|
| CLASSIFICATION AND RECOGNITION OF MINERAL FLOTATION FROTH IMAGES BASED ON PRINCIPAL COMPONENT ANALYSIS |
|
Received:March 05, 2005
|
| View Full Text View/Add Comment Download reader |
| DOI： |
| KeyWord:Flotation; Froth images; Principal component analysis; BP neural network |
|
|
| Hits: 3070 |
| Download times: 18 |
| Abstract: |
| Flotation froth images belong to a special kind of texture images. But its classification is not very ideal through using neural network analysis of the feature parameters obtained by neighboring gray level dependence matrix approach and other similar approaches owing to the different rank of froth images is so similar. Better classification could be reached with the use of principal component analysis approach to transform the original feature parameters into new feature parameters favorable to BP neural network for the implementation of the classification of flotation froth images. Comparison tests showed that the classification based on the transformed feature parameters is much accurate than that on the original feature parameters. |
| Close |