基于机器视觉的汽车远近光灯的自适应切换算法

低续航车辆由于里程所限,多行驶于无中分隔离带的市区和市郊区域。针对该路况下夜间行驶中,驾驶员易出现会车等情况下未按法规正确进行远近光灯切换的现象,提出了一种基于视觉的远近光灯自适应切换算法。该算法利用图像处理技术对有无路灯场景、无路灯会车场景和无路灯且前车距本车较近场景这三种情况进行识别,并给出与场景规则最相符的灯光使用指示。实验表明,该算法能实时有效的给出远近光灯的切换策略,既保障了行车的安全性,也避免了远光灯不必要的开启,节约了汽车的电能。

Adaptive High-Low beam switching algorithm based on machine vision

Due to the limitation of travel distance for some cars, they are often driving in the urban and suburban areas without isolation belt. When driving on the road at night time, the drivers are easy to not switch the light according to laws and regulations properly. To solve this problem, an adaptive light switching algorithm is proposed based on machine vision. The algorithm using image processing technology to identify the presence of street scenes, no street light with passing vehicle scene and no street light with a vehicle ahead in near range. This algorithm is able to identify those three scenarios and give corresponding instructions according to road rules. Experiments show that the algorithm can effectively give strategy for high beam low beam switch, which guarantee the running safety and also avoid unnecessary high beam open to save electric energy.