Skip to main content

Table 1 The comparison results of state-of-the-art image segmentation algorithms

From: High throughput automatic muscle image segmentation using parallel framework

Method F1-score (%) Precision (%) Recall (%)
  mean std mean std mean std
ISO [37] 80.50 0.0993 89.88 0.0589 74.29 0.1369
gPb [29] 79.04 0.0780 91.23 0.0515 70.11 0.0962
Proposed method 84.61 0.0134 85.99 0.0035 85.11 0.0181
  1. Our results reported here are computed by setting the score threshold of the selected candidate regions to 0.26 (see Fig. 6). Our proposed method shows significantly higher recall than others. These entries in boldface are means the best results