
GEViCAM GN-3000 GN Series GigE Vision Smart Camera
Offering Wholesale, OEM, USA / International, GSA Government / Educational, Reseller / Retail, & Quantity Prices and Specifications / Datasheets, Installation Guides, and Operational Manuals.
Price Documents
→ Hardware Neural network (CogniMem) built-in
→ 1024 Neurons with 256 Byte per Neuron
The GEViCAM GN-3000 is comprised of Gigabit Ethernet smart cameras with a built-in high performance hardware Neuron chip. They are designed on a common platform and comply with the GigE Vision standard for plug-and-play compatibility as well as a proprietary high performance SDK. GN-3000 uses a 1/3" WVGA (748 x 480) CMOS with global shutter. Even though the smart camera is a stand alone system, the real time data output is available in 10-bit, or 8-bit (MSB) at 27 MHz. The frame rate is 60 fps for full resolution and at faster rates in ROI scanning. The neural network chip, CogniMem* is a radial based or KNN nonlinear recognition engine with 1024 neurons, all connected together for simultaneous parallel access. Each neuron is 256 in length by one byte deep, as if each neuron has 256 dendrites of 8-bit data and all connected to the common data source. Imager output is directly connected to CogniMem and the hardware neural network performs the recognition tasks in 1.4 μs after reception of vector data. The vector data conversion is implemented from the direct video input during V blank (FVAL= low) in the each frame.










