Visible + SWIR Dual-Band Camera: Specifications, Applications and Comparison

Visible and Near-Infrared Imaging (400–1700 nm) with Si+InGaAs Sensors

Capturing images across visible and short-wave infrared (SWIR) wavelengths requires combining different detector materials, since traditional silicon-based sensors rapidly lose sensitivity beyond ~950 nm. The 400–1700 nm spectral range of the GD9L0-130CN-G2 / GD9L0-33CN-G2 camera covers the full visible band and the SWIR region, bridging this gap. It achieves this via a Si+InGaAs dual-sensor design (integrating a silicon photodiode array with an InGaAs array) that enables broad-band imaging in one device. In the visible range (approximately 400–700 nm) silicon excels with high quantum efficiency, while the InGaAs portion detects longer wavelengths up to 1.7 μm that silicon can’t capture. By unifying these technologies, the camera can reveal features invisible to conventional visible-light cameras – for example, moisture or silicon material variations that only appear under SWIR illumination – while still imaging normal visible details. This dual-band approach provides a more complete picture in inspection tasks, eliminating the need for two separate camera systems to cover the different spectra.

Key Features and Specifications of the GD9L0 Series Cameras

The GD9L0-130CN-G2 and GD9L0-33CN-G2 are high-performance industrial cameras that offer dual-band spectral imaging alongside advanced on-board processing and robust design. Below are their notable technical features:

1. Dual-Band Imaging (0.4–1.7 μm): Both models use Si+InGaAs sensor arrays to simultaneously capture visible through SWIR wavelengths, delivering imaging across a broad 400–1700 nm spectrum. This allows one camera to detect standard visible features as well as SWIR-specific details (e.g. water content or silicon penetration) in real time. The sensor arrays have a 5 μm × 5 μm pixel pitch and come in two resolution options: 1280 × 1024 (≈1.3 MP) and **640 × 512 (≈0.33 MP)**.
2. Resolution and Frame Rates: The 1.3 MP model (GD9L0-130CN-G2) delivers up to 1280×1024 at 38 frames per second, while the 0.33 MP model (GD9L0-33CN-G2) can reach 640×512 at 152 fps for high-speed imaging. Both sensors use a global shutter exposure (7 μs minimum shutter, up to 0.1 s) so that fast motion is frozen without distortion. The camera outputs 8-bit, 10-bit or 12-bit monochrome images, providing flexibility in dynamic range versus data size.
3. High Dynamic Range and Low Noise: The dual-band sensor achieves a dynamic range of about 56 dB and a signal-to-noise ratio (SNR) up to 52 dB, which is impressive for a SWIR-capable device. This high dynamic range means the camera can capture detail in both dark and bright areas of a scene simultaneously, an important feature for applications like wafer inspection or outdoor imaging with variable lighting. The low noise floor (corresponding to 52 dB SNR) indicates clean, high-clarity images even when observing low-reflectance targets. On-board image processing further improves quality: the camera firmware includes image correction and noise reduction algorithms to enhance the raw sensor data in real-time.
4. Embedded Processing (FPGA+ARM with Linux): Uniquely, these cameras contain an internal processing system based on an FPGA coupled with dual ARM cores. Specifically, an 85K logic-cell FPGA plus a dual-core Cortex-A9 ARM processor (with 1 GB RAM) are built into the camera. This embedded hardware runs a Linux operating system, allowing users to deploy custom image processing algorithms directly on the camera. In practice, this means tasks like image filtering, feature extraction, or even simple AI inference can be done at the edge, reducing the data load and latency back to the host PC. The FPGA handles high-speed tasks (e.g. real-time pixel corrections, HDR merging), while the ARM cores manage image analysis or communication stacks. The camera’s firmware provides a SDK for both Linux and Windows to support user-developed applications and algorithm integration on this platform.
5. Interface and I/O: The GD9L0 series uses a Gigabit Ethernet data interface, supporting up to 1000 Mb/s throughput over standard CAT5e/CAT6 cables up to 100 m. This GigE link ensures high-speed, uncompressed video streaming with ease of integration into existing networks or vision systems. The cameras use a custom network protocol (TCP/IP based) for image transmission along with a provided viewer software/SDK. In addition, a 15-pin D‐sub GPIO port provides rich I/O for industrial integration – including 2 opto-isolated inputs, 3 opto-isolated outputs for triggers/actuators, an RS-485 serial interface, and 2 differential encoder inputs for reading encoder signals. This allows precise hardware synchronization (e.g. triggering from motion encoders or PLC signals) and control of external devices like strobed lights or actuators directly from the camera.
6. Power Efficiency: These cameras are designed for low power draw and heat dissipation. They accept a 12 V DC supply (with tolerance for typical 10–14 V ranges) and consume under 6 W of power. Despite the dual sensor technology, <6 W is quite power-efficient – achieved through careful thermal design and the absence of power-hungry cooling elements. (Notably, many SWIR cameras use thermoelectric coolers, but the GD9L0 series operates uncooled while still maintaining stability.) The efficient power usage simplifies integration on robots or portable systems and means less heat generation in confined setups.
7. Lens Mount and Mechanical Design: For optics, the cameras use a standard C-Mount lens interface, making them compatible with a wide range of machine vision lenses, including those corrected for NIR/SWIR wavelengths. The housing is a compact metal enclosure, roughly 64 mm × 64 mm × 56.5 mm in size and about 300 g in weight. Importantly, the body features mounting threads on five sides (all faces except the front lens opening), which gives flexibility in mounting orientation on industrial equipment or optical setups. This five-sided mounting design allows engineers to secure the camera from the top, bottom, or either side, which is very useful in tight spaces or when aligning multiple cameras. The enclosure is rated IP40 (protected from small objects >1 mm but not sealed against water) and the camera is specified for operating temperatures from -10 °C to 55 °C (non-condensing humidity 20–95%). These specs indicate robust operation in typical indoor lab or factory environments.

