智能车摄像头图像畸变矫正的研究

摘要

The geometric model of smart car camera is built, and the truth of image distortion is discussed. Against the longitudinal distortion and transversal distortion, non-uniform lines collection and linear compensating are proposed, and the conversion of coordinates formula is deduced. The experiment result shows that this image rectifying method is efficient and can supply accurate image information to extract routing parameter for smart car. For the simple theory, high adaptability and ideal effect of this method, it has high-use value on engineering implementation.%建立了智能车摄像头成像的几何模型,并对图像畸变原理进行了分析,针对图像的纵向畸变和横向畸变分别提出了非均匀行采集和线性补偿的矫正方法,并推导出了坐标变换公式.实验证明,这种方法能够对摄像头图像进行有效矫正,为智能车导航路径参数的提取提供了准确的图像信息.该方法原理简单,适应性强,效果理想,在工程上具有较高的实用价值.