The Spectral CIS online acquisition platform is equipped with a broadband transmission-based spectral reconstruction algorithm, enabling high-quality image capture and analysis even under low-light conditions. It supports real-time spectral mapping of the target scene, outputting a 3D datacube that combines 2D spatial imagery with rich spectral information.

Users can interactively extract spectral data from any pixel or selected region within the image, view complete spectral response curves, and perform detailed single-point, multi-point, or regional feature analysis with export options. Offline spectral reconstruction is also supported, allowing flexible post-processing and review of hyperspectral datasets.

The platform accepts multiple data formats and is fully compatible with VIS, NIR, and VIS-NIR series devices. It supports both Windows and Linux environments and provides SDK/API access for system integration and secondary development.

Real-Time Spectral Mapping 

The system enables real-time linkage between spatial imagery and spectral data, allowing users to simultaneously visualize both the 2D image and its corresponding spectral response. This function facilitates rapid identification and instant analysis of target regions.

Feature Recognition

Built-in algorithms extract and identify spectral features at the pixel level, enabling precise target detection and classification. Users can apply default or customized rules to automatically locate anomalies or specific material signatures.

Visualized Analysis

Based on spectral differences, the platform generates intuitive visual outputs such as pseudo-color maps and contrast overlays. This enhances the clarity of target features and improves both analytical speed and accuracy.

Model Training & Development

The platform allows users to import labeled datasets for training spectral recognition models used in classification or regression tasks. With visualized training workflows and performance feedback, users can efficiently build and validate robust models.

Model Iteration & Optimization

Support is provided for model updates and transfer learning to adapt to varying tasks and datasets. Existing models can be incrementally improved to enhance generalization capabilities and ensure consistent performance across applications.