96 contiguous spectral bands captured simultaneously across the whole scene, 30 frames per second. No scanning, no motion blur — the snapshot hyperspectral platform Easepect deploys for defence, research and industrial inspection.
400–1000 nm contiguous coverage
Snapshot — no scanning, no motion blur
First-class SDK for research + robotics
Runs on Jetson for real-time deployment
Traditional hyperspectral systems scan line-by-line and freeze the scene; any motion ruins the cube. Living Optics captures the full spectrum in a single snapshot — image moving objects, flying platforms and live conveyor processes without motion artefacts. Every pixel in every frame carries its own 96-band signature.
Easepect supports Living Optics customers across Southeast Asia with scoping workshops, loan units, airframe integration and model-training support — so you arrive at a deployable pipeline, not just a beautiful cube.
Snapshot hyperspectral.
96 contiguous spectral bands captured simultaneously across the scene at 30 fps. No scanning, no motion blur. The reference snapshot hyperspectral camera.
Python + ROS bindings.
First-class Python API and ROS 2 driver. Stream raw cubes, run pre-trained classifiers or slot in your own PyTorch / scikit-learn models.
Starting-point classifiers.
Pre-trained models for material discrimination, vegetation, polymer sorting and food quality. Fine-tune on a few hundred labelled samples.
NVIDIA Jetson deployment.
Move the trained model to a Jetson for real-time production use — conveyor inspection, UAV onboard inference, ISR.
Define the target material or signal. Easepect runs a scoping workshop — bring samples, leave with a data plan.
Benchtop, drone or conveyor — image samples and live processes to build a labelled spectral dataset.
Start from a pre-trained classifier or fine-tune your own on the captured cube. Python / PyTorch workflow.
Move the trained model to edge compute (Jetson) or host PC for real-time production use.
| Dimension | Hyperspectral (Living Optics) | Multispectral (MicaSense) |
|---|---|---|
| Bands | 96 contiguous (400–1000 nm) | 5 fixed bands |
| Resolution | ≈ 6 nm spectral | Wide bands |
| Capture mode | Snapshot, every frame | Snapshot, every frame |
| Motion-safe | Yes — 30 fps full cube | Yes |
| Use case | Research, novel materials, defence | Routine agronomy NDVI / NDRE |
| Indicative cost | SGD 5–6 figures | SGD 4–5 figures |
For routine agronomy NDVI / NDRE, multispectral is the better-value pick. For novel materials, defence applications, polymer sorting or industrial inline inspection, only hyperspectral resolves the signal.
Multispectral gives you 5–10 pre-selected broad bands. Living Optics gives you 96 contiguous bands across 400–1000 nm in a single snapshot. You can discriminate materials multispectral cannot see — similar polymers, mineral phases, camouflage, early-stage disease — because the full spectral fingerprint is there.
Traditional hyperspectral systems build the cube line-by-line as the sensor or scene moves. Any motion blurs the cube. Living Optics captures every band of the 2D scene simultaneously at 30 fps — so you can image moving objects, live processes and flying platforms without stitching.
Not always. Pre-trained classifiers ship for common tasks (plastics sorting, vegetation stress, material discrimination). For novel targets you fine-tune on a few hundred labelled samples — we run the workshop as part of the sale.
Yes — the UAV payload variant mounts on DJI M300 / M350 class airframes. Typical flight: 80 m AGL, 3 m/s, full spectral cube recorded per frame for airborne hyperspectral surveys.
Living Optics is research / industrial-grade hyperspectral — priced for capital procurement. Typical SpectralCamera + SDK is in the SGD high-five to low-six figures. Evaluation loans are available to qualified buyers.
On-site demos, loaner units and scoping workshops. Tell us the target material or signal — we design the evaluation.
