Genetic algorithm and mathematical morphology based binarization method for strip steel defect image with non-uniform illumination

• We proposed EOBMM to reduce the non-uniform illumination in strip steel defect image.
• We put forward BMBGA to decide the threshold value of defect image binarization.
• We combined EOBMM and BMBGA to present strip steel defect image binarization method.
• The proposed binarization method is effective and efficiency in real-time environment.

In order to precisely extract the image shape feature for the defect detection and classification, the strip steel image needs to firstly be binarized effectively. In this paper, the intelligent information processing, including mathematical morphology and genetic algorithm, is introduced to the strip steel defect image binarization. In order to eliminate the effect of non-uniform illumination and enhance the detailed information of the strip steel defect image, an enhancement operator based on mathematical morphology (EOBMM) is proposed firstly. And then, the binarization method based on genetic algorithm (BMBGA) is applied to the binarization of the strip steel defect image processed by EOBMM. The experiment results show that our method is effective and efficiency in the strip steel defect image binarization and outperforms the traditional image binarization methods, Otsu and Bernsen.