EMVA 1288 machine-vision testing methods, including measurements at 1,550 nm. At 30°C in low-gain mode, the report gives a typical dark current of approximately 200,308 electrons per second per pixel, equivalent to about 32 fA, and temporal dark noise of roughly 211 electrons. It reports linearity within ±3.5% and a system gain of about 18.7 electrons per digital number.
The authors calculate an average detectivity of 1 × 1012 Jones from 400 to 2,000 nm at 303 K. Detectivity is a measure used to compare a detector’s ability to distinguish a signal against noise; it should be considered alongside other system characteristics rather than treated as a complete measure of camera performance. In particular, the whitepaper notes that colloidal quantum dot devices can have lower quantum efficiency than thick epitaxial detector materials. In photon-starved applications, that limitation may matter. Where active illumination is practical, the source suggests it may help address sensitivity requirements.
The distinction is important for machine-vision planning: laboratory sensor measurements do not by themselves establish performance in a specific inspection line. Engineers should assess the camera with the target material, wavelength, optics, illumination and required acquisition conditions.
Comparing eSWIR with established detector options
InGaAs sensors are widely used for SWIR imaging, but the whitepaper describes practical systems as generally focused on the 900–1,700 nm range, with resolution and cost constraints. Beyond that band, alternatives include indium antimonide (InSb) and Type II superlattice (T2SL) sensors.
The paper’s comparison of three commercial cameras illustrates different design trade-offs, rather than a universal ranking. The listed CQD camera has a 1920 × 1080 array and 15 µm pixels, while the compared T2SL camera has 320 × 256 resolution and 30 µm pixels. The compared InSb model offers 1280 × 1024 resolution. The source also reports that the CQD camera operates at 30°C with a single-stage thermoelectric cooler, compared with cooling to −73°C for the T2SL camera and cryogenic cooling for the InSb camera.
Cooling requirements, power, weight, spectral range and detector sensitivity all affect system design. The reported comparison indicates that room-temperature operation and high pixel count may make CQD eSWIR worth assessing where more heavily cooled systems are impractical. It does not remove the need to match detector performance to the application.
Industrial applications to evaluate
SWIR imaging is already used in inspection tasks such as semiconductor wafer and die defect detection, moisture assessment in packaged products, glass imaging, produce bruise detection and inspection of lumber or coated films. Extending response toward 2,000 nm could support existing workflows or enable evaluation of additional wavelength-dependent inspection methods.
Potential application areas identified in the whitepaper include:
- Chemical inspection and hydrocarbon or gas detection
- Plastic sorting and wafer defect inspection
- Imaging through haze, fine dust, smoke or mist
- High-speed thermal imaging and laser beam profiling
- Medical imaging, crop monitoring and surveillance
These are application areas to investigate, not guaranteed outcomes. A successful deployment depends on whether the target produces useful contrast in the available spectral band and whether the complete imaging setup meets operational requirements.
Practical implications for system designers
For integrators and plant teams, the central opportunity is a different balance of resolution, spectral reach and cooling needs. A high-definition focal plane array can provide more spatial detail than lower-resolution SWIR formats, while direct CMOS integration and solution processing offer a route toward scalable fabrication. At the same time, quantum efficiency and photon availability remain important considerations.
When assessing eSWIR cameras, define the required wavelength band and spatial resolution first. Then test representative samples under realistic illumination and operating conditions, and compare sensitivity, noise, cooling, power and integration requirements against alternative detector technologies. The practical value of eSWIR is ultimately determined by whether its extended spectral response reveals information that improves the inspection or monitoring task.