BUILDING AND PREPROCESSING OF IMAGE DATA USING INDEXES OF REPRESENTATIVENESS AND CLASSIFICATION APPLIED TO GRANULAR PRODUCT CHARACTERIZATION

The characterization of granular products using image analysis is comp lex, as defining sample size is a very difficult task (should one use weight or number of particles?) and because of the diversity of the da ta which can be extracted from the image. A three-step procedure is ap plied: data extraction, data preprocessing and sample classification. We deal with the second step, once the image data have been extracted and gathered into histograms with a large number of intervals. The met hod we propose allows both the building of optimal size samples and th e creation of data vectors appropriate for the third step. The origina lity of the method lies in the supervision of the data processing by t aking into account the final goal, the discrimination into classes. In dices of stability and discrimination are created to build new histogr ams. To determine the optimal sample size, indices of representativene ss and classification are used. This process has been tested on mill p roduct images which are divided into three classes. The optimal sample size given by the representativeness index is 18 images, whereas it d rops to 13 using the classification index. For this example the featur es, if considered independently, are not informative enough to solve t he problem (the best classification performance is 60%), It is necessa ry to develop a strategy where features are combined. This strategy is presented in a separate paper. (C) 1997 John Wiley & Sons, Ltd.