应用形态学top-hat变换的支票方形印鉴自适应分割

Adaptive algorithm based on morphological top-hat transformation to segment Chinese square seals in bank checks

为了准确分割支票中的方形印鉴图像,为印鉴真伪自动识别提供无失真的二值待验图像和预留图像,提出一种基于top-hat变换的自适应形态学分割算法。根据支票印鉴的色彩信息和拓扑结构特点得到方形印鉴的灰度图像,将印鉴灰度图像分成4个子区域,以便准确分割印鉴中具有不同笔划结构和背景均匀度的印鉴成份。对每个子区域中的字符和印框分别做top-hat变换增强图像细节后,再用Otsu方法二值化。由于top-hat变换中所用结构元素的尺寸对分割结果有很大影响,本文根据每个子区域中印框和字符的结构特点分别迭代地计算适用于它们的最佳尺寸结构元素。分割了350枚不同支票方形印鉴,分割准确率为99.43%。结果表明,该方法可以准确地分割笔划复杂且分布密集的方形支票印鉴图像,有效地减少分割结果中容易出现的笔划粘结和残缺失真,降低了对原始印鉴图像的质量要求。

To segment a binary seal image out of a check image without distortion, an adaptive morpho- logical segmentation algorithm based on a top-hat transformation is proposed to accurately extract a binary Chinese square seal from a bank check. A gray scale square seal is extracted from a coloured bank check according to the colour information. Different Chinese characters have different stroke fea- tures and background evenness. To respectively process each character in the square seal, the seal is divided into four sub-squares. Then,tl^e background across each sub-square of the grey scale seal im- age is smoothed by the top-hat transformation. The size of the structuring element in the top-hat transformation can have a significant influence on the segmentation. So an adaptive method is pro- posed to iteratively estimate the proper size of the structuring element according to the local fore- ground area. In each sub-square, the optimal size of the structuring elements, respectively, for the imprint frame and the character are estimated. With their optimal structuring elements, the character and the imprint frame in each sub-square are filtered by the top-hat transform and binarized by the Ot- su＇s method. The experimental result shows that when 350 different square seals are segmented in bank checks,only 2 segmented seals have distortions. It is proved that the proposed algorithm can cor- rectly segment Chinese characters with intricate and dense strokes in a bank check square seal. Adhe- sion and incompleteness distortions in the segmentation results are reduced, even when the original square seal has poor quality.