Combining Mahalanobis and Jaccard Distance to Overcome Similarity Measurement Constriction on Geometrical Shapes

In this study Jaccard Distance was performed by measuring the asymmetric information on binary variable and the comparison between vectors component. It compared two objects and notified the degree of similarity of these objects. After thorough preprocessing tasks; like translation, rotation, invariance scale content and noise resistance done onto the hand sketch object, Jaccard distance still did not show significance improvement. Hence this paper combined Mahalanobis measure with Jaccard distance to improve the similarity performances. It started with the same pre-processing tasks and feature analysis, shape normalization, shape perfection and followed with binary data conversion. Then each edge of the geometric shape was separated and measured using Jaccard distance. The shapes that passed the threshold value were measured by Mahalanobis distance. The results showed that the similarity percentage had increased from 61% to 84%, thus accrued an improved average of 21.6% difference.

[1]  Chaur-Chin Chen,et al.  Similarity Measurement Between Images , 2005, COMPSAC.

[2]  Jerry M. Mendel,et al.  The Linguistic Weighted Average , 2006, 2006 IEEE International Conference on Fuzzy Systems.

[3]  Yan Zhang,et al.  On the Euclidean distance of images , 2005, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Beiji Zou,et al.  Shape-Based Trademark Retrieval Using Cosine Distance Method , 2008, 2008 Eighth International Conference on Intelligent Systems Design and Applications.

[5]  Hua Li,et al.  Similarity Measurement Based on Trigonometric Function Distance , 2006, 2006 First International Symposium on Pervasive Computing and Applications.

[6]  Jitendra Malik,et al.  Shape matching and object recognition using shape contexts , 2010, 2010 3rd International Conference on Computer Science and Information Technology.

[7]  Nualsawat Hiransakolwong,et al.  Feature Selection Using Euclidean Distance and Cosine Similarity for Intrusion Detection Model , 2009, 2009 First Asian Conference on Intelligent Information and Database Systems.

[8]  Daud Mohamad,et al.  Investigating Jaccard Distance similarity measurement constriction on handwritten pen-based input digit , 2010, 2010 International Conference on Science and Social Research (CSSR 2010).

[9]  Yang Mingqiang,et al.  Shape Matching and Object Recognition Using Chord Contexts , 2008, 2008 International Conference Visualisation.

[10]  H. Nemmour,et al.  New Jaccard-Distance Based Support Vector Machine Kernel for Handwritten Digit Recognition , 2008, 2008 3rd International Conference on Information and Communication Technologies: From Theory to Applications.

[11]  Jun Guo,et al.  Efficient Computation of Mahalanobis Distance in Financial Hand-Written Chinese Character Recognition , 2007, 2007 International Conference on Machine Learning and Cybernetics.