Home
News

Enhancing System Performance with IP Cores

Embedded Advisor | Wednesday, January 27, 2021

Fremont, CA: One of the most significant advantages of integrating IP cores directly into field-programmable gate arrays (FPGA) is the readymade functionality for system integrators and camera manufacturers, which includes image processing functions like contrast enhancement and noise reduction, and the integration of machine vision standards into cameras or image acquisition devices. Also, the use of IP cores leads to saving time and money as it provides an alternative to programming FPGAs, which involves complicated VHDL or Verilog programming languages.

Take the case of JPEG compression. This is one scenario where IP cores are simpler than programming an FPGA. To program an FPGA, one must understand the JPEG algorithm and how to implement it into an FPGA. This requires simulation, verification, and integration into a platform. On the other hand, purchasing a JPEG compression core externally and integrating it into the frame grabber is more straightforward and cheaper. Some of the most commonly used IP cores for image processing include improving the quality of an image, reducing image noise, image classification, and adapting to lighting or motion.

Another added benefit for end-users looking to deploy IP cores is in the data speeds. A system can run up to four CoaXPress 2.0 lanes at 12.5 Gbps per lane. The camera, however, produces more data than this. When all of this information is sent to the frame grabber when the PC is running software, it may not be able to keep up with the data rate. One option is to wait for the data or lower the frame rate. But by keeping the data in the frame grabber or the camera, the system will keep up with the speed of data.

By implementing a custom IP core into the frame grabber, the machine vision system will be able to identify the cells in the image and colorize them based on circularity, making it easier to identify the type of cells. IP cores also add value to laser tracking, whether for defense or manufacturing. In the case of laser tracking, processing must be done as quickly as possible, before the next frame comes in, as it requires immediate adjustments. In the case of additive manufacturing, IP cores provide feedback for closed-loop laser power control. As the laser heats powdered metal into solid metal, the machine vision system counts the number of pixels appearing above a certain threshold. This helps the system to recognize a constant number of pixels in the area heated by the laser.