Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This brief analyzes the performance of the hardware-based tone mapping operators for compression of high dynamic range images. The bottlenecks of a tone mapping system are determined and a high-performance field programmable gate array (FPGA) implementation of an operator is introduced. The operator utilizes polynomial mapping technique, adaptive to the pixel values; hence preserving high contrast areas. The technique is further optimized for the presented resource-efficient FPGA implementation. We show that the timing optimization does not reduce the image quality, by obtaining high peak signal-to-noise ratio of the resulting images. The timing comparison to the similar implementations shows 2.5 times increase in the achieved throughput, irrespective of the hardware platform.
Jan Wienold, Geraldine Cai Ting Quek, Dong Hyun Kim
, ,
, ,