Skip to main content

Table 5 Relationships between the identified and manual counting of 10 cultivars

From: Wheat ear counting using K-means clustering segmentation and convolutional neural network

Sites Data set Samples size Slope Intercept RMSE (ears) R2 RRMSE (%) Bias (ears)
Yuanyang SM159 5 0.61 127.37 12.47 0.98 3.96 3.60
XN20 5 0.81 67.88 6.91 0.96 2.00 3.00
XN511 5 0.92 32.27 8.97 0.99 2.79 8.00
YM11 5 0.71 92.03 7.29 0.98 2.30 0.40
ZM119 5 0.84 36.91 10.14 0.81 3.99 − 4.40
ZM136 5 0.96 0.33 11.67 0.89 3.87 − 10.60
ZM158 5 0.78 74.03 20.35 0.92 8.13 18.80
ZM318 5 0.94 18.50 5.64 0.98 1.91 1.40
ZM32 5 0.70 86.44 20.45 0.95 6.05 − 15.40
ZM36 5 0.60 110.48 10.24 0.97 3.79 2.00
All cultivars 50 0.85 45.09 12.43 0.92 4.15 0.68
  1. ZM119, ZM136, and ZM32 overestimated the number ears