Table 2 Performance comparison of the pork color scoring models based on different image feature parameter

From: Rapid construction method for a precision pork color scoring model based on standard color board images

Models

Function_Pre

R2

RMSE

Function_Cal

CS_1_ L*

CS = −0.175 L + 12.824

0.97

0.26

CS = −0.175 L + 11.332

CS_2_ L*

CS = −0.207 L + 12.654

0.97

0.16

CS = −0.207 L + 12.764

CS_3_ L*

CS = −0.132 L + 9.613

0.50

0.34

CS = −0.132 L + 9.407

CS_1_ L*a*

CS = −0.145 L + 0.037a + 10.213

0.97

0.24

CS = −0.145 L + 0.037a + 9.590

CS_2_ L*a*

CS = −0.156 L + 0.101a + 9.368

0.99

0.13

CS = −0.156 L + 0.101a + 9.392

CS_3_ L*a*

CS = −0.113 L + 0.131a + 7.156

0.79

0.26

CS = −0.113 L + 0.131a + 7.144

CS_1_ L*a*b*

CS = −0.086 L + 0.165a-0.273b + 8.742

0.96

0.26

CS = −0.086 L + 0.165a-0.273b + 7.868

CS_2_ L*a*b*

CS = −0.136 L + 0.136a-0.079b + 8.774

0.99

0.13

CS = −0.136 L + 0.136a-0.079b + 8.785

CS_3_ L*a*b*

CS = −0.106 L + 0.142a-0.084b + 7.439

0.81

0.25

CS = −0.106 L + 0.142a-0.084b + 7.419

  1. The 95% confidence intervals for the model regression coefficients are provided in Table S1.
  2. R2 the fitting degree of the Function_Pre., RMSE root mean squared error of the Function_Pre, Function_Pre pork color scoring equation fitted by ridge regression method using different image feature parameters, Function_Cal calibrated pork color scoring equation fitted by ridge regression method using different image feature parameters; Models were calibrated by mixed pig herd.
Back to article page