摘要
与传统的经过图像处理的识别方法不同,利用CCD采集的道路灰度图像,采用抛物线模型作为目标函数去拟合车道线,应用遗传算法对抛物线参数进行优化,通过二进制编码、多点交叉及变异等遗传过程,得到各个参数的最优解,从而识别车道线.试验结果表明,本方法能有效消除复杂环境下噪声的不良影响,准确识别车道线,大大提高了识别的实时性.
Lane line recognition is the most intelligent vehicle research of key technology. Different from the traditional after image processing method of identification, this study obtained each parameter of the optimal solution, thus the recognition of lane line, by using CCD acquisition way gray level image, using parabolic model as a target function to fit lane line, and using genetic algorithm to optimize the parameters of the parabola, through the binary coding, multipoint crossover and mutation genetic process. The test results show that this method can effectively eliminate the noise effects in the complex environment, and accurately identify lane line, greatly improving the real-time recognition.
出处
《湖北工业大学学报》
2013年第1期56-59,共4页
Journal of Hubei University of Technology
关键词
车道识别
图像处理
遗传算法
抛物线拟合
lane recognition
image processing
genetic algorithm
parabolic fitting