基于SLIC-DPC算法的车辆检测研究  

Research on the vehicle detection based on the SLIC-DPC algorithm

作　　者：阮孟丽 陶兆胜[1] 占伟豪 王丽华[1] 王彪[1] RUAN Meng-li;TAO Zhao-sheng;ZHAN Wei-hao;WANG Li-hua;WANG Biao(School of Mechanical Engineering,Anhui University of Technology,Anhui Maanshan 243032,China)

机构地区：[1]安徽工业大学机械工程学院,安徽马鞍山243032

出　　处：《齐齐哈尔大学学报（自然科学版）》2021年第2期41-45,54,共6页Journal of Qiqihar University(Natural Science Edition)

基　　金：安徽省高校自然科学项目(KJ2016A812)。

摘　　要：针对密度峰值聚类算法存在对噪声敏感,在图像较大时检测时间较长等问题,以停车场车位为检测研究对象提出一种简单线性迭代聚类-密度峰值聚类算法。首先对线性迭代聚类算法中颜色与空间距离值进行分析;接着运用主成分分析算法对图像的特征矩阵进行降维;其次改进线性迭代算法的距离函数以处理像素边界和形状;最后将算法运用于停车场车辆检测,得到最终检测图像。实验结果表明,本文方法在车辆形状保留度、噪声处理等方面效果显著,且运算时间大大降低。Aiming at the problems of noise sensitivity and long detection time in the density peak clustering algorithm,a simple linear iterative clustering density peak clustering algorithm is proposed in this paper.Firstly,the distance between color and space in the linear iterative clustering algorithm is analyzed.Secondly,the principal component analysis algorithm is used to reduce the dimension of the feature matrix.Then,improve the distance function in the linear iterative algorithm,and process the pixel boundary and shape through the distance function.Finally,the algorithm is introduced into the image segmentation and applied to the detection about the vehicle in parking lot to get the final detection image.The experimental results show that the proposed method has significant performance in the vehicle shape retention and the noise processing,and the operation time is greatly reduced.

关 键 词：车辆检测 聚类算法 DPC SLIC SLIC-DPC 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U491.7[自动化与计算机技术—计算机科学与技术]