Skip to main content

Table 1 Errors generated by chamfer masks

From: Optimum design of chamfer masks using symmetric mean absolute percentage error

Neighborhood

Metric

Local weights

SMAPE (%)

MAE (%)

3Ă—3: (a,b)

Euclidean

\((1,\sqrt {2})\)

3.9566

7.6120

 

Verwer

(0.9481,1.3408)

2.6641

5.4741

 

Butt et al.

(0.9619,1.3604)

2.0174

3.9609

 

Proposed

(0.9612,1.3593)

1.9790

4.0366

  

(99,140)/102.995

1.9791

4.0354

5Ă—5: (a,b,c)

Euclidean

\((1,\sqrt {2},\sqrt {5})\)

1.3557

2.6751

 

Verwer

(0.9801,1.4060,2.2044)

1.0050

2.0304

 

Butt et al.

\((0.9866,\sqrt {2},2.2062)\)

0.6823

1.3582

 

Proposed

\((0.9865,\sqrt {2},2.2059)\)

0.6796

1.3685

  

(805,1154,1800)/816

0.6786

1.3665

7Ă—7: (a,b,c,d,e)

Euclidean

\((1,\sqrt {2},\sqrt {5},\sqrt {10},\sqrt {13})\)

0.6498

1.2913

 

Verwer

(0.9902,1.4092,2.2363,3.1420,3.5960)

0.4924

0.9897

 

Butt et al.

\((0.9935,\sqrt {2},\sqrt {5},3.1419,\sqrt {13})\)

0.3261

0.6543

 

Proposed

\((0.9935,\sqrt {2},2.2220,3.1417,3.5830)\)

0.3261

0.6543

  

(444,632,993,1404,1601)/446.87

0.3268

0.6514