CAJ | 학술논문

针对当前车底阴影分割算法在复杂环境下鲁棒性较差以及最大类间方差（maximum between-class variance，MBCV）多阈值分割算法不能自动确定阈值个数的问题，提出利用峰值自适应方法自动确定 MBCV 多阈值分割算法中阈值个数；然后，以阈值的个数为粒子群优化算法（particle swarm optimization，PSO）中粒子的维数，提出了一种改进的 PSO-MBCV 算法的车底阴影分割。实验结果表明，该算法能有较低的误分类误差，能有效地分割出车底阴影。
침대당전차저음영분할산법재복잡배경하로봉성교차이급최대류간방차（maximum between-class variance，MBCV）다역치분할산법불능자동학정역치개수적문제，제출이용봉치자괄응방법자동학정 MBCV 다역치분할산법중역치개수；연후，이역치적개수위입자군우화산법（particle swarm optimization，PSO）중입자적유수，제출료일충개진적 PSO-MBCV 산법적차저음영분할。실험결과표명，해산법능유교저적오분류오차，능유효지분할출차저음영。
The current segmentation algorithms of bottom shadow of vehicle have poor robustness,mean-while,the multilevel thresholds segmentation algorithm of maximum between-class variance (MBCV)method does not determine automatically the number of the thresholds.Therefore,firstly,the peak adaptive method based on image histogram is used to determine the number of thresholds;then,the number is considered as the particle dimension of the particle swarm optimization (PSO)algorithm,and the bottom shadow of vehicles based on an improved PSO-MBCV algorithm is proposed.The results show that the misclassification error (ME)can be deduced and the bottom shadow of vehicles can be effectively segmented.