|
|||||
|
|
| 多特征多阈值级联AdaBoost行人检测器 | |
| 引用本文： | 崔华, 张骁, 郭璐, 袁超, 薛世焦, 宋焕生. 多特征多阈值级联AdaBoost行人检测器[J]. 交通运输工程学报, 2015, 15(2): 109-117. doi: 10.19818/j.cnki.1671-1637.2015.02.012 |
| 作者姓名： | 崔华 张骁 郭璐 袁超 薛世焦 宋焕生 |
| 作者单位： | 长安大学 信息工程学院, 陕西 西安 710064 |
| 摘 要： |  为了满足更快、更准、更鲁棒的行人检测需求, 考虑交通监控视频图像质量不高与局部特征不明显的缺点, 采用简单的行人特征来实现行人检测。除矩形度、高宽比、轮廓复杂度、宽度比、行人面积特征外, 特定选用了对遮挡等干扰具有强鲁棒性的头部圆形度这一简单的局部特征。考虑交通监控视频图像中行人的尺寸变化, 引入区域划分策略划分图像区域。兼顾高检测率和低误检率, 根据分类误差最小原则与正样本分类率最大原则训练多个单特征多阈值AdaBoost行人检测器。为了优化多个行人检测器级联后的检测性能, 在兼顾检测性能和检测速度的基础上, 定义了以贡献率作为行人检测器的级联规则, 依据贡献率大小确定的级联次序为基于高宽比、宽度比、矩形度、行人面积、轮廓复杂度和头部圆形度的行人检测器, 依次进行级联, 建立了新的多特征多阈值级联AdaBoost行人检测器。选用3个交通场景对行人检测器进行测试, 并与单级AdaBoost行人检测器与现有2种级联AdaBoost行人检测器进行比较。分析结果表明: 在3个交通场景的检测中, 相比其他几种行人检测器, 多特征多阈值级联AdaBoost行人检测器具有较高检测率、较快的检测速度和较低误检率, 检测率最低为96.70%, 误检率最高为0.67%, 检测时间小于5s, 满足交通场景中对行人检测实时性和可靠性的要求。 |
| 关 键 词： | 交通图像处理 行人检测 特征提取 AdaBoost分类器 区域划分 级联规则 |
| 收稿时间： | 2014-11-13 |
Cascade AdaBoost pedestrian detector with multi-features and multi-thresholds |
|
| CUI Hua, ZHANG Xiao, GUO Lu, YUAN Chao, XUE Shi-jiao, SONG Huan-sheng. Cascade AdaBoost pedestrian detector with multi-features and multi-thresholds[J]. Journal of Traffic and Transportation Engineering, 2015, 15(2): 109-117. doi: 10.19818/j.cnki.1671-1637.2015.02.012 | |
| Authors: | CUI Hua ZHANG Xiao GUO Lu YUAN Chao XUE Shi-jiao SONG Huan-sheng |
| Affiliation: | School of Information Engineering, Chang'an University, Xi'an 710064, Shaanxi, China |
| Abstract: | In order to meet the practical demand for pedestrian detection with high speed, high accuracy and strong robustness, in view of the poor quality and unapparent local image features of traffic videos, some simple pedestrian features were chosen for pedestrian detection.Besides rectangle degree, ratio of height to width, shape complexity, normalized width, and pedestrian area, head density was applied because it is a simple local feature and has strong robustness for occlusion interference.Considering the size changing of pedestrian in the image, region division strategy was introduced into image region division.An improved training algorithm based on the minimum principle of classification error and the maximum principle of positive sample classification rate was implemented by considering both high detection rate and low false detection rate, thus several single-feature AdaBoost pedestrian detectors with multi-thresholds were obtained.To optimize the detection performance of cascade pedestrian detectors, the cascade rulewas obtained in term of the contribution rate.The contribution rate was defined by weighing detection performance and detection speed.The cascade order was the detectors based on ratio of height to width, normalized width, rectangle degree, pedestrian area, shape complexity and head density.The pedestrian detectors were sequentially cascaded according to the cascade order, thus a cascade AdaBoost pedestrian detector with multi-features and multi-thresholds was constructed.The proposed pedestrian detector was tested by using 3traffic scenes, and compared with singlecascade-level AdaBoost pedestrian detector and 2existed cascade AdaBoost pedestrian detectors.Analysis result indicates that in 3traffic scenes, compared with the other pedestrain detectors, the proposed pedestrain detector has higher detection rate, higher detection speed and lower false detection rate, the minimum detection rate is 96.70%, the maximum false detection rate is0.67%, and the detection time is less than 5s.So the detector satisfies the real-time and reliable requirements of pedestrian detection in traffic scene. |
| Keywords: | traffic image processing pedestrian detection feature extraction AdaBoost classifier region division cascade rule |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |
| 点击此处可从《交通运输工程学报》浏览原始摘要信息 | |
| 点击此处可从《交通运输工程学报》下载全文 | |