Applications in Industry and Research

The ability to image across both visible and SWIR bands makes these cameras versatile for many inspection, detection, and sensing applications. Some example use cases include:

1. Semiconductor Inspection: In microelectronics manufacturing, SWIR imaging can see through silicon wafers and reveal subsurface features or defects that visible light cannot show. For instance, alignment marks buried under silicon, or cracks in silicon substrates, become visible in SWIR. The GD9L0’s dual-band camera can simultaneously capture the top-layer visible appearance of a wafer and SWIR details like circuit patterns on the backside or voids in bonded silicon stacks. This improves semiconductor testing and quality control.
2. Solar Cell and Silicon Material Inspection: Short-wave IR is highly effective for examining silicon photovoltaic cells and wafers. Crystalline silicon is relatively transparent at wavelengths above ~1100 nm, so SWIR imaging can detect micro-cracks, inclusions, or soldering defects in solar cells that are invisible under visible light. The camera’s 0.4–1.7 μm range covers both the standard visual inspection of solar cell surfaces and SWIR analysis of silicon bulk quality and polycrystalline structure. This dual inspection helps in solar panel production (identifying defective cells) and in sorting silicon materials by purity or crystal type.
3. Glass and Polymer Inspection: Many glass and plastic materials have different absorption/reflection characteristics in SWIR compared to visible light. For example, certain coatings, residues or internal stress patterns in glass cover plates (such as smartphone touch screens or optical components) are revealed under SWIR illumination. The GD9L0 camera can be used to inspect glass or polymer layers for contaminants, measure film thickness uniformity, or see through an optically opaque layer (to visible) that becomes translucent in SWIR. This is useful in quality control for display glass, coated optics, or multi-layer plastic films. The camera’s high dynamic range also helps to image glass, which can have both very bright reflections and dark absorption regions simultaneously.
4. Automated Recycling and Sorting:Garbage classification systems benefit from the SWIR band because different materials (plastics, paper, organic matter, textiles, etc.) have distinct spectral signatures in the NIR/SWIR range. For instance, certain plastics that look similar in visible light can be differentiated by SWIR reflectance, enabling sorting machines to identify and separate them on a conveyor. A dual-band visible-SWIR camera can both “see” the object normally (for shape/size via visible light) and analyze material composition via SWIR. This improves accuracy in automated recycling facilities by identifying foreign objects or sorting waste streams by material type.
5. Agriculture and Food – Fruit Defect Detection: Perhaps one of the most striking applications is detecting internal defects or bruises in fruits and produce. Bruises often do not show a visible mark on the surface initially, but the damaged tissue has higher water content and structural changes that strongly absorb **SWIR light (around 1000–1400 nm)**. Using the GD9L0 camera, a fruit can be imaged in dual bands to find “invisible” bruises: the visible-light image shows the fruit’s surface as normal, while the SWIR image will reveal dark spots where bruising or rot is present. This allows automated sorting systems to remove fruits with internal damage before they are visible to the human eye. It’s a powerful technique for quality grading of apples, pears, citrus, etc., improving consumer satisfaction and reducing waste.