Development of a verification platform for intelligent surveillance camera systems

Surveillance cameras account for a large portion of the security devices market and are used in an increasing number of areas. Recently, surveillance cameras have been evolved into high performance intelligent surveillance camera systems using many image processing algorithms including face and object recognition, image reconstruction, super resolution, and tracking. Most image processing algorithms for intelligent cameras require hardware accelerators for real-time processing in an embedded environment. Therefore, a certain type of verification platform is indispensable for efficient control of software and hardware partitions. In addition, the intelligent surveillance camera systems require active processing that uses a network with security server computers. In this paper, we presented a verification platform that makes it easy to verify software and hardware partitioning and integration. A server computer receives input images from client cameras through Ethernet and performs many different analysis works by referring to database images. The presented platform uses the Cortex-A8 32-bit RISC processor and the Xilinx Virtex5-LX330 FPGA.