基于Hough变换的视觉识别方法在仪表自动检测中的应用研究

Application Research on Visual Identification Approach Based on Hough Transformation in Meter Auto-examination

利用图像形态骨架细化方法和Hough变换对指针式仪表表头读数的识别进行研究。将仪表表头图像预处理后,进行形态骨架细化,然后用Hough变换检测表头图像的所有刻度线和指针。在此基础上提出一个识别表头读数值的算法。实验表明该方法的有效性,在指针式仪表的自动检测中有一定的实用意义。

The image, skeketons thining approach and Hough transformation to research on the identification of pointer meters's result are applied. After the pretreatment of the meters outfit image, the framework is thin, then all the graduation lines and pointers of the meters outfit image are checked. On this basis, a arithmetic approach to identify the result of meters outfit is proposed. The experiment concerned testifies the feasibility of the approach and shows that it has applied significance in the auto-examination of pointer meters.