A comparative study on point cloud down-sampling strategies for deep learning-based crop organ segmentation

Table 6 Quantitative comparison of both semantic segmentation and instance segmentation performances across five down-sampling strategies on the PSegNet network

PSegNet	Semantic Segmentation (%)				Instance Segmentation (%)				Score	AveDiff (%)
PSegNet	Precision	Recall	F1	IoU	Cov	WCov	mPrec	mRec	Score	AveDiff (%)
FPS	90.03	89.38	89.68	81.90	79.63	85.50	84.30	78.18	7	2.12
RS	91.71	91.04	91.35	84.62	78.79	85.67	83.54	72.80	0	2.01
UVS	93.10	92.62	92.67	86.78	80.12	86.77	85.37	74.99	35	0.39
VFPS	92.26	92.14	92.18	85.99	78.21	85.67	85.05	72.27	15	1.48
3DEPS	92.24	91.86	92.04	85.74	79.03	85.92	84.37	73.48	15	1.36

ISSN: 1746-4811