Improved algorithm for digital segment codes´ flaw detection

According to the problems that there are low detection rate and high false reject rate based on traditional image processing algorithms, this article proposes a method that combines SSDA and character contour searching algorithms to recognize the digital segment codes´ large flaws and extract characters in the VFD image, and by image subtraction, maximum between-cluster variance (Otsu) and mathematical morphology algorithms, we can easily judge if there is any other little flaw in it. The results demonstrate the performance of our proposed algorithm through simulation, and we validate it with many experiments using this new method compared with other existing methods.