作者: Snehamoy Chatterjee , Ashis Bhattacherjee
DOI: 10.1016/J.ENGAPPAI.2010.11.009
关键词: Feature selection 、 Quality monitoring 、 Test data 、 Data mining 、 Image (mathematics) 、 Computer science 、 Genetic algorithm 、 Artificial neural network 、 Iron ore 、 Feature (computer vision) 、 Set (abstract data type) 、 Quality (business)
摘要: Measuring the quality parameters of materials at mines is difficult and a costly job. In this paper, an image analysis-based method proposed efficiently cost effectively that determines material. The features are extracted from samples collected mine modeled using neural networks against actual grade values generated by chemical analysis. dimensions reduced applying genetic algorithm. results showed only 39 out 189 sufficient to model parameter. was tested with testing data set result revealed estimated in good agreement real (R^2=0.77). developed then applied case study iron ore. show image-based algorithm can be alternative for estimating site. effectiveness verified it on limestone deposit performed equally well deposit.