一种新的消失点检测算法ANewVanishingPoint-电子与信息学报.PDFVIP

第28 卷第8 期 电 子 与 信 息 学 报 Vol.28No.8 2006 年8 月 Journal of Electronics Information Technology Aug.2006 一种新的消失点检测算法 陈付幸 王润生 ( 国防科学技术大学ATR 国家重点实验室 长沙 410073) 摘 要 该文在随机抽样一致性算法基础上，提出了一个基于预检验的随机抽样一致性Preview model Parameters Evaluation RANSAC(PERANSAC)消失点估计算法：该算法在原始RANSAC算法消失点检验前，加入一个预检验步 骤，在保证计算结果精度不变的前提下，过滤掉大量偏差较大的消失点，减少了检验的计算量，大大提高了算法 的整体效率。大量的实验结果表明，该算法的计算精度与RANSAC算法精度保持一致，计算速度远高于RANSAC 算法。 关键词 消失点, Robust, 随机抽样一致性算法, 预检验的随机抽样一致性算法 中图分类号：TP391 文献标识码：A 文章编号：1009-5896(2006)08-1458-05 A New Vanishing Point Detecting Algorithm Chen Fu-xing Wang Run-sheng (National Key Laboratory of ATR , National University of Defence Technology, Changsha 410073, China) Abstract Preview model Parameters Evaluation RANSAC algorithm (PERANSAC) is given in vanishing point detecting. A preview model parameters evaluation selection is added in the RANSAC algorithm. With guaranteeing the same confidence of the solution as RANSAC, a very large number of erroneous vanishing point obtained from contaminated samples are discarded in the preview evaluation selection. The time of evaluating the quality of the vanishing point is reduced. RANSAC efficiency is significantly improved. PERANSAC algorithm is evaluated on real-world images, a significant increase in speed is shown and the solutions are same as RANSAC. Key words Vanishing point, Robust, RANSAC, PERANSAC [7-9] [10] 1 引言 例如高斯球变换 ，Ho