Kodierung und Dekodierung der Spektralen Verteilung von Licht mit einem Neuronalen Netzwerk
Anmelder: Intelligent Vision GmbH 🇩🇪
Details
- Veröffentlichungs-Nr.
- EP4749253
- Anmeldetag
- 20. April 2023
- Veröffentlichung
- 27. Mai 2026
- Rechtsraum
- EP
- IPC
- G01J3/28
Abstract
The particular spectral distribution (310) of light (210) is encoded to a spectral distribution identifier (250/270). The light (210) is separately filtered by a set of filters (120-n), that together comply with uniqueness conditions. The filtered light (220-n) is measured (130) to obtain a provisional intensity vector ({B}, 230-n). To compensate for variations, computing functions (140, 150, 160, 170) use an intensity reference value (L_DATA) to accommodate light variations and use pre-determined calibration data ({CAL}) to accommodate filter variations. The computing functions thereby turn the provisional intensity vector to the spectral distribution identifier (250/270). Pre-determined calibration data ({CAL}) are obtained from a neural network that operates with simulated data.
Anmelder
- Firma
- Intelligent Vision GmbH
- Land
- 🇩🇪 Deutschland