All data is compliant and reusable, supporting customers in rapidly implementing hyperspectral

Описание: I. Core Project Purpose
Using dual-band hyperspectral cameras (400–1000 nm and 900–1700 nm), complete spectral data of Fabric A294 is collected. Combined with the component values of Fabric A294 obtained via the customer’s chemical method, machine learning and deep learning algorithms are used for training and modeling.
This project aims to conduct technical experiments and industrialization for non-destructive and rapid identification of Fabric A294 components, providing an efficient, non-destructive and accurate solution for fabric testing.
II. Sample Information
Sample Type: Fabric A294 (customer-supplied sample)
Purpose: Testing and experiment
Quantity: 3 types (Sample Image: sample.png)
The samples are qualified, highly representative, and properly stored throughout the process to ensure reliable experimental data.


III. Equipment Configuration & Acquisition Process
Testing Equipment Configuration
Complete acquisition equipment is provided, including:
400–1000 nm and 900–1700 nm hyperspectral cameras
Optical dark box (with 350–2500 nm light source and movable sample stage)
Low-reflectance black tray
Identification labels
The system fully meets the requirements for high-precision hyperspectral data acquisition of Fabric A294.
Standardized Acquisition Process

Sample Placement
Place Fabric A294 samples according to the standard diagram (Dark Box Image: darkbox.png) to ensure consistent acquisition conditions and improved data comparability.
Data Acquisition
Use reflection mode to accurately collect reflectance data of Fabric A294 samples in 400–1000 nm and 900–1700 nm bands, ensuring data integrity.
Equipment Calibration
Fix lens aperture at F1.4 (maximum), adjust camera height, focal length and exposure time to ensure full sample coverage, clear imaging and non-overexposed data, guaranteeing overall data quality.
IV. Deliverables
A complete dataset is provided for each Fabric A294 sample, directly usable for modeling and industrial application:
Raw data (400–1000 nm & 900–1700 nm): .dat, .hdr
Reflectance data (corresponding bands): .dat, .hdr
Hyperspectral images: .png
Actual sample placement photos: .jpg
Data and equipment display files:
400-1000 curve.png, 400-1000-03 curve.png, 900-1700 curve.png, 900-1700-03 curve.png, Chassis Image 1.png, Accessory Image 2.png, Image 3.png
Acquisition Reference Standard: Q/EX C 0628-2025
All data is compliant and reusable, supporting customers in rapidly implementing hyperspectral non-destructive identification technology.