Method of Automatic Identification of Rifling Mark Based on Similarity
- 715 Downloads
This paper provides an effective method for rifling marks identification based on similarity. Here, secondary edges of rifling marks are the object of study. In detail, first, we extract feature of each curve, all of which compose a cross section, by the method of quadratic curve fitting. Second, the similarity of every two secondary edges from different two bullets is computed, through the improved Euclidean distance calculation. Third, these similarities are compared by program in MATLAB. Finally, doing two experiments, it is shown that if their similarity is much lower, less than 0.4, they are from the same gun, otherwise, they are from different guns. So it is proved this method can correctly identify the rifling marks of bullets whether from the same gun.
KeywordsSimilarity rifling mark automatic identification
Unable to display preview. Download preview PDF.
- 1.Zheng, Q., et al.: Measurement Error Analysis and Data Processing, pp. 168–170. Press of Beijing University of Aeronautics and Astronautics, Beijing (2007)Google Scholar
- 2.He, X.: Modern Statistical methods and Application, pp. 263–267. China Renmin University Press, Beijing (2007)Google Scholar
- 3.Wang, M.: MATLAB and Scientific Computing. Electronic Industry Press, Beijing (2003)Google Scholar
- 5.Zheng, X., et al.: Nonlinear Optimization, pp. 144–145. National Defense University Press, Changsha (2003)Google Scholar