Browsing by Author "Mery, D"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    A real time visual sensor for supervision of flotation cells
    (PERGAMON-ELSEVIER SCIENCE LTD, 1998) Cipriano, A; Guarini, M; Vidal, R; Soto, A; Sepulveda, C; Mery, D; Briseno, H
    This paper describes an expert system for the supervision of flotation plants based on ACEFLOT, a real time analyzer of the characteristics of the froth that is formed on.:the surface of flotation cells. The ACEFLOT analyzer is based on image processing and measures several physical variables of the froth, including colorimetric, geometric and dynamic information. On the other hand, the expert system detects abnormal operation states and suggests corrective actions, supporting operators on the supervision and control of the flotation plant. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.
  • Thumbnail Image
    Item
    Classification of potato chips using pattern recognition
    (WILEY, 2004) Pedreschi, F; Mery, D; Mendoza, F; Aguilera, JM
    An approach to classify potato chips using pattern recognition from color digital images consists of 5 steps: (1) image acquisition, (2) preprocessing, (3) segmentation, (4) feature extraction, and (5) classification. Ten chips prepared for each of the following 6 conditions were examined: 2 pretreatments (blanched and unblanched) at 3 temperatures (120 degreesC, 150 degreesC, and 180 degreesC). More than 1500 features were extracted from each of the 60 images. Finally, 11 features were selected according to their classification attributes. Seven different classification cases (for example, classification of the 6 classes or distinction between blanched and unblanched samples) were analyzed using the selected features. Although samples were highly heterogeneous, using a simple classifier and a small number of features, it was possible to obtain a good performance value in all cases: classification of the 6 classes was in the confidence interval between 78% and 89% with a probability of 95%.
  • Thumbnail Image
    Item
    Neural network method for failure detection with skewed class distribution
    (BRITISH INST NON-DESTRUCTIVE TESTING, 2004) Carvajal, K; Chacon, M; Mery, D; Acuna, G
    The automatic detection of flaws through non-destructive testing uses pattern recognition methodology with binary classification. In this problem a decision is made about whether or not an initially segmented hypothetical flaw in an image is in fact of law. Neural classifiers are one among a number of different classifiers used in the recognition of patterns. Unfortunately, in real automatic flaw detection problems there are a reduced number of flaws in comparison with the large number of non-flatus. This seriously limits the application of classification techniques such is artificial neural networks elite to the imbalance between classes. This work presents a new methodology for efficient training with imbalances in classes. The premise of the present work is that if there are sufficient cases of the smaller class, then it is possible to reduce the Size of the larger class by using the correlation between cases of this latter class, with a minimum information loss. It is then possible to create it training set for a neural model that allows good classification. To test this hypothesis a problem of great interest to the automotive industry is used, which is the radioscopic inspection of cast aluminium pieces. The experiments resulted in perfect classification of 22936 hypothetical flaws, of which only 60 were real flat-vs and the rest were false alarms.
  • Thumbnail Image
    Item
    Segmentation of colour food images using a robust algorithm
    (ELSEVIER SCI LTD, 2005) Mery, D; Pedreschi, F
    In this paper, a robust algorithm to segmenting food image from a background is presented using colour images. The proposed method has three steps: (i) computation of a high contrast grey value image from an optimal linear combination of the RGB colour components; (ii) estimation of a global threshold using a statistical approach; and (iii) morphological operation in order to fill the possible holes presented in the segmented binary image. Although the suggested threshold separates the food image from the background very well, the user can modify it in order to achieve better results. The algorithm was implemented in Matlab and tested on 45 images taken in very different conditions. The segmentation performance was assessed by computing the area A(z) under the receiver operation characteristic (ROC) curve. The achieved performance was A(z) = 0.9982. (C) 2004 Elsevier Ltd. All rights reserved.