Computer vision algorithms, such as those found in the OpenCV library, are typically both compute-intensive and memory-bandwidth-intensive. When implementing these algorithms on embedded devices, two questions arise: How to achieve real-time performance while staying within the available power budget? How to address the need for programmability? Standard CPUs and GPUs often have to be ruled out for power-efficiency reasons. So Application Specific Processors (ASIPs) become the alternative choice.

This webinar introduces the new Vision Processor Starter Kit from Synopsys, featuring Processor Designer, the industry’s leading ASIP design tool. The Vision Processor Starter Kit gives a head-start for the design of an Embedded Vision ASIP. It features a small RISC core with a C/C++ compiler, a ported OpenCV library, examples on how to build programmable accelerators, the RTL design flow, and the FPGA-based rapid prototyping flow. Using the example of Canny Edge Detection, the webinar demonstrates how the Vision Processor Starter Kit can be used to design an optimized programmable architecture.

Bo Wu is the Technical Marketing Manager in the Systems Group at Synopsys, supporting Processor Designer. Bo came to Synopsys from CoWare as part of the System Level Design Product Team. Prior to CoWare, Bo held different senior engineering positions focusing on system designs for wireless applications at Cadence, AT&T Wireless, and Nortel. Bo has a Ph.D. Degree in Electrical and Computer Engineering from University of Victoria in Canada, and a Master and a Bachelor degree in Electronic Engineering from Tsinghua University in Beijing, China.