Vision Components has launched two Linux-based, smart machine vision cameras and a COM built around a Xilinx Zynq SoC, each supporting up to 4.2MP video.
Over the last decade, smart cameras for machine vision have been transitioning from DSPs to systems that combine DSPs or FPGAs with ARM or x86 processors running Linux. The latest to join the Linux camp is Ettlingen, Germany based machine vision manufacturer Vision Components, which with its latest “VC Z” cameras has switched from a DSP-based system to a tuxified ARM/FPGA combo. Thanks to the Xilinx Zynq, the company was able to accomplish this with a single system-on-chip. The VC Z is available in a VCSBC nano Z computer-on-module, which also appears to act as the foundation for the new VC nano Z and VC pro Z cameras.
VCSBC nano Z (left), VC nano Z, and VC pro Z
(click image to enlarge)
The company’s VC cameras, which are still available, incorporate freely programmable DSPs combined with the proprietary VCRT operating system. The new VC Z devices switch to the Zynq running Linux, thereby using less board space while becoming more easily programmable, says Vision Components.
The VC Z cameras also said to “provide a considerable speed boost, if required,” says Vision Components. Since the FPGA can now be used for image processing, “this process can be executed up to ten times faster than without FPGA support,” says the company. The Linux-based VC Z units are said to “independently execute all image processing routines.”
It is not clear which of the Xilinx Zynq-7000 series SoCs is used here, but all the models, like this one, are based on dual 866MHz Cortex-A9 cores running Linux. The ARM cores are tightly linked with varying levels of FPGA circuitry.
The VCSBC nano Z COM and the two cameras are all available with the Zynq SoC and five different CMOS sensors with global shutter and resolutions of up to 4.2-megapixels. Each offers a battery-backed RTC, trigger and flash trigger outputs, and up to 12 programmable inputs and outputs, depending on the model. Further details are as follows:
- VCSBC nano Z — This 65 x 40mm module incorporates a gigabit Ethernet controller, as well as 12 programmable I/Os, an I2C interface, and both an opto-decoupled trigger input and a flash trigger output. The COM is available with one or two remote image sensor boards.
- VC nano Z — The 80 x 45 x 30mm nano Z camera offers a 10/100Mbps port, as well as two inputs, four outputs, a trigger input, and a flash output. It supports existing VC nano connectors and cables.
- VC pro Z — The higher-end pro Z is also larger than the nano Z, measuring 90 x 58 x 36mm. It features a gigabit Ethernet port, as well a serial interface, an encoder, and dual external lighting interfaces. There are also four inputs, four outputs, a trigger input, and a flash trigger output. The camera has an M12 connector, and offers options including a lens, integrated lighting system, and an autofocus module.
Vision Components is now offering the “recently extended” VC Lib image processing library free of charge, “without any additional costs for runtime licenses or other fees,” says the company. The library encompasses more than 300 VC Lib functions for intelligent VC cameras. Vision Components also provides a range of consulting services.
“We see the combination of Linux as an operating system and an ARM processor as a future-proof approach that opens up new possibilities for our users,” stated VC CEO and founder Michael Engel. “Our VC Z cameras are now for instance suitable for high-speed or line scan applications that were not easy to implement with former models.”
The new VC Z products appear to be available now, at unstated prices. More information may be found at the VC Z product page. Vision Components will be demonstrating the VC Z series at the VISION conference, Nov. 4-6 in Stuttgart, Germany.