Combining Gidel’s hardware and software variants enables easy and fast implementation of demanding vision on the edge applications with high-speed robust processing capabilities at less than 20W. Gidel will demo real-time image acquisition, compression, and processing from 100 Megapixel 10GigE camera.
At the Embedded World show in Nuremberg, Germany, held on April 9-11, 2024, Gidel will be demonstrating that high-resolution, high-speed, and/or multi-camera imaging with high-bandwidth camera interfaces like 10GigE can be achieved with compact, low-power embedded systems.
Gidel, a market leader in high-bandwidth imaging solutions and FPGA processing, will presenting its modular embedded vision platforms consisting of the award-winning FantoVision edge computers and Gidel’s unique image-processing and compression libraries for real-time optimization of the image stream on the frame grabber’s FPGA.
FantoVision: Jetson and FPGA computer with high-bandwidth camera interfaces
Thanks to Gidel’s high-bandwidth camera interface technology, FantoVision edge computers allow system developers to design embedded vision applications on an NVIDIA Jetson SoM with high-resolution and/or high-speed cameras. The platform offers various hardware options. Camera interfaces include two 10GigE ports, 4x PoCXP-12, or Camera Link. Users can also choose from different FPGA chips for image pre-processing or various NVIDIA Jetson SoM (Xavier NX, Orin NX 16Gb).
With Gidel’s FantoVision edge computers, engineers now have a turn-key yet modular solution to implement high-bandwidth machine vision camera interfaces on an embedded platform.
FantoVision: Jetson and FPGA computer with high-bandwidth camera interfaces
Thanks to Gidel’s high-bandwidth camera interface technology, FantoVision edge computers allow system developers to design embedded vision applications on an NVIDIA Jetson SoM with high-resolution and/or high-speed cameras. The platform offers various hardware options. Camera interfaces include two 10GigE ports, 4x PoCXP-12, or Camera Link. Users can also choose from different FPGA chips for image pre-processing or various NVIDIA Jetson SoM (Xavier NX, Orin NX 16Gb).
With Gidel’s FantoVision edge computers, engineers now have a turn-key yet modular solution to implement high-bandwidth machine vision camera interfaces on an embedded platform.
Gidel image-processing libraries reduce image data and free up CPU capacity
FantoVision edge computers can achieve maximal performance thanks to their ability to offload image correction and compression tasks to the FPGA built in Gidel’s proprietary frame grabbers. This can be achieved easily with Gidel’s ready-to-use image processing libraries including high dynamic range (HDR), compression, and more.
Image processing and compression tasks can be performed simultaneously and at full speed on the FPGA with virtually no latency. This frees up processing load on the CPU that in turn can be dedicated to the application-specific tasks as AI inference.
New HDR and high-quality compression algorithm
Visitors at Gidel’s Embedded World booth will be able to see the company’s new HDR algorithm that outperforms most existing solutions in terms of speed and precision.
Gidel will also be demonstrating its proprietary Quality+ compression algorithm that allows compression ratios of 1:10 and beyond while maintaining high image quality (high SNR). This innovation eliminates the traditional dilemma of compromising between image quality and compression ratio.
With the award-winning FantoVision edge computer platform and advanced image pre-processing and compression capabilities, Gidel presents to Embedded World visitors a fast and easy solution for demanding imaging applications with challenging size, weight, and power constraints. Developers of mobile high-resolution imaging systems such as aerial imaging and mapping should not miss a visit at Gidel’s booth 2-559.