Fig. 16From: Large-scale monocular FastSLAM2.0 acceleration on an embedded heterogeneous architectureGlobal processing time and overall speedupBack to article page