Automatic texture classification in manufactured paper

The automatic classification of manufactured paper must be seen as an integral part of the paper making industry. Currently the human element plays a pivotal role in the quality assessment of manufactured paper. This renders the inspection results unreliable as the human element is susceptible to different moods, social pressures and fatigue among others [1]. The system presented in this thesis replicates the actions of the human element in the quality assessment of manufactured paper and also expresses the subjective judgement for an objective figure of merit. This is achieved through the application of texture analysis in the characterisation of the surface appearance of paper for quality. However, texture analysis techniques individually give unsatisfactory classification performance. This thesis has shown that the use of multiple features from different techniques in combination leads to enhanced classification performance over the use of features from any single method alone. Techniques from computer image analysis that were found useful for characterising the paper surface included the co-occurrence matrices, the grey level run length method, the specific perimeter method and first order statistics. A supervised neural network classifier was used for classification. Confusion matrices and the loss matrices have been used for the first time for interpreting the paper classification results. An intelligent feature selection strategy (intuition) has been found to be a powerful tool in paper classification. Furthermore, a combination of features from different techniques performed better than features from a single technique. A classification performance of 87% has been achieved on two classes of paper, the "good" and "poor" quality paper. The results suggest that classifying paper is a difficult problem. This thesis has examined the suggestion that an automated paper classification system based upon multiple texture based features trained to match the performance of the human visual system.

[1]  Paul Scheunders,et al.  Wavelets for texture analysis, an overview , 1997 .

[2]  Ren C. Luo,et al.  The analysis of natural textures using run length features , 1988 .

[3]  S. Iyengar,et al.  Multi-Sensor Fusion: Fundamentals and Applications With Software , 1997 .

[4]  M. Loew,et al.  Fractal dimension of low-resolution medical images , 1996, Proceedings of 18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[5]  Belur V. Dasarathy,et al.  Image characterizations based on joint gray level-run length distributions , 1991, Pattern Recognit. Lett..

[6]  Benoit B. Mandelbrot,et al.  Fractal Geometry of Nature , 1984 .

[7]  Ronald W. Schafer,et al.  Multilevel thresholding using edge matching , 1988, Comput. Vis. Graph. Image Process..

[8]  T. Southard,et al.  Detection of simulated osteoporosis in maxillae using radiographic texture analysis , 1996, IEEE Transactions on Biomedical Engineering.

[9]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[10]  Ioannis Pitas,et al.  Multimodal decision-level fusion for person authentication , 1999, IEEE Trans. Syst. Man Cybern. Part A.

[11]  Robert M. Hodgson,et al.  Texture Measures for Carpet Wear Assessment , 1988, IEEE Trans. Pattern Anal. Mach. Intell..

[12]  Stanley J. Reeves,et al.  Sequential algorithms for observation selection , 1999, IEEE Trans. Signal Process..

[13]  R. J. Trepanier,et al.  Specific perimeter : a statistic for assessing formation and print quality by image analysis , 1998 .

[14]  Simon Haykin,et al.  Neural networks , 1994 .

[15]  Doyle E. Wilson,et al.  Tissue characterization for beef grading using texture analysis of ultrasonic images , 1993 .

[16]  Robert King,et al.  Textural features corresponding to textural properties , 1989, IEEE Trans. Syst. Man Cybern..

[17]  Robert M. Pap,et al.  Handbook of neural computing applications , 1990 .

[18]  Karl Sammut,et al.  Face image matching using fractal dimension , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[19]  Richard W. Conners,et al.  A Theoretical Comparison of Texture Algorithms , 1980, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  Dennis Gabor,et al.  Theory of communication , 1946 .

[21]  Bidyut Baran Chaudhuri,et al.  Texture Segmentation Using Fractal Dimension , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[22]  Azriel Rosenfeld,et al.  A Comparative Study of Texture Measures for Terrain Classification , 1975, IEEE Transactions on Systems, Man, and Cybernetics.

[23]  Yuxin Liu,et al.  Image feature extraction and segmentation using fractal dimension , 1997, Proceedings of ICICS, 1997 International Conference on Information, Communications and Signal Processing. Theme: Trends in Information Systems Engineering and Wireless Multimedia Communications (Cat..

[24]  R. Day.  Visual Spatial Illusions: A General Explanation , 1972, Science.

[25]  Bidyut Baran Chaudhuri,et al.  An efficient approach to estimate fractal dimension of textural images , 1992, Pattern Recognit..

[26]  I. Introduction.  Nonlinear Operators for Improving Texture Segmentation Based on Features Extracted by Spatial Filtering , 1990 .

[27]  G. MallatS..  A Theory for Multiresolution Signal Decomposition , 1989 .

[28]  Barbara Hubbard,et al.  The World According to Wavelets , 1996 .

[29]  Richard O. Duda,et al.  Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.

[30]  Mike J. Chantler,et al.  The response of texture features to illuminant rotation , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[31]  Jorge Herbert de Lira,et al.  Two-Dimensional Signal and Image Processing , 1989 .

[32]  Ming-Ting Sun,et al.  Computation reduction for discrete cosine transform , 1998, ISCAS '98. Proceedings of the 1998 IEEE International Symposium on Circuits and Systems (Cat. No.98CH36187).

[33]  Roger Woods,et al.  Implementation of the 2D DCT using a Xilinx XC6264 FPGA , 1997, 1997 IEEE Workshop on Signal Processing Systems. SiPS 97 Design and Implementation formerly VLSI Signal Processing.

[34]  S. Mitra,et al.  Analysis of texture images using robust fractal description , 1994, Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation.

[35]  G. Longobardi,et al.  Neural networks for the optical recognition of defects in cloth , 1996 .

[36]  Fabrizio Argenti,et al.  Fast algorithms for texture analysis using co-occurrence matrices , 1990 .

[37]  J. Dubois,et al.  Evaluation Of The Grey-level Co-occurrence Matrix Method For Land-cover Classification Using Spot Imagery , 1990 .

[38]  Geoffrey M. Henebry,et al.  Multi-scale texture in SAR imagery: Landscape dynamics of the Pantanal, Brazil , 1994, Proceedings of IGARSS '94 - 1994 IEEE International Geoscience and Remote Sensing Symposium.

[39]  James M. Keller,et al.  Texture description and segmentation through fractal geometry , 1989, Comput. Vis. Graph. Image Process..

[40]  R. Kohler.  A segmentation system based on thresholding , 1981 .

[41]  Michael Unser,et al.  Feature extraction and decision procedure for automated inspection of textured materials , 1984, Pattern Recognit. Lett..

[42]  R. J. Trepanier.  Off-line paper formation quality testing and its dependency on forming technology , 1987 .

[43]  Azriel Rosenfeld,et al.  Histogram modification for threshold selection , 1977 .

[44]  F. Ade,et al.  Characterization of textures by ‘Eigenfilters’ , 1983 .

[45]  F. Arduini,et al.  Multifractals and texture classification , 1992 .

[46]  George J. Vachtsevanos,et al.  A comparison of fractal dimension algorithms using synthetic and experimental data , 1999, ISCAS'99. Proceedings of the 1999 IEEE International Symposium on Circuits and Systems VLSI (Cat. No.99CH36349).

[47]  F. Pedersen,et al.  Principal component analysis of dynamic PET and gamma camera images: a methodology to visualize the signals in the presence of large noise , 1993, 1993 IEEE Conference Record Nuclear Science Symposium and Medical Imaging Conference.

[48]  Xiaoou Tang,et al.  Texture information in run-length matrices , 1998, IEEE Trans. Image Process..