针对现有车牌识别技术中的车牌图像定位处理问题,提出一种基于多重分形理论和数学形态学的复合边缘检测算法.该算法利用多重分形理论中分形维数特征进行图像二值化处理,再利用数学形态学检测连续的特征边缘.试验结果表明,在抗噪性能方面,该方法比经典的边缘检测算子具有更好的检测效果.%For the problem of vehicle plate image location in the existing plate recognition technology,we presented a compounding algorithm of edge detection based on the multi-fractal theory and mathematical morphology. This algorithm uses fractal dimension features in multi-fractal theory to threshold plate images and uses mathematical morphology to detect edge features. Experiment results showed that this algorithm achieved a better anti-noise per-formance than classical edge detection algorithms.
展开▼