“多段式”近光光型拐点检测算法研究  

Research on the Detection Algorithm for "Multi-segment" Low-beam Light Inflection Points

在线阅读下载全文

作  者:吴杰 沈琪琪 WU Jie;SHEN Qiqi

机构地区:[1]常州星宇车灯股份有限公司,江苏常州213000 [2]常州信息职业技术学院网络空间安全学院,江苏常州213000

出  处:《科技创新与应用》2024年第32期31-34,39,共5页Technology Innovation and Application

摘  要:汽车近光拐点的检测对于前照灯的配光检测十分重要,为了准确地检测近光拐点的位置,该文提出一种基于点到折线组距离和的汽车近光光型拐点检测算法,尤其适合用于“多段式”近光光型的拐点检测。首先,需要对近光光型图进行预处理,包括平滑滤波、目标区域裁剪和二值化,再利用轮廓跟踪法,从左侧起始点对预处理图像进行边缘点提取得到边缘点和链码,然后将边缘点经过两次筛选得到候选拐点及其组合成的折线组,最后计算所有边缘点到每个折线组的距离和,其最小值所对应的候选拐点即为近光光型的拐点。实验验证,该方法可以很好地检测出“多段式”近光光型拐点以及截止线,且具有旋转不变性,具有很高的实用价值。The detection of the inflection point of the low beam is very important for the light distribution detection of the headlight.In order to accurately detect the position of the low beam inflection point,this paper proposes an inflection point detection algorithm for the low beam light pattern based on the sum of the distance from the point to the broken line group,which is especially suitable for the inflection point detection of the"multi-segment"low beam light pattern.First,it is necessary to pre-process the low-beam light pattern,including smoothing filtering,target area clipping and binarization.Then use the contour tracking method to extract edge points from the pre-processed image from the starting point on the left to obtain edge points and chain codes.Then,the edge points are filtered twice to obtain candidate inflection points and their combined broken line groups.Finally,the sum of the distances from all edge points to each broken line group is calculated.The candidate inflection point corresponding to the minimum value is the inflection point of the low-beam light pattern.Experimental results show that the method in this paper can well detect the inflection point and cut-off line of the"multi-segment"low-beam light pattern,and is rotationally invariant and has high practical value.

关 键 词:近光拐点 配光检测 点到折线组 距离和 多段式 

分 类 号:U463.65[机械工程—车辆工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象