基于数学形态学的灰度图像车牌字符提取算法

在深入研究数学形态学的基础上,提取了一种灰度图像车牌字符提取算法。该算法构造4种结构元素,采用数学形态学的膨胀和腐蚀算子依次求出梯度算子对车牌字符进行边缘检测处理和迭代阈值分割,最后结合数学形态学的区域填充方法弥合字符中的空隙。与传统边缘检测算子相比较,实验结果表明该算法具有较强的提取字符能力和良好的抗噪能力,并保护了字符的边缘。其计算量小,在保证处理效果的同时,保证了处理速度,具有一定的可用性和可行性。

License Plate Character Extraction Algorithm for Gray Image Based on Mathematical Morphology

Through studying the mathematical morphology thoroughly,a novel license plate character extraction algorithm for the gray image is proposed.Through dilation operator and erosion operator of the mathematical morphology,the presented algorithm constructs four structural elements to gain the morphology-gradient operators in turn,and employs them to extract the edges of the license plate characters,then combines the iterative algorithm to segment the characters.At last,this algorithm adopts the area filling method of the mathematical morphology to fill the gaps of the characters.Compared with traditional edge detection operators,the experimental results indicate that the algorithm has very strong ability to extract the characters and reduce noises,and protect the edges of the characters better.It requires fewer calculations,which guarantees the processing effect and processing speed.Therefore,it has very high usability and feasibility.