Band 20, Heft 3

A generalized Bayesian inference nets model (GBINM) is proposed to aid researchers to construct Bayesian inference nets for various applications. The benefit of such a model is well demonstrated by applying GBINM in constructing a hierarchical Bayesian fuzzy inference nets (HBFIN) to diagnose five important types of cardiovascular diseases (CVD). The patients' medical records with doctors' confirmed diagnostic results obtained from two hospitals in China are used to design and verify HBFIN. Bayesian theorem is used to calculate the propagation of probability and address the uncertainties involved in each sequential stage of inference nets to deduce the disease(s). The validity and effectiveness of proposed approach is witnessed clearly from testing results obtained.

In this paper, a novel approach for word extraction and character segmentation from the handwritten Bangla document images is reported. At first, a modified Run Length Smoothing Algorithm (RLSA), called Spiral Run Length Smearing Algorithm (SRLSA), is applied for the extraction of words from the text lines of unconstrained handwritten Bangla document images. This technique has helped to overcome some of the drawbacks of standard horizontal and vertical RLSA techniques. SRLSA technique has been applied on the Bangla handwritten document image database CMATERdb1.1.1 and the success rate of the word extraction is found to be 86.01%. In the second part of the work, we have presented a useful solution to the problem on how best word images of handwritten Bangla script can be segmented into constituent characters. Moreover, the technique can segment the words having discontinuity in Matra, a prominent feature of Bangla script. It also optimizes the trade-off between under/over segmentation as Matra region and segmentation points are estimated more precisely. As a result, better word segmentation accuracy is achieved with minimal data loss. Here, a success rate of 92.48% is observed on a dataset of 750 handwritten Bangla words which is 3.35% higher than that of our earlier techniques.

In the present study, ordinary kriging and support vector machine (SVM) have been used to develop three dimensional site characterization model of an alluvial site based on standard penetration test (SPT) results. The SVM is a novel type of learning machine based on statistical learning theory, uses regression technique by introducing ε -insensitive loss function has been adopted. The knowledge of the semivariogram of the SPT values ( N ) is used in the ordinary kriging method to predict the N values at any point in the subsurface of the site where field measurements are not available. The comparison between the SVM and ordinary kriging model demonstrates that the SVM model is superior to ordinary kriging model in predicting N values in the site.

Defect detection is a major concern in quality control of various products in industries. This paper presents two different machine-vision based methods for detecting defects on periodically patterned textures. In the first method, input defective image is split into several blocks of size same as the size of the periodic unit of the image and chi-square histogram distances of each periodic block with respect to itself and all other periodic blocks are calculated to get a dissimilarity matrix. This dissimilarity matrix is subjected to Ward's hierarchical clustering to automatically identify defective and defect-free blocks. The second method of defect detection is based on Universal Quality Index which is a measure of loss of correlation, luminance distortion and contrast distortion between any two signals. Quality indices of a periodic block with respect to itself and all other periodic blocks are calculated to get a similarity matrix containing quality indices. Specific variances of the periodic blocks are derived from the quality index matrix through orthogonal factor model based on eigen decomposition. These variances are subjected to Ward's hierarchical clustering to automatically identify defective and defect-free blocks. Results of experiments on real fabric images with defects show that the defect detection methods based on chi-square histogram distance and universal quality index yield a success rate of 98.6% and 97.8% respectively.