Specim, Spectral Imaging Ltd.
- Home
- Companies
- Specim, Spectral Imaging Ltd.
- Products
- SpecimCUBE - High-Performance ...
SpecimCUBE - High-Performance Processing Platform
SpecimCUBE is a high-performance processing platform to run classification models created by SpecimINSIGHT in real-time. It is based on Xavier, a system on a chip developed by Nvidia and includes optimized software to meet industry requirements for throughput, latency and jitter.
Most popular related searches
SpecimCUBE receives the data from Specim FX cameras and processes it in real-time based on a classification model. The sorting results are streamed over GigE Vision to the target systems (sorting machine, existing machine vision systems).
Specim Cube provides excellent performance for real-time sorting applications with the flexibility to switch and run different models.
PROTOCOLS
- GigE Vision compliant. Streaming and controls. Supported format RGB8(application results), Mono12 or Mono12packed (detector image)
- JSON-RPC TCP/IP protocol for controlling CUBE – alternative for GigE Vision
- Specim UDP alternative to GigE Vision network protocol for data streaming
- FTP For transferring the datasets
- AIE Advanced Automatic Image Enhancement
CameraLink
- Supported cameras: Specim FX10 CL / Specim FX17 CL / Specim GX17
- Operating system: Specim CameraOS
- Operating modes: Real-time processing, data collection
- Real time processing: Using applications created with SpecimINSIGHT
- User interface: CameraOS – webUI
- SoC: NVIDIA Jetson Xavier AGX
- Storage: 1 TB
- Data format: Specim dataset with ENVI compatible files (for training)
- Camera Interface: 2 x CameraLink (SDR connector)
GigE Vision
- Supported cameras: Specim FX10e / Specim FX17e / Specim FX50e
- Operating system: Specim CameraOS
- Operating modes: Real-time processing, data collection
- Real time processing: Using applications created with SpecimINSIGHT
- User interface: CameraOS – webUI
- SoC: NVIDIA Jetson Xavier AGX
- Storage: 1 TB
- Data format: Specim dataset with ENVI compatible files (for training)
- Camera Interface: GigE Vision (Ethernet connector)