Research Article
Feature Extraction with Ordered Mean Values for Content Based Image Classification
Table 27
Confusion matrix for feature extraction by binarization with Niblack’s local threshold technique.
| a | b | c | d | e | f | g | h | i | ← classified as |
| 45 | 3 | 8 | 12 | 1 | 19 | 8 | 3 | 1 | ∣ a = Tribals | 5 | 35 | 12 | 7 | 0 | 10 | 1 | 2 | 28 | ∣ b = Seabeach | 11 | 16 | 28 | 19 | 0 | 3 | 6 | 1 | 16 | ∣ c = gothicstr | 14 | 6 | 20 | 44 | 0 | 6 | 2 | 0 | 8 | ∣ d = Bus | 0 | 0 | 0 | 0 | 100 | 0 | 0 | 0 | 0 | ∣ e = Dinosaur | 15 | 13 | 3 | 5 | 0 | 54 | 0 | 5 | 5 | ∣ f = Elephant | 10 | 0 | 6 | 4 | 0 | 0 | 79 | 1 | 0 | ∣ g = Roses | 5 | 1 | 0 | 0 | 0 | 6 | 0 | 88 | 0 | ∣ h = Horses | 3 | 31 | 14 | 8 | 0 | 5 | 0 | 1 | 38 | ∣ i = Mountains |
|
|