收录:
摘要:
In this paper, we analyse the existing C code for the Bayesian deconvolution algorithm, and propose an improved scheme for a CPU platform. The improved algorithm is implemented on an FPGA and functional verification is performed. A complete FPGA hardware acceleration system is then constructed to realize data communication and computation between the FPGA and the upper-level computer. Four-way parallel computation on the FPGA platform is designed and optimized to improve the computational efficiency. The algorithm is implemented through dedicated hardware servers to protect the intellectual property of the software algorithm. © 2023 SPIE.
关键词:
通讯作者信息:
电子邮件地址:
来源 :
ISSN: 0277-786X
年份: 2023
卷: 12645
语种: 英文
归属院系: