Reconstruction of broken handwritten digits based on structural morphological features

Abstract In this paper, a new method of reconstructing broken handwritten digits is developed. The conditional dilation algorithm is used to bridge small gaps. Spurious segments introduced during extraction of digit fields are detected and deleted based on the morphological structural analyses of digit fields. A set of structural points of digits are detected along the outer contours of digits. The preselected broken points of the digits are determined based on the minimum distance between two structural points. The correction rules of the preselected broken points are based on the structural morphological analyses and the stroke extension. The reconstruction and recognition of handwritten internally broken digits are also discussed. Experimental results are given showing the effectiveness of the method.

[1]  Theodosios Pavlidis,et al.  A shape analysis model with applications to a character recognition system , 1992, [1992] Proceedings IEEE Workshop on Applications of Computer Vision.

[2]  B. Ripley,et al.  Pattern Recognition , 1968, Nature.

[3]  Hong Yan,et al.  An efficient algorithm for smoothing binary image contours , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[4]  Hong Yan,et al.  Handwritten digit recognition using an optimized nearest neighbor classifier , 1994, Pattern Recognit. Lett..

[5]  Hong Yan,et al.  Separation of single-touching handwritten numeral strings based on structural features , 1998, Pattern Recognit..

[6]  Theodosios Pavlidis,et al.  A Shape Analysis Model with Applications to a Character Recognition System , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Hong Yan,et al.  Linking broken character borders with variable sized masks to improve recognition , 1996, Pattern Recognit..

[8]  Hong Yan,et al.  Structural primitive extraction and coding for handwritten numeral recognition , 1998, Pattern Recognit..

[9]  Donggang Yu,et al.  An efficient algorithm for smoothing, linearization and detection of structural feature points of binary image contours , 1997, Pattern Recognit..

[10]  Hong Yan,et al.  Prototype optimization for nearest neighbor classifiers using a two-layer perceptron , 1993, Pattern Recognit..

[11]  Ching Y. Suen,et al.  A fast parallel algorithm for thinning digital patterns , 1984, CACM.

[12]  Hong Yan,et al.  Separation of single- and double-touching handwritten numeral strings based on structural analysis , 1999 .