Skip to main content

Table 2 Accuracy comparison between active contours, rosette tracker, incremental learning, graph segmentation, and DAPD algorithm

From: Developmental normalization of phenomics data generated by high throughput plant phenotyping systems

Method/algorithm Accuracy (%) (A1, A2, A3, A4 and our dataset)
Precision Recall Jaccard Dice
Active contours 42.25 (26.05) 99.66 (26.18) 42.19 (25.66) 59.34 (26.20)
Rosette Tracker 54.29 (18.11) 99.97 (17.79) 54.29 (19.22) 70.37 (15.00)
Incremental learning 89.87 (13.68) 91.94 (2.96) 83.90 (14.44) 89.72 (12.36)
DAPD 93.95 (5.42) 97.03 (6.03) 90.76 (7.36) 94.90 (5.78)
  1. Each cell shows the mean and standard deviation of the 4 metrics (precision, recall, Jaccard, and dice) across the 5 datasets (A1, A2, A3, A4, and our dataset)