Table 3 Success rates (%) across different forest maps.

From: Autonomous navigation in unstructured outdoor environments using semantic segmentation guided reinforcement learning

Method

Sensor modality

Map-A

Map-B

Map-C

21

Vision (RGB)

73.3

68.3

70.0

46

Vision (RGB)

85.0

78.3

80.0

24

Vision (RGB)

80.0

75.0

76.7

25

Vision (RGB)

83.3

80.0

81.7

53

Vision + LiDAR

81.7

88.3

76.7

54

LiDAR-only

83.3

86.7

78.3

Our

Vision (RGB) Only

90.0

86.7

83.3

  1. Bold numbers represent the highest success rates (%) for each forest map. For Maps A and C (columns 3 and 5), our proposed method achieved the best success rate. For Map B (column 4), the method cited in [53] (row 6) achieved the best success rate. The last column shows the average success rate across the three maps, with the best value highlighted in bold.