Optimization of the SLAMBench workload targetting a CPU-GPU-FPGA system (using OpenCL). Mainly focused in Altera FPGA devices.