基于双曲映射变换的低照度道路图像增强方法研究

Research on low illumination road image enhancement method based on hyperbolic mapping transformation

在智能驾驶领域中,针对低照度环境下获取视觉图像中道路环境交通特征信息难的问题,提出一种基于双曲映射变换的低照度道路图像增强方法。采用最大熵法对HSV颜色空间中的V分量进行阈值分区,利用改进的双曲正切S型函数和双曲正割累积分布函数分别增强明暗区域的亮度,采取非线性变换函数LC补偿光照,进而用双尺度均值滤波法增强图像中道路特征信息。通过自建数据集对比验证相关算法,结果表明:本方法增强后的图像平均亮度提高了111.54%、信息熵提高了11.8%,平均梯度提高了159.75%,低照度环境下获取视觉图像中交通特征信息显著提升。

In the field of intelligent driving,aiming at the difficulty of obtaining the traffic characteristics of road environment in visual images under low illumination,a low illumination road image enhancement method based on hyperbolic mapping transformation is proposed.The maximum entropy method is used to divide the threshold value of the V component in the HSV color space.The improved hyperbolic tangent S-type function and hyperbolic secant cumulative distribution function are used to enhance the brightness of the light and dark regions respectively.The nonlinear transformation function LC is used to compensate for the illumination,and then the dual scale mean filtering method is used to enhance the road feature information in the image.By comparing and verifying the relevant algorithm with self-built datasets,the results show that the average brightness of the enhanced image is increased by 111.54%,the information entropy is increased by 11.8%,the average gradient is increased by 159.75%,and the traffic feature information in the visual image obtained in low illumination environment is significantly improved.