A Fast Thinning Algorithm of Square Hmong Character Handwriting Using Template Matching Mechanism

According to the similar features of word-information principles between Chinese and Square Hmong character, this manuscript proposes a fast thinning algorithm based on template matching mechanism to obtain the skeleton of the square Hmong character handwriting accurately. First, the index templates are optimized in this presented algorithm by analysing the relative positions between the centre pixel and neighbouring pixels and the information of the handwriting nodes. Then, the initial shinning operation is implemented based on considering the stroke trend with the refinement of square Hmong character handwriting binary image. Furthermore, a secondary shinning operation is used to deal with short-stemmed and non-single-pixel problems which could be appealed after refinement. The experimental results show that the proposed algorithm owns the ability to accurately define the framework of the square Hmong character handwriting skeleton on the basis of effectively reducing the complexity of the algorithm, and provide effective pre-processing for the character recognition of Square Hmong Character Handwriting